The Staking Plans Book by Tom Whitaker FREE PDF FOR REFERRALS 2019
The Staking Plans Book by Tom Whitaker FREE PDF FOR REFERRALS 2019
1
The
Staking Plans
Book
The Complete Guide to Staking
for Sports Betting Systems
Money Management Methods to
Make More Profit from Winning Strategies with an
An Innovative Research Methodology
By Tom Whitaker
1
Copyright © 2019 Tom Whitaker
All rights reserved. No part of this book may be reproduced or utilized in any form or by any means,
electronic or mechanical, including photocopying, recording or by any information storage or retrieval system,
without permission in writing from the Publisher, except in the case of brief quotations embodied in critical
reviews and certain other non-commercial uses permitted by law. Inquiries should be addressed to the
Permissions Department via [email protected]
ISBN: 9781694390684
www.thestakingbook.webs.com
[email protected]
2
“Uncertainty is simply a lack of information.
When you have enough data, the right answer is obvious.”
3
Chapter List
PART I 22 Coup Master Staking 183
THE RESEARCH METHODOLOGY 23 Labouchere Staking 189
1 Introduction 11 24 Reverse Labouchere Staking 194
2 Comparing Staking Plans 12 25 Bookies Bank Staking 197
3 Analysing Staking Plans 16 26 Phantom Bet Staking 203
27 Bookies Bank Version 2 Staking 207
PART II 28 Bookies Bank Version 2 Staking with
THE STAKING PLANS Phantom Losing Bets 213
4 Level Staking 75 29 Recovery Type 1 Staking 218
5 Percentage Staking 78 30 Recovery Type 2 Staking 224
6 Target Profit Staking 85 31 Recovery Type 3 Staking 229
7 Retirement Staking 93 32 Stop at a Winner Staking 236
8 1326 Staking 102 33 Up-Down Staking 241
9 1234 Staking 108 34 i-TSM (The Staking Machine) 246
10 4321 Staking 113 35 Whitaker Staking 257
11 1234 Winners Staking 118 36 Ranking of Staking Plans 264
12 Fibonacci Staking 125
13 Martingale Staking 130 PART III
14 Parlay Staking 134 THE CONCLUSION
15 Rolling Doubles Staking 141 37 The Best Staking Plans 281
16 Square Root Staking 146 38 Practical Staking and Betting Issues 284
17 Secure Staking 149 39 Betaminic and Big Data Betting 288
18 Kelly Staking 156 40 Staking Plan Analysis Service 294
19 LP28 Staking 162 41 Free Betting Systems Books Offer 295
20 D’alembert Staking 169 42 Conclusion: The Final Word 297
21 Professional Staking 174
4
Contents
1 Introduction 11
2 Comparing Staking Plans 12
2.1 The Aim of Staking Plans 12
2.1.1 Value and Edge 12
2.1.2 How Bookmakers Calculate Odds 12
2.2 Data Sets 13
2.3 Types of Betting Systems 14
2.4 Betting Banks and Bet Points 15
3 Analysing Staking Plans 16
3.1 Level Stakes 16
3.1.1 Methodology to Compare Level Staking Plans 16
3.1.2 Staking Plan Profiles 16
3.1.3 Monte Carlo Simulations and Profit Ranges 17
3.1.4 Defining Luck 19
3.1.5 Profit Taking and Doubling Your Betting Bank 20
3.1.6 Betting Legs 21
3.1.7 Lowest Troughs and Reducing The Effects of “Luck” 23
3.1.8 Setting Stake Size and Avoiding Bankruptcy 27
3.1.9 Average Number of Bets to Double Bank 30
3.1.10 Varying the Re-Calculation Period 31
3.2 Results for Level Stakes with Betting System Type 3: Evens 32
3.3 Summary for Level Stakes Betting on Type 3 Evens Betting Systems 37
3.4 Applying the Staking Plans to Real Betting Systems 37
Colossus 05 Against the Trend Overs from Betaminic.com
3.4.1 Key Summary Data 42
3.5 Level Stakes Betting on Type 1 Low Odds Systems 43
3.5.1 Results for Level Stakes Profiles 13-19 45
3.5.2 Monte Carlo Simulation Graphs for Level Stakes Profiles 13-19 46
3.5.3 Summary for Level Stakes Betting on Type 1 Low Odds Betting Systems
3.5.4 Applying the Staking Plans to a Real Type 1 Betting Systems 48
Colossus 03 Free Scoring Favourites from Betaminic.com
3.5.5 Comparing Different Betting systems with Different Staking Plans 51
3.6 Level Stakes Betting on Type 2 Odds-On Systems 52
3.6.1 Applying the Staking Plan to a Real Type 2 Betting System 52
Pro 1 Away Day Blues from Betaminic.com
3.7 Level Stakes Betting on Type 4 Over-Evens Systems 55
3.7.1 Applying the Staking Plan to a Real Type 4 Betting System 56
Betaminic 5 Over 2.5 Against the Trend from Betaminic.com
3.8 Level Stakes Betting on Type 5 Medium Odds Systems 60
3.8.1 Applying the Staking Plan to a Real Type 5 Betting System 61
5
Colossus 17 Dog Draw from Betaminic.com
3.9 Level Stakes Betting on Type 6 High Odds Systems 64
3.9.1 Applying the Staking Plan to a Real Type 6 Betting System 64
Pro 8 Underestimated Underdog v2 from Betaminic.com
3.10 Level Stakes Betting on Type 7 Very High Odds Systems 68
3.10.1 Applying the Staking Plan to a Real Type 7 Betting System 68
Horse-Racing Tipster “Boris”
3.11 Level Stakes Betting on Type 8 Mixed Odds Systems 71
3.11.1 Applying the Staking Plan to a Real Type 8 Betting System 73
6 Betaminic Strategies Together
4 Level Staking 75
4.1 Level Staking Plan Summary 75
5 Percentage Staking 78
5.1 Analysis of Percentage Staking 79
5.2 Percentage Staking Plan Summary 83
5.3 Applying to Real Betting Systems 83
5.4 Recommended Staking Plan Leaderboard 84
6 Target Profit Staking 85
6.1 Analysis of Target Profit Staking 87
6.2 Target Profit Staking Plan Summary 90
6.3 Applying to Real Betting Systems 90
6.4 Recommended Staking Plan Leaderboard 92
7 Retirement Staking 93
7.1 Analysis of Retirement Staking 98
7.2 Retirement Staking Plan Summary 100
7.3 Applying to Real Betting Systems 101
8 1326 Staking 102
8.1 Analysis of 1326 Staking 103
8.2 1326 Staking Plan Summary 106
8.3 Applying to Real Betting Systems 107
9 1234 Staking 108
9.1 Analysis of 1234 Staking 109
9.2 1234 Staking Plan Summary 112
9.3 Applying to Real Betting Systems 112
10 4321 Staking 113
10.1 Analysis of 4321 Staking 114
10.2 4321 Staking Plan Summary 116
10.3 Applying to Real Betting Systems 117
11 1234 Winners Staking 118
11.1 Analysis of 1234 Winners Staking 119
11.2 1234 Winners Staking Plan Summary 121
11.3 Applying to Real Betting Systems 122
11.4 1326, 1234, 4321 and 1234W comparison 123
6
12 Fibonacci Staking 125
12.1 Analysis of Fibonacci Staking 126
12.2 Fibonacci Staking Plan Summary 129
12.3 Applying to Real Betting Systems 129
13 Martingale Staking 130
13.1 Analysis of Martingale Staking 131
13.2 Martingale Staking Plan Summary 133
13.3 Applying to Real Betting Systems 133
14 Parlay Staking 134
14.1 Analysis of Parlay Staking 135
14.2 Parlay Staking Plan Summary 139
14.3 Applying to Real Betting Systems 139
15 Rolling Doubles Staking 140
15.1 Analysis of Rolling Doubles Staking 142
15.2 Rolling Doubles Staking Plan Summary 144
15.3 Applying to Real Betting Systems 145
16 Square Root Staking 146
16.1 Analysis of Square Root Staking 147
16.2 Square Root Staking Plan Summary 149
16.3 Applying to Real Betting Systems 149
17 Secure Staking 150
17.1 Analysis of Secure Staking 151
17.2 Secure Staking Plan Summary 152
17.3 Applying to Real Betting Systems 153
17.4 Recommended Staking Plan Leaderboard 155
18 Kelly Staking 156
18.1 Analysis of Kelly Staking 157
18.2 Kelly Staking Plan Summary 160
18.3 Applying to Real Betting Systems 160
19 LP28 Staking 162
19.1 Analysis of LP28 Staking 163
19.2 LP28 Staking Plan Summary 167
19.3 Applying to Real Betting Systems 167
19.4 Recommended Staking Plan Leaderboard 168
20 D’alembert Staking 169
20.1 Analysis of D’alembert Staking 170
20.2 D’alembert Staking Plan Summary 172
20.3 Applying to Real Betting Systems 173
21 Professional Staking 174
21.1 Analysis of Professional Staking 176
21.2 Professional Staking Summary 181
21.3 Applying to Real Betting Systems 182
22 Coup Master Staking 183
7
22.1 Analysis of Coup Master Staking 184
22.2 Coup Master Staking Plan Summary 188
22.3 Applying to Real Betting Systems 188
23 Labouchere Staking 189
23.1 Analysis of Labouchere Staking 190
23.2 Labouchere Staking Plan Summary 192
23.3 Applying to Real Betting Systems 193
24 Reverse Labouchere Staking 194
24.1 Analysis of Reverse Labouchere Staking 195
24.2 Reverse Labouchere Staking Plan Summary 197
24.3 Applying to Real Betting Systems 197
25 Bookies Bank Staking 198
25.1 Analysis of Bookies Bank Staking 199
25.2 Bookies Bank Staking Plan Summary 201
25.3 Applying to Real Betting Systems 202
26 Phantom Bet Staking 203
26.1 Analysis of Phantom Bet Staking 204
26.2 Phantom Bet Staking Plan Summary 206
26.3 Applying to Real Betting Systems 207
27 Bookies Bank Version 2 Staking 208
27.1 Analysis of Bookies Bank Version 2 Staking 209
27.2 Bookies Bank Version 2 Staking Plan Summary 211
27.3 Applying to Real Betting Systems 212
28 Bookies Bank Version 2 Staking with Phantom Losing Bets 213
28.1 Analysis of Bookies Bank Version 2 Staking with Phantom Losing Bets 214
28.2 Bookies Bank Version 2 with Phantom Losing Bets Staking Plan Summary 217
28.3 Applying to Real Betting Systems 217
29 Recovery Type 1 Staking 218
29.1 Analysis of Recovery Type 1 Staking 219
29.2 Recovery Type 1 Staking Plan Summary 223
29.3 Applying to Real Betting Systems 223
30 Recovery Type 2 Staking 224
30.1 Analysis of Recovery Type 2 Staking 225
30.2 Recovery Type 2 Staking Plan Summary 227
30.3 Applying to Real Betting Systems 228
31 Recovery Type 3 Staking 229
31.1 Analysis of Recovery Type 3 Staking 231
31.2 Recovery Type 3 Staking Plan Summary 234
31.3 Applying to Real Betting Systems 235
32 Stop at a Winner Staking 236
32.1 Analysis of Stop at a Winner Staking 237
32.2 Stop at a Winner Staking Plan Summary 239
32.3 Applying to Real Betting Systems 240
8
33 Up-Down Staking 241
33.1 Analysis of Up-Down Staking 242
33.2 Up-Down Staking Plan Summary 244
33.3 Applying to Real Betting Systems 245
34 i-TSM (The Staking Machine) Staking 246
34.1 Analysis of i-TSM Staking 249
34.2 i-TSM Staking Plan Summary 255
34.3 Applying to Real Betting Systems 256
34.4 Recommended Staking Plan Leaderboard 257
35 Whitaker Staking 258
35.1 Analysis of Whitaker Staking 261
35.2 Recommended Staking Plan Leaderboard 263
36 Final Ranking of Staking Plans 264
36.1 Final Ranking of Type 1 Systems: Low Odds Range 1.01-1.49 265
Artificial Data Ranking: AD 1.25 265
Real Betting System Ranking: Colossus 3 (Betaminc) 266
36.2 Final Ranking of Type 2 Systems: Odds-On Range 1.50-1.99 267
Artificial Data Ranking: AD 1.75 267
Real Betting System Ranking: Pro 1 Away Day Blues (Betaminc) 268
36.3 Final Ranking of Type 3 Systems: Evens Range 1.90-2.10 269
Artificial Data Ranking: AD 2.00 269
Real Betting System Ranking: Colossus 05 Against the Trend Overs (Betaminc)
36.4 Final Ranking of Type 4 Systems: Over-Evens Range 2.00-2.50 271
Artificial Data Ranking: AD 2.25 271
Real Betting System Ranking: Betaminic 5 Over 2.5 Against the Trend 272
36.5 Final Ranking of Type 5 Systems: Medium Odds Range 2.50-4.50 273
Artificial Data Ranking: AD 3.50 273
Real Betting System Ranking: Colossus 17 Dog Draw (Betaminc) 274
36.6 Final Ranking of Type 6 Systems: High Odds Range 5.00-10.00 275
Artificial Data Ranking: AD 7.50 275
Real Betting System Ranking: Pro 8 Underestimated Underdog v2 (Betaminc)
36.7 Final Ranking of Type 7 Systems: Very High Odds Range 10.00-30.00 277
Artificial Data Ranking: AD 15.00 277
Real Betting System Ranking: Horse-Racing Tipster “Boris” 278
36.8 Final Ranking of Type 8 Systems: Types 1-6 Mixed 279
Artificial Data Ranking: ADmx 1.25-7.50 279
Real Betting System Ranking: 6 Betaminic Strategies Mixed Together 280
37 The Best Staking Plans 281
37.1 Recommended Staking Plan Table 281
37.2 Staking Plan Overall Ranking Chart 282
38 Practical Staking and Betting Issues 283
38.1 Simultaneous Betting (Placing multiple bets at the same time.) 283
38.1.1 Multiple Banks / Split Banks 283
9
38.1.2 Presume Lost 283
38.1.3 Parallel Betting (Column Betting) 283
38.1.4 Liability Limiting 283
38.1.5 Daily Percentages 284
38.1.6 Re-Calculate Stakes when bank is doubled 284
38.2 Record Keeping 286
38.3 Bank Size 286
38.4 Profit Taking 286
38.5 Betting Bots 286
38.6 Getting The Best Odds 287
38.7 Bookmakers and Exchanges 287
39 Betaminic and Big Data Betting 288
39.1 Three Recommended Betting Systems from Betaminic.com 291
40 Staking Plan Analysis Service 294
41 Free Betting Systems Books Offer 295
42 Conclusion: The Final Word 297
10
1 Introduction
With the advent of the internet, it has become easier to bet anytime, anywhere and
with a wider choice of bookmakers. It has also become easier to get large amounts of
information on many sports and with that bettors have developed many new systems of their
own or been able to follow winning systems through tipping services or other websites. I
myself have used big data to research patterns in sports betting odds and have developed
value spotting strategies that give an edge over the bookmaker’s odds which I wrote about in
my previous books Winning Sports Betting Strategies and Big Data Betting. The next question that
comes to bettors that have a working system is how much to bet and with what staking plan
to use in order to maximize the profits of their systems while managing risk levels to avoid
losing their betting banks.
This is what led me to research these staking plans. I wanted to find the best money
management tactics to maximize potential. There are a variety of staking plans out there and
each staking plan has a number of settings that can be adjusted that make it an almost
overwhelming challenge to analyse and truly understand which is the best plan for a system.
There also comes the question of when to make withdrawals from your betting bank and take
some of the profit out of your system without destabilizing it. In addition, we have to check if
the staking plan will work in the future if our system produces similar results.
This book seeks to compare staking plans and their variations in a new and innovative
way. I do this by reducing the essence of the staking plan to one key question and display that
as numbers which can be ranked and easily understood “How long until this staking plan
doubles my bank without increasing risk?” It is meant to make it easier for people to find
the right staking plan for their system based on their risk level. For ease of analysis, I also
break down betting systems into 7 types based on their odds from low odds favourite systems,
through evens 50-50 systems, up to high odds long shot systems. In this way you can instantly
compare the research and ranking tables to your systems if you know which type you are
using.
Also, I will introduce the tool I use to analyze staking plans, The Staking Machine
(TSM), and show you how to use it for deeper and more accurate research into your own
systems if you want to. I will also show how I used this tool to find the best staking plans for
betting systems researched with Betaminic’s Big Data analysis tool “The Betamin Builder”
and how you can use them yourself.
This book will give you a better understanding of staking plans, rank them in order of
effectiveness and also introduce you to winning betting strategies that you can use them with.
11
2 Comparing Staking Plans
2.1 The aim of staking plans
The aim of a staking plan is to increase the profit of a working system. I define a working
system as one that makes profit over time using simple level stakes. Simple level stakes are
when you bet the same stake on every bet. The first and most important thing to check is
if your system makes profit with simple level stakes. There is no staking plan in existence that
can take a loss making system and turn a profit. If someone does make profit from a loss
making system, it is purely luck and a short term win, in the long run, it will lose. Systems that
make long-term profit on level stakes work because they are betting on value bets where the
bettor has an edge.
Value is when you are betting on odds that are above the true odds. Edge is the advantage
you have when consistently finding value odds because your information or judgement is
superior to the betting market. Some professional bettors say that if you can get an edge of
5% or more over a year, then you are doing well.
In order to beat the bookies at their own game, we need to understand how they make their
odds. Bookmakers decide odds in a variety of ways. The most common being to look at past
data and to make a model of possible outcomes based on that data. The probabilities of those
outcomes are converted to odds. These are deemed the true odds. Then the bookmaker
reduces the true odds by their margin so they can make a profit regardless of the result. Once
the odds are made available, some bookmakers then adjust their odds if many bettors back
one outcome, possibly due to team news in football or ground conditions in horse racing and
so on. These are called crowd-adjusted market odds. Some bookmakers do this by watching
how their “sharp” bettors place bets, their customers who often bet first and bet correctly,
then they adjust their odds based on what the sharp bettors do. Some bookmakers watch
other bookmakers’ odds to make sure they are not higher than them, or not too much higher.
They often look to the big Asian betting markets for signs of movement, since the larger
volumes on those markets from professional gamblers lead to the first market shifts
happening there.
For example, a bookmaker’s model might suggest there is a 50% probability that Arsenal will
beat Chelsea in their next game. That equates to true decimal odds of 2.0. This means that if
this fixture was played 100 times, then 50 times Arsenal would win and 50 times it would be a
draw or a Chelsea win. The bookie then lowers the odds to 1.9 to put their margin in and
make profit. This means that if we backed this kind of bet 100 times we would make a loss
12
since the odds are not fair. If, however, we have a different model where we know that
Chelsea have a European Game on the next Tuesday and they will be resting key players for
that vital game and that in the past such fixtures were often unexpected losses for the away
team that had a Champions League game the following week, and in fact there is a 60%
chance that Arsenal will win, then we perceive the odds to be value odds since the true odds
would be 1.67. We then bet before the team sheet comes out and the crowd-adjusted market
odds get lower. This is an example of finding value and having an edge.
I have not worked at a bookmaker, so I do not know exactly how they make their models, and
I imagine that different bookmakers use slightly different models. But my guess would be that
they have a variety of probabilities based on recent form and other statistics and then they
weight those according to importance and combine them into a complete model. But the key
point from all of this is that the initial or opening odds are based on models. Those
models can be very good, but there are cases where the odds are not correct. It is those cases
where value can be found. If your models or judgement are better, you can have an edge
over the bookmakers. Once you have an edge, you can make long term profit on level
stakes. We then want to know if another staking plan can increase those profits.
First, I will analyse staking plans in seven different odds ranges or “brackets”. It was my
hypothesis which I have proved in my research to be true, that different brackets perform
better with different staking plans. I will analyse staking plans using artificial data sets of
10,000 bets with a constant 5% edge to make a comparable ranking where all odds are at the
mid-point of that range. For example, for Type 1 “Low Odds Systems” with a range
1.01~1.49, I will create a data set where the odds for every bet are 1.25. Then I will also apply
those staking plan settings to real data from actual systems which have edges ranging from 2-
10%. By the end of the book, you will see a ranking of which staking plans work best with
each odds bracket. It will also make it more comparable to your own betting system at a
glance by comparing your average odds to the different types of example systems here.
13
2.3 Types of Betting Systems
For the purposes of comparison I have chosen 7 types of betting systems to analyse. It
doesn’t matter if you are betting on horses, football, tennis, golf, greyhounds and so
on. Sports betting odds are odds. The only important information is what the odds were
and if you won or not. With the odds and win/lose data we can analyse a betting system. Note
that casino odds are not in this. Casino game odds are stacked in favour of the house and on
level stakes over a long period of time, you cannot win at roulette, blackjack or other casino
table games. These staking plans will not help with those kind of games, that is why I focus on
sports betting.
If you bet on a variety of odds ranges and sports, it can be a good idea to separate your results
into odds ranges. This is because the characteristics of the frequency of wins are different
depending on the type. Even if you only bet on horse racing, it might be better to apply a
different staking plan to odds of different ranges. A very simple example of what I mean is
that if you are using a recovery staking plan that seeks to recover a target amount of 100
pounds on the next bet, then high odds would mean a low stake is needed, but low odds
would mean a very large stake would be needed. Such fluctuations in stakes can make it hard
to set minimum and maximum staking levels and also to set stop losses and so on. You may
also want to separate your bets by sport since your edge may vary according to sport, so the
pattern of your system may vary again. But it is more important to vary stake settings by odds
brackets than to separate by sports.
14
2.4 Betting Banks and Bet Points
A betting bank is basically how much you have to bet. It could be 250 pounds, it could be
33,000 pounds. A betting bank is usually divided up into 100 points. In this way, when I talk
about betting 1 point. I basically mean betting 1% of your betting bank. This makes it easier
for people of different betting levels to compare risk. So 1 point of a 250 pounds starting bank
would be £2.50. 1 point of a 33,000 pound bank would be £330. Some people talk about
betting 1 point of a 200 point bank, but in my mind, that is the same as betting 0.5 points
from a 100 point bank. It introduces an unnecessary variable to the research. In this book, all
research is based on 100 point banks. This is one step towards making comparisons easier.
When a staking plan needs a bigger bank than 100 points to stay out of bankruptcy, then I
reduce the stake below 1 point. In some cases of truly awful staking plans, this can mean stake
of 0.01 points, which equates to needing a bank of 10,000 points for 1 point betting just to
stay in the game.
15
3 Analysing Staking Plans
3.1 Level Stakes
Level Stakes is the first and most important of all staking plans to consider. If your betting
system does not make profit at level stakes, then it will not make profit with any other system.
But level stakes is more than just betting the same amount each time. We need to know how
much to bet based on our betting bank. Staking plans seek to maximize the profit of a system,
but should we be betting 1 point per bet from a 100 point betting bank or 5 points or 10
points? Also, how often should we change the bet level? Are we going to bet 5 points on each
bet for eternity? At some point we want to increase out bets, but when and at what rate?
For each variation of a staking system I am going to call it a profile. Each betting system type
will be analysed with different staking plan profiles.
First, let’s look at Betting System Type 3: Evens. If we have a 5% edge, 10,000 randomized
bets of past data and set the average odds at 2.00. Let’s bet 1 point of our 100 points bank on
each bet. (Level Stakes Profile 1) What results will get? We will get 500 points profit after
10,000 bets. That gives us a Return On Investment (ROI) of 5%. We bet 10,000 points in
total and got 500 points profit. That’s 500 points divided by 10,000 points total outlay. ROI is
an important figure to understand because it tells us how much profit we gain on average for
each bet. It tells us how efficiently we are betting. What if we re-adjust our 1 point bet every
100 bets so that we bet 1 point (or 1% basically) of our current betting balance. (Level Stakes
Profile 2) Then, if we have made profit and our betting bank is 120 points, our stakes can
increase at fixed intervals to take advantage of our larger balance, 1.2 points per bet in the case
of a 120 point balance. Conversely, if the balance has gone down to 92 points, then we only
bet 0.92 points on each bet in the next series. What results do we get then?
We get 9,194 points profit. This has increased our profit by over 1,800%. I analyse these
staking patterns with a wonderful piece of software called “The Staking Machine” made by
Dave Morris. The program gives us some very interesting data about our two Level Stakes
Profiles.
16
Largest Smallest Largest
Staking Cumulative Single Single Average Single
Plan Profit ROI Stake Stake Stake Profit
Level Stakes Profile 1 500 5 1 1 1 1
Level Stakes Profile 2 9,194 7.68 73.76 0.97 11.98 73.76
First we can see that Profile 1 made 500 profit and Profile 2 made 9,194 points profit. Profile
2 also had a better ROI of 7.68. So it was using our money more efficiently. Profile 2
increased the stakes every 100 bets to 1% of the betting bank at that point. The highest stake
reached was 73.76. So for a period of 100 bets were betting that much. Interestingly, the
lowest stake was 0.97. This was for the 3rd set of 100 bets. The second set of 100 bets actually
increased the stake to 1.1 point per bet, but went on a bad run leaving the betting balance at
97 points for bet 201, we then made it back to a balance of 105 at the end of that 100 bet run.
Profile 1 never changed its stake, so its average is 1 point per bet, but Profile 2 has an average
of 11.98 points per bet over the 10,000 bet set.
Total Largest Losing Largest Losing
Staking of All Highest Lowest Sequence Peak Sequence Peak
Plan Stakes Bankrupt Peak Trough to Trough in £ to Trough in %
Level Stakes Profile 1 10,000 No 602 90 12 -7.27
Level Stakes Profile 2 119,791 No 9,442 89 369 -3.91
Interestingly, Profile 2 bet a total of 119,791 points compared to Profile 1. That is a much
larger volume of betting. The bankrupt column tells us that the betting bank never went below
zero in this 10,000 bet series. In the above table we can see that the highest peak of the profit
for Profile 1 was 602. Since we finished on 500, that means we have actually come down from
its best profit level. Profile 2 reached 9,442, and then came down to its final total of 9,194.
Profile 2’s lowest profit level was 89 points and that came at bet 220. Profile 1 had a longest
losing run of 12 bets in a row at 1 point stake per bet, leading to a £12 loss. This equated to
7.27% of the betting balance at that time. Profile 2 also experience the same 12 bet losing row,
but since it was using higher stakes it led to a £369 loss. That equated to 3.91% of the total
betting bank at that time.
Importantly, once we start making increases in our stake over time periods, the order of bets
becomes significant. If we had a good run for our first 100 bets, and then a bad run in our
next 100 bets, we end up betting more on losing bets than on winning bets and we could end
up with a lower profit level than if we had bet the same stakes for all the 10,000 bets. How can
we test this? We can do something called a Monte Carlo Simulation. This is where we take
our betting results and shuffle them to see if the outcome would have been different if the
fixture list had been different. If we have a betting system, we can hope for similar results in
the future from similar picks, but we cannot guarantee when those wins will come, so timing
becomes an important variable. If I shuffle the data 1,000 times and see 1,000 different
17
permutations of how the order of wins and losses could have come and thus a variation of
stakes at different times, what range of results would we get? The results come out as
follows:
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 2
The profits range from 6,500 to 9,800. The average profit was 9,195 and the average ROI was
5%. We see that all of the simulations end with a higher profit than Profile 1 and none of
them experienced bankruptcy. We can see from the results that our “actual results” can be
considered lucky, since they are above the average total profit for that staking plan and the
betting data set.
Number of
1 8 36 47 52 69 32 3 2
occurrences
Points profit 6,422 7,152 7,736 8,208 8,247 8,539 8,977 9,488 9,780
*selection of results, not all.
18
3.1.4 Defining Luck
Once we start changing the stake over time, the order of wins and losses becomes important.
“Luck” or the random order of the bet list becomes a factor. What if we re-adjust every 20
bets? (Level Stakes Profile 3).The total profit becomes a lower amount of 8,734 and a higher
ROI of 7.74%.
If we run 1,000 Monte Carlo Simulations, we then see this distribution of profit results.
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 3
19
Average Actual No. of
Staking Total Average Total Actual No. of Bankrupt Bankruptcy
Plan Profit ROI Profit ROI Simulations Simulations Chance %
Level Stakes Pr. 1 500 5 500 5 1,000 0 0
Level Stakes Pr. 2 8,250 5 9,195 8 1,000 0 0
Level Stakes Pr. 3 8,763 5 8,735 8 1,000 0 0
The average profit goes down to 8,763 and gives an ROI of 5%. It seems that reducing the
gaps between re-adjusting the stakes smoothed out the results and our actual result was closer
to the average.
One problem I see here is testing what interval to re-adjust the stake seems like an almost
infinite variable. 5 bets, 10 bets, 20 bets, 50 bets? Also, if we have a good run in the beginning,
then we don’t know how the system reacts in cases of long losing runs that may come later. I
also wonder when would we take out profit in this case? If I take out profit, I reduce the
betting bank and then the stake reduces, too. If I take out profit at the wrong point, I am
destabilizing my system. How can I include the idea of profit taking into this research and
also solve the problem of when to adjust stakes?
For this reason I thought of another way. Instead of increasing the bet every 10 or 100 bets, I
decided to take profit and reset the balance. The purpose of any betting system is to make
money. Doubling your bank is a good milestone to aim for. So I researched what would
happen if each time the 100 point bank doubles to 200 points, we take out 100 points. In this
way we take profit equal to our start bank and we also reset the betting bank so we can see
what will happen if we start betting from scratch at various points during the lifetime of the
betting data. This has the advantage of testing if a staking plan performs consistently over
time in a variety of good and bad runs. It also gives the option of actually withdrawing money
or simply increasing the stake of the system in line with the points balance each time it
doubles itself. For example, if you started with £100 and bet £1 per point, and then if the
points doubled from 100 to 200 which meant you doubled £100 to £200, then instead of
taking out the £100 when the points reset to 100, but let it ride, then you would in effect be
betting £2 on each unit in the next bets instead of £1, because your betting bank would be
£200 now. This means this profit taking methodology allows you to let the bank balance
continue or withdraw at milestones when the bank doubles.
20
3.1.6 Betting Legs
Let’s see what happens to our data when we add profit taking to the staking plan profiles.
Below we can see profit trend graphs before and after profit taking with the actual data set.
You can see the peaks where it reaches 200 points and then has 100 points profit withdrawn
and is reset back to near 100 points. You can also see how during different periods the
progression from 100 points to 200 points is smooth and quick while at other times it is quite
volatile and takes longer to reach its 200 point target. This should give you a good visual image
of good luck and bad luck and how it affects our staking plans. I am going to call each period
of bets taken to double the bank a leg. You can see the end of each leg at the peaks of the
graph when they reach 200 points.
21
Level Stakes Profile 3 Level Stakes Profile 3 with Profit Taking
The right hand graph shows 6 peaks indicating 600 points profit created or 6 times when a
100 point bank would double.
Looking at the cumulative profit column (green shading), we see Level Stakes Profile 2 Profit
Taking ( LSP2-PT) has only 665 points profit instead of 9,194 points in Level Stakes Profile 2.
Please note here that this is actually a similar profit level. It is just that in LSP2 we let the
profits ride, but in LSP2-PT we took out 100 points profit each time the bank doubled. This is
very important to understand when looking at these figures. In fact, in this case LSP2-PT
made more actual point profit than LSP2, this is because the stakes each profile used were
different. LSP2 re-calculated the stake according to the total betting balance so far every
100 bets, but LSP2-PT re-calculated the stake every 100 bets based on that leg’s current
point balance. An interesting thing to note here is that the ROI (blue shading) for LSP2 and
LSP3 come down closer to the actual edge of 5% for LSP2-PT and LSP3-PT. This is because
the constant resetting of the betting points balance to 100 gives us a more balanced outcome
of results and reduces the effect of lucky/unlucky runs at the beginning of a betting series.
My analysis of staking plans is going to be based on this idea of profit taking and
resetting the points balance to 100 each time the bank doubles. So it is important to
understand the merits of this for getting a clear image of the characteristics of each staking
plan. Look at the table below.
22
Comparing LSP2 and LSP2 TP Leg Leg Leg Leg Leg Leg Leg Total
1 2 3 4 5 6 7*
Bet Number in 10,000 series 3,006 4,220 5,206 6,511 8,833 9,768 10,000
LSP2-PT Number of bets taken
to double the bank 3,006 1,214 986 1,305 2,322 935 232
LSP2-PT profit withdrawn each
leg 1-6 + Leg 7 current profit (points) 100 100 100 100 100 100 65 665
LSP2-PT balance if profit not
withdrawn (includes starting bank) £ 200 400 800 1,600 3,200 6,400 10,560 10,560
LSP2-PT stake per leg if profit not
withdrawn £ 1 2 4 8 16 32 64
LSP2 profit at bet number (points) 102 272 645 1,248 2,586 5,953 9,194 9,194
LSP2 balance at bet number
(includes starting bank) £ 202 372 745 1,348 2,686 6,053 9,294 9,294
*Note: Leg 7 is not complete and only 65 points towards its 100 point target.
Look at the green shaded row. This shows the number of bets that were needed to double the
bank. Here you can get a very easy to understand reference for how long it might take to
double your bank with this staking plan. The shortest legs were 935 and 986 bets. The longest
legs were 3,006 and 2,322 bets. The middle ground seems to be about 1,214 or 1,305 bets to
double the bank. The average number of bets needed to double the bank can be found by
dividing the number of bets for 6 legs (9,768) by the number of legs (6), which gives us an
average of 1,628 bets per leg. Looking at the data ourselves, it would seem fair to think that
1,100 bets would be considered normal and the 2,300 and 3,000 were “unlucky” bad runs.
Look at the blue and red shaded rows. These show how the balances would look like if we
had bet with a £100 starting balance with LSP2 PT and we let our profits ride. As you can see
we make more profit from LSP2 PT. Again this is because we balanced out the betting series
into legs that reduce the influence of good and bad runs.
The Lowest Trough column tells us how low our betting balance went. We can see from the
Lowest Trough column (green) that the PT profiles have lower troughs. This information is
far more useful than the non-PT profiles because those non-PT profiles are just telling us the
lowest level at the beginning of our 10,000 bet series. Once we clear a certain level, we never
23
go back down there unless there is an extremely bad run of results. The PT profiles split up
the 10,000 bet run into 7 legs of bet series. Here we have 7 different “starts” in one series and
this lowest trough could be from any of those legs. It gives us a much better idea of how low
the point balance might go if we started using this staking plan with this betting system. From
this we can think about increasing our starting bet from 1% of the betting bank to 2% or even
3%. But each time we do that we will need to do more Monte Carlo Simulations to see if
increasing our staring stake settings would work in a variety of different “luck” scenarios.
Looking at the Largest Losing Sequence Peak to Trough in % column (blue), the numbers are
much higher. Again, this is a better representation of the performance of the staking plan over
time since it shows the worst drop out of all the legs.
A great advantage of this method is that when your bank does double, you can choose to
withdraw any currency amount of profit and then continue with the next leg calculating the
stakes based on the remaining balance or even let it all ride on. You do not have to withdraw
100% of your profit each time. The purpose is to give you a clear idea of how long it will
take this system to double its bank and how low the bank might go. We can also see our
likely total profit range from Monte Carlo Simulations.
The best news from this table is there is 0% of bankruptcy found in 1,000 simulations (green
column). This is a very safe staking plan. If we look at the ROI from the non-PT profiles (red)
we can see that the Monte Carlo Simulations bring the ROI down from 8% to 5%, which is
correct since this is our edge. The PT profiles (blue) actual ROI and Monte Carlo Simulation
ROI are almost the same. This means that the profit taking method of using staking plans
reduces the range of profit result possibilities and reduces the “luck” factor of our long term
profits.
24
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 1 and Level Stakes
Profile 1 PT. Of course, with level stakes betting and no adjustment in stakes, the result is the
same whatever the order of bets.
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 2
25
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 2 PT
We can see here that the range of results in LSP2-PT is much reduced in comparison to LSP2.
LSP2 ranged from 6,422 to 9,780 which is a +/- 20% range from the median 8,101. But
LSP2-PT only ranged from 575 to 718 points profit which is only a +/- 10% range from the
median 646. This again supports the concept that profit taking, even if only on paper to
calculate stakes, is a good method to reduce the impact of luck on betting runs.
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 3
26
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 3 PT
We can see here that the range of results in LSP3 PT is also reduced in comparison to LSP3,
but not as much with Profile 2. LSP3 ranged from 7,850 to 9,450 which is a +/- 7% range
from the median 8,650. But LSP2 PT only ranged from 615 to 690 points profit which is only
a +/- 5.8% range from the median 652. The reason for this is that the stakes were adjusted
every 20 bets which reduced the chances of betting too much on losing runs. This result still
supports the concept that profit taking, even if only on paper to calculate stakes, is a good
method to reduce the impact of luck on betting runs.
We can see that the lowest trough for LSP1-PT was 86. It never went below 86 points in any
of the 7 legs. It only went 14 points under the 100 point balance. Maybe we only need 20
points to work with. That means we could bet 5 points per bet instead of 1 point per bet with
no re-adjustment of stakes. Let’s re-test for 2,3,4, and 5 points. (Level Stakes Profile 4,5,6,7
PT).
27
Largest Smallest Largest
Staking Cumulative Single Single Average Single
Plan Profit ROI Stake Stake Stake Profit
Level Stakes Pr. 1 PT 500 5 1 1 1 1
Level Stakes Pr. 2 PT 665 4.7 1.94 0.8 1.41 1.94
Level Stakes Pr. 3 PT 643 5.02 1.93 0.89 1.28 1.93
Level Stakes Pr. 4 PT 1,000 5 2 2 2 2
Level Stakes Pr. 5 PT 1,500 5 3 3 3 3
Level Stakes Pr. 6 PT 2,000 5 4 4 4 4
Level Stakes Pr. 7 PT 2,500 5 5 5 5 5
Straightaway we can see that Profiles 5, 6 and 7 went bankrupt. (bright red) The Lowest
Trough column tells us by how much. Profile 5 touched a balance of 0, which means its next
bet of 3 points (which it won) would have taken it -3 points into debt. Profile 6 went to -36
points, so its next bet of 4 points would have taken it even further down to -40. And Profile 7
went down to -20 and its next bet of 5 points would have taken it to -25. If this betting bank
was all of your betting money, then it would be a serious problem for it to reach 0. Some
people have the ability to refill their betting banks, but for the purposes of research in this
book, I will consider bankruptcy as discounting it from being a valid staking plan, even if it
dips even slightly under 0. One more check remains with Profile 4 and that is a Monte Carlo
Simulation to see if its lowest trough of 32 points was just “luck” or not.
28
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 4 PT
29
Average Actual No. of
Staking Total Average Total Actual No. of Bankrupt Bankruptcy
Plan Profit ROI Profit ROI Simulations Simulations Chance %
Level Stakes Pr. 1 500 5 500 5 1,000 0 0
Level Stakes Pr. 2 8,250 5 9,195 8 1,000 0 0
Level Stakes Pr. 3 8,763 5 8,735 8 1,000 0 0
Level St. Pr. 1 PT 500 5 500 5 1,000 0 0
Level St. Pr. 2 PT 651 5 665 5 1,000 0 0
Level St. Pr. 3 PT 648 5 643 6 1,000 0 0
Level St. Pr. 4 PT 1,000 5 1,000 5 1,000 45 4.5
But now the bad news, after running 1,000 simulations of shuffling the order of bets, there
were 45 of 1,000 simulations that had bankruptcy and lost the entire bank. That’s a 4.5%
chance of going bankrupt in our next 10,000 bets. Some people might think that 4.5% is a low
risk for the chance of getting 1,000 points profit, which if you had let ride, would have meant
102,400 points profit after 10,000 bets. But it may take a long time to do 10,000 bets and I
would rather put my efforts into something where I only have to worry about keeping my
edge at 5% and not having to worry about my staking plan. All of these staking plans that say
0% chance of bankruptcy still require your actual betting system to keep its edge, so I think it
is best to work with staking plans that do not show any bankruptcies in testing. So I will
discount any staking plan profiles that generate bankrupt Monte Carlo Simulations.
Another way to look at this data is to see the chance of success (non-bankruptcy) as a
percentage and also to convert that chance to odds. For many experienced gamblers, the odds
of 1.05 appear much less of a “sure thing” than the more positive looking 95.5% chance of
success probability. Backing 1.05 odds is like backing 1 goal or more in football games. It
seems like a safe bet, but over time people will get caught out by 0-0 draws. Look at the “No.
of Times bank Doubled” column (yellow). If you compare these figures to the yellow
highlighted part in the previous table’s “Average Total Profit” column for the Profit Taking
staking plans, then you can see the correlation. “The Number of Times Bank Doubled” is
simply the “Average Total Profit” divided by 100. If I divide the total number of bets made by
the “Number of Times Bank Doubled” figure, then we get the “Average No. of Bets to
Double Bank” column results (blue). This is a key figure I will use to rank the staking
plans and the betting systems we apply them to.
30
Ranking by Average Total Profit or Number of Times Bank Doubled would also lead to the
same result. The reason I choose the Average No. of Bets to Double Bank figure is that you
can take this number and immediately apply this to your own betting system if it has a similar
edge and odds range. If you average 100 bets per week, then it can be reasonably expected to
take about 15 weeks with Level Stakes Profile 2 Profit Taking to double your bank with no
risk of the staking plan leading to bankruptcy. (There is, however, still the risk of bankruptcy
from losing your edge.) The more past data you have on your system, be it back testing or
actual results, then the more chance of the staking analysis of being correct. If your edge is
higher, then it will be quicker at doubling the bank.
So currently Profile 2 is our best at level stakes. One more variable we have not exhausted yet
is the period between re-calculating stakes according to the balance. Let’s check stake re-
calculation after 10, 20, 50, 100, 200, 500 and 1,000 bets to find the right balance. Maybe a
pattern will emerge. Here is a summary of the settings we are testing in 12 profiles. From here
on in this book all profiles will be Profit Taking profiles so I will drop the PT tag.
31
3.2 Results for Level Stakes with Betting System Type 3: Evens
Looking at the results we can immediately see a pattern. The higher the re-calculation
interval is, the higher the average profit. However, once the re-calculation interval reaches
1,000 in Profile 12, there is a 0.4% risk of bankruptcy. That is very low, but it is a warning sign
that we are pushing the limits of our bank with that staking plan. Therefore we can see
another pattern. The higher the re-calculation interval is, the more chance of
bankruptcy. Why is this? This is because if we have a good run and our betting bank balance
goes high and we re-calculate stakes to a higher level and then bet for the next 1,000 bets at
that level in a bad run, it could really deplete the betting bank with no re-adjustment of stake
levels even with the bank balance plummeting.
Looking at this next table, we see another pattern emerging in the “Range as % from
Average” column (blue). This shows that as the re-calculation interval increases, the
spread of results or standard deviation from the average profit increases. In the case of
Profile 11 and 12, this leads to some simulations showing a profit lower than Profile 1’s simple
staking plan since they go under 500 points to 495 and 430 respectively (orange). But it is
important to note that the lowest profits of intervals 10-200 (green) are all higher than not re-
adjusting stakes at all in LSP1. Also, the highest possible profit also increases as the re-
calculation interval increases (purple). Profile 8, with an interval of 10 bets between each
stake re-adjustment, has a highest profit of 685, but Profile 11, with an interval of 500 bets,
reaches 780 in one of its simulations. The pattern that has emerged is that there is a chance
32
of higher profits with higher stake re-calculation intervals, but there is a higher risk of
bankruptcy.
Average
Non- No. of Non- No. of Range
Bkrptcy Bets to Bkrptcy Times "Let It Lowest Highest as %
Staking Chance Double Chance Bank Ride" Profit Profit from
Plan as % Bank as Odds Doubled Balance Range Range Average
LSP1(s1-r0) 100 2,000 1.00 5 3,200 500 500 0.00%
LSP8(s1-r10) 100 1,546 1.00 6.47 8,865 615 685 5.49%
LSP3(s1-r20) 100 1,543 1.00 6.48 8,926 614 690 5.91%
LSP9(s1-r50) 100 1,541 1.00 6.49 8,988 598 698 7.45%
LSP2(s1-r100) 100 1,536 1.00 6.51 9,114 568 718 11.28%
LSP10(s1-
100 1,529 1.00 6.54 9,305 515 770 19.44%
r200)
LSP11(s1-
100 1,511 1.00 6.62 9,836 495 780 20.07%
r500)
LSP12(s1-
99.6 1,506 1.01 6.64 9,973 430 790 27.52%
r1000)
LSP4(s2-r0) 95.5 1,000 1.05 10 102,400 1,000 1,000 0.00%
LSP5(s3-r0) 56.4 667 1.77 15 3^6* 1,500 1,500 0.00%
LSP6(s4-r0) 15.7 500 6.37 20 104^6 2,000 2,000 0.00%
LSP7(s5-r0) 1.1 400 90.91 25 3^9 2,500 2,500 0.00%
*NB: Large Numbers are abbreviated. 3^6 = 3,000,000. The ^6 part mean to add 6 zeroes.
Everbody knows that with higher profits there is higher risk. My aim with this staking book is
to remove or reduce the risk from the staking plan. (I cannot remove the risk from betting
systems.) For this reason I am going to discount profiles that have a lowest profit range
that is under the simple level stakes settings of that plan. There is no point using a
staking plan if it might end up with you making less money than doing nothing. A good
staking plan should only increase the profits of a betting system without increasing its risk. I
also see that dips into lowest profit levels that are below the basic profile as a sign that we are
pushing the limits of this staking plan. Staking plans should enhance our profits. Following
this logic, Profile 10 (s1-r200) is now our number one profile for level stakes in Type 3
Evens Betting Systems.
For analysis purposes I include the Monte Carlo Simulation Profit Frequency Graphs for
Profiles 8-12 below.
33
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 8 LSP8(s1-r10)
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 9 LSP9(s1-r50)
34
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 10 LSP10(s1-r200)
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 11 LSP11(s1-r500)
Higher intervals between stake readjustment lead to graphs with more extreme results at either
end.
35
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 12 LSP12(s1-r1000)
This shows the widest profit spread and extreme results. Also, the bell curve becomes more
uneven, heavy on the lower profit side with a steep up-ramp and a more gradual decline.
36
3.3 Summary for Level Sakes Betting on Type 3 Evens Betting Systems
Level Staking Profile 10 (s1-r200) is the best choice for Type 3 Evens Betting Systems.
Average
Non- No. of Non- No. of Range
Bkrptcy Bets to Bkrptcy Times "Let It Lowest Highest as %
Staking Chance Double Chance Bank Ride" Profit Profit from
Plan as % Bank as Odds Doubled Balance Range Range Average
LSP10(s1-
100 1,529 1.00 6.54 9,305 515 770 19.44%
r200)
I have a number of actual working football betting systems created using an enormous
database of over 100,000 past matches and bookmaker odds from 54 leagues since 2012.
These were made with a free online tool called the Betamin Builder from the website
www.betaminic.com. Basically, the tool helps you find patterns of value odds where the
bookmakers are wrong and can then take advantage of that trend. How to use the tool and
analysis of these systems can be found in two of my other books Winning Sports Betting Strategies
and Big Data Betting. One of the strategies is called “Against the Trend Overs” which is
strategy No. 5 from Winning Sports Betting Strategies. You can also view its results and statistics
on the Betaminic site which lists 30 strategies researched by myself and many other publicly
shared betting systems based on big data. One of the reasons I wrote his book is because I
have betting systems already, but I wanted to know the best staking plan to use with each
system. On the Betaminic site, this strategy I will use to test this staking plan is named
“Colossus 05 Against the Trend Overs”. As the name suggests, when certain conditions are
met in specified leagues, there is value to be found in betting on teams that have not been
doing well recently. This betting system fits the Type 3 Evens Betting System model.
37
The above screenshot is from the Betaminic site and gives the current statistics of this system.
It has 2,819 bets since 2012 which averages out to 34 bets per month. The average odds are
1.99 but the win rate is 54%. That gives an edge of 6.77% (which is displayed as yield, but also
can be read as the ROI). The Max DD (maximum drawadown) of -16.11 is the biggest losing
run it has had in its entire existence from a peak to a trough. Over 6 years it has made 190.79
points profit. We can see more details in these charts from TSM below.
From my data, I know the majority of the odds range from 1.8 to 2.2 but there are extremes in
there of 1.52 and 2.72, but it still fits the Type 3 Evens Betting System.
38
This shows the frequency of winning and losing sequences. The longest winning runs were
13,14,15 and 16 which each happened just once. The longest losing runs were 11 and 12
lossses in a row which happened once each.
TSM is even able to calculate the edge for different odds ranges and we can see this strategy
does better with the odds that are over evens with an edge of 8% versus odds on bets which
had an edge of 6%.
39
This graph tells us the strategy’s cummulative drawdown over the course of its 2,819 bets. We
can see how many points negative we went at worst, which never went over -14 points. It is
those deep spikes that can lead to bankruptcy if the stakes were increased too much after a
good run. Let’s see if Profile 10 makes more profit than Profile 1 when applied to an actual
betting strategy without going bankrupt.
Profile 1 Profile 10
From the profit progression graphs we can instantly see that we have two peaks instead of one,
so we have doubled our bank more than twice.
Profile 10 makes 28% more profit than simple level stakes (246 from 191) and has a slightly
better ROI of 6.9% to Profile 1’s 6.77%.
40
Total Largest Losing Largest Losing
Staking of All Highest Lowest Sequence Peak Sequence Peak
Plan Stakes Bnkrpt Peak Trough to Trough in £ to Trough in %
LSP1(s1-r0) Col5 2,819 No 200 89.68 12 -7.35
LSP10(s1-r200)Col5 3,565 No 199 90.78 16.5 -14.78
Monte Carlo Simulations show the actual results of my system were slightly under the average
total profit of 250, but 247 is still pretty close to the average profit. The average profit of the
real data has been increased by 30.89% This is very interesting because when we were
analysing our artificial data, Profile 1 increased the average profit from 500 to 654 which is a
30.80% increase. That is a fantastic figure. It suggests that this staking plan can
transpose its increases to real data to achieve similar levels. We are at the stage where we
can say, use this staking plan to increase your profits by 30.8% on average, and that is exactly
what a staking plan should be. It is a method to increase profits of a working system without
increasing risk.
Monte Carlo Simulation Profit Frequency Graph for Level Stakes Profile 10 LSP10(s1-r200)
applied to “Colossus 05 Against the Trend Overs” from Betamininc.com
41
Average
No. of Non- No. of Range
Bets to Bankruptcy Times "Let It Lowest Highest as %
Staking Double Chance as Bank Ride" Profit Profit from
Plan Bank Odds Doubled Balance Range Range Average
LSP1(s1-r0)
1,476 1.00 1.91 376 191 191 0.00%
Col5
LSP10(s1-
1,128 1.00 2.5 566 192 278 17.41%
r200) Col5
The above table shows us that Profile 10 definitely has a positive effect on the final outcome
with all possible results ending higher than profile 1 (192~278 instead of 191). We also see
that it will take an average 1,128 bets to double my bank. At an average of 34 bets per month,
that will take me 33 months. That number should let you know very clearly how much
discipline a professional sport bettor needs. We still, however, need to test other staking plans
to see if we can get better performance out of this strategy, but if this was the best staking
plan for it, I would use this strategy as part of a larger portfolio of multiple strategies, such a
multi-strategy system is one method of reducing risk by not putting all your eggs in one basket
as it were.
In order to make it easier to compare multiple betting systems and staking plans I have
condensed the key information into this format of table.
Range
Average No. Ave. "Let It Lowest Highest as %
Profile
of Bets to Total Ave. Ride" Profit Profit from
Double Bank Profit ROI Balance Range Range Average
Level Stake Profile 10
1,128 250 7 566 192 278 17.41%
LSP10(s1-r200) Col5
Level Stake Profile 1
1,476 191 7 376 191 191 0.00%
LSP1(s1-r0) Col5
Level Stake Profile 10
1,529 654 5 9,305 515 770 19.44%
LSP10(s1-r200) AD
Level Stake Profile 1
2,000 500 5 3,200 500 500 0.00%
LSP1(s1-r0) AD
*AD = Artificial Data / Col5 = “Colossus 05 Against the Trend Overs” real betting system
42
Ave.
Range
Bets Ave. "Let It Low High Bnkrpt Bets Mnths
Staking Ave. as %
to Total Ride" Profit Profit Chnce per to
Plan ROI from
Dbl Profit Bal. Range Range % Mnth Dble
Ave.
Bank
LSP10(s1- 17.41
1,128 250 7 566 192 278 0 34 33
r200) Col5 %
LSP1(s1-
1,476 191 7 376 191 191 0.00% 0 34 43
r0) Col5
LSP10(s1- 19.44
1,529 654 5 9,305 515 770 0
r200) AD %
LSP1(s1-
2,000 500 5 3,200 500 500 0.00% 0
r0) AD
*AD = Artificial Data / Col5 = “Colossus 05 Against the Trend Overs” real betting system
The average number of bets to double bank column becomes the best way to rank and
compare staking plans and systems with different numbers of bets, yields and characteristics.
Next, I repeat the research for betting systems with low odds, 1.01~1.49, Type 1. I again
create an artificial data set of 10,000 bets, all with odds of a median 1.25 and an edge of 5%.
Then I apply the profit-taking profiles we used with Type 3. After that I look at the patterns
and customize new profiles according to the results.
43
Results
From the results we can immediately see that even using 5% of the bank as an initial stake
does not give us bankruptcy. If we run 1,000 Monte Carlo Simulations on Profile 7 (5%) and 6
(4%) we get bankruptcy on 2 and 1 of them respectively. That is just a 0.2% or 0.1% chance
of bankruptcy, but it is enough to discard profile 6 and 7.
From this table we see no immediate pattern emerge from increasing the interval of re-
calculation. But we do see that there is a definite improvement from having some level of re-
calculation since profiles 13-19 are all higher than profile 5 (green).
45
There is a slight trend that the largest losing sequence in % increases in size as the stake
change interval increases (blue), except for profile 18 (orange)
3.5.2 Monte Carlo Simulation Graphs for Level Stakes Profiles 13-19
46
LSP17(s3-r200) Monte Carlo Simulation LSP18(s3-r500) Monte Carlo Simulation
3.5.4 Summary for Flat Betting on Type 1 Low Odds Betting Systems
Range
Ave. Bets Ave. "Let It Lowest Highest Bnkrpt
Staking Ave. as %
to Double Total Ride" Profit Profit Chance
Plan ROI from
Bank Profit Balance Range Range %
Average
LSP18(s3-r500) 476 2,102 6 212,642,715 1682 2417 17.48% 0.2
LSP17(s3-r200) 478 2,094 5 201,172,267 1824 2274 10.74% 0.1
LSP16(s3-r100) 483 2,069 5 169,165,038 1880 2220 8.22% 0
LSP19(s3-r1000) 485 2,061 5 160,039,878 1695 2463 18.63% 0
LSP15(s3-r50) 488 2,051 6 149,322,486 1893 2175 6.87% 0
LSP14(s3-r20) 491 2,038 5 136,455,584 1948 2128 4.42% 0
LSP13(s3-r10) 492 2,033 5 131,807,406 1968 2097 3.17% 0
LSP5(s3-r0) 667 1,500 5 3,276,800 0 0 0.00% 0
LSP1(s1-r0) 2,000 500 5 3,200 500 500 0.00% 0
In conclusion, Profile 16 is the best for Type 1 low odds betting systems. This means we can
bet 3% of our starting bank and re-calculate the stake based on the betting bank balance after
47
each 100 bets. Its profit range deviates only 8% from the average. Compared to simple level
stakes betting of profile 1, we have increased profits by 413.8%. (500 to 2,069 points)
Once more, I will take an actual working betting strategy created with the Betamin Builder
Tool. This one is called “Colossus 03 Free Scoring Favourites”. It backs teams with high
scoring averages when one is a favourite. It has an edge of 7.67% and a data set of 1,433 bets
to analyse.
Its average odds are 1.31 which puts it firmly in the Type 1 betting system range. It has a win
rate of 82%. It gives about 19 bets a month on average. Over 6 years it has made 109 points
profit.
From the data, I know the majority of the odds range from 1.1 to 1.5 but there are extremes in
there of 1.05 and 1.93, but it still fits the Type 1 Low Odds Betting System.
48
This shows the frequency of winning and losing sequences. The longest winning runs were 30,
31 and 32 which each happened just once. The longest losing runs were 5 in a row which
happened once and 4 in a row which happened twice.
The edge for this strategy is about the same in the two odds bands.
49
This graph shows that the worst runs led to nearly -8 points on two occasions.
Let’s apply the staking plans from profile 1, 5 and 16 to this strategy.
50
Ave.
Range
Bets Ave. "Let It Low High Bnkrpt Bets Mnths
Staking Ave. as %
to Total Ride" Profit Profit Chnce per to
Plan ROI from
Dbl Profit Bal. Range Range % Mnth Dble
Ave.
Bank
LSP16(s3- 30.40
316 454 8 2,326 257 533 0 19 17
r100) Col3 %
LSP5(s3-
434 330 8 985 330 330 0.00% 0 19 23
r0) Col3
LSP1(s1-
1,303 110 8 214 110 110 0.00% 0 19 69
r0) Col3
From the results we can see that Profile 16 applied to an actual Type 1 Low Odds system
leads to similar results. With the artificial data there was a 413.8%. increase in profits and with
real data there is a 412.7% increase. That is a very close correlation. Again we can say that if
we take this profile and apply it real systems of this betting type, we can increase the profits
over 4 times.
When using different betting systems that have different betting volumes, average odds, total
profit and so on, it becomes more and more difficult to compare them. In the table below we
can clearly see which of the two real betting strategies is better.
Ave.
Range Ave. Ave.
Bets Ave. "Let It Low High Bnkrpt
Staking Ave. as % Bets Mnths
to Total Ride" Profit Profit Chnce
Plan ROI from per to
Dbl Profit Bal. Range Range %
Ave. Mnth Dble
Bank
LSP16(s3- 30.40
316 454 8 2,326 257 533 0 19 17
r100) Col3 %
LSP10(s1- 17.41
1,128 250 7 566 192 278 0 34 33
r200) Col5 %
Despite Colossus 3 having fewer bets per month and betting at lower odds, it actually
performs better than Colossus 5. This is because the characteristics of its low odds system
allow a higher percentage of the bank to be bet each time in its staking plan. It then only takes
316 bets on average to double its bank whereas Colossus 5 takes 1,128 bets on average.
Colossus 3 only has about 19 bets per month, but even so, it takes an average of 17
months to double its bank versus 33 months for the Colossus 5. With this kind of data
we can compare unlike betting systems and find out the best way to bet smart. Our
question should always be, how long will it take me to double my bank? Once we know that
answer, we know which staking plan and which betting system to use. Of course, 17 months is
a long time to wait to double your bank, but we will see that time get shorter as we look at
more kinds of staking plans and betting systems.
51
3.6 Level Stakes Betting on Type 2 Odds-On Systems
Now you have seen my methodology and how I arrive at the key data shown in the tables. If I
repeat this level of detailed explanation for all the betting system types and for all the staking
plans, then this book will go over 1,000 pages. So I will just show the results for the other
types.
Profile 10 (green) emerges as the best profile again, which is where we start with 1% of the
bank as the initial stake and re-calculate the stake based on the balance after each 200 bets.
Profile 6, 7, 4 and 11 (red) all had bankruptcies in their Monte Carlo simulations and profile 12
(orange) had a lowest profit level under the basic stakes level of profile 1. Profile 10 gives us a
34.8% increase in profits over simple level stakes.
I will apply this staking plan to another strategy discovered on Betaminic.com. This one is
called “Pro 1 Away Day Blues” and is introduced and analysed further in my book Big Data
Betting. This strategy looks for teams that are on form at home who are coming up against a
team that has been awful away recently. There are 1,982 bets in this data set.
52
This betting strategy data shows most bets, 854 of 1,982 are in the odds-on region, but some
are over evens.
As an odds-on strategy, it is unsurprising to see the winning runs being longer than the losing
runs with 2 occasions having 11 wins in a row while there were also 2 occasions of having 7
losses in a row.
53
The odds broken down by range suggest that 1,400 of the 1,982 bets that were under evens
actually had a range of 1-3% while it was the bets with higher odds that had much higher
yields. This strategy only bets on home wins, and so that means some quite high priced home
odds were being backed. It is very likely that those games were where the away team had a big
reputation and “name”, but recent form suggested they were not as good as their pedigree,
and this strategy is very good at finding those matches where the market has put too much
focus on the reputation of the club rather than the cold facts of recent form. The tips are
based on a database of statistics, so the computer cares nothing for the club stature. That is a
real advantage of big data betting.
54
Applying Profile 10 to this strategy shows an improvement to an average profit of 192 which
is better than on basic level stakes which makes 145 points, but there are 2 of the 1,000
simulations that end up with lower profits than simple level stakes of 119 and 135.
Level Stakes Profile 10 applied to the artificial data led to a 34.8% increase in average profit.
The same staking plan settings applied to this real strategy with its larger range of odds and
actual results led to a 32.41% increase. That is a similar improvement and another sign that we
can take staking plans that are tailored to betting system types and expect similar effects.
This artificial data set has 10,000 bets that all have odds of 2.25 exactly and an edge of 5%.
As the average odds increase, the win rate decreases and the losing runs increase. This
means there is less ability to bet with higher percentages of the betting bank. Here we have to
bet with just 1% of the betting bank. We then looked at different re-adjustment intervals for
the stake and found that re-calculating every 100 bets was the best setting. Profile 10 had 1
result of 499 as its total profit, which put it under the simple level stakes results, and it also
had a range of 19% variation from the average total profit. From this I chose profile 2 as the
better option. It increased profits by 27.4% over simple level stakes.
55
3.7.1 Applying the Staking Plan to a Real Type 4 Betting System
I will apply this staking plan to a strategy created by the Betaminic team themselves on
Betaminic.com. This one is called “Over 2.5 Against the Trend” and is analysed further in
my book Winning Sports Betting Strategies. This strategy is as the name suggests, finding value in
overs bets where teams are not expected to score many. This tends to push the odds over
evens. There are 705 bets in this data set. While the name and concept is similar to the
“Colossus 5 Against the Trend Overs”, the filters used to find match selections are different
and there is no overlap between the systems, so they can be used together.
The yield of 20.17% is amazing and it has a very low drawdown of -12. Unfortunately it only
has an average of 9 bets per month, but it is a good one to add to a set of strategies as part of
a multi-strategy system.
The odds are mainly in the range for Type 4 betting systems with a few bets above and below
the target range.
56
The edge is similar just under or over evens.
With most of the bets being over evens, it would be expected that the losing runs would be
longer than the winning runs, but in this data set the edge is quite high and the longest
winning run of 10 bets is longer than the longest losing run of 8 bets which occurred twice.
57
The cumulative drawdown has a few spikes but never goes over -10 points.
Results
All the results (152~228 average total profit) ended up higher than simple level stakes (143).
58
Ave. Range Ave. Ave.
Ave. "Let It Low High Bnkrpt
Staking Bets Ave. as % Bets Mnths
Total Ride" Profit Profit Chnce
Plan to Dbl ROI from per to
Profit Bal. Range Range %
Bank Ave. Mnth Dble
LSP2(s1-
r100) 365 193 21 381 152 228 19.69% 0 9 41
BtmO2.5
LSP1(s1-
r0) 493 143 21 269 143 143 0.00% 0 9 55
BtmO2.5
The artificial data improved its average total profit by 27.4% and the same staking plan applied
to this real betting strategy led to an increase of 34.97%. This is another good transition
from artificial research data to actual real-world application.
This artificial data set has 10,000 bets that all have odds of 3.50 exactly and an edge of 5%.
There are some interesting results here. For the first time, even simple level stakes leads to
bankruptcy in 4.9% of 1,000 simulations (Profile 1). However, if the stake is adjusted
according to the balance, then it can avoid bankruptcy. Basically, when the bank takes hits
from a bad losing run, the stakes are reduced and it has chance to continue and come back up.
Intervals of 20, 50 and 200 (orange) did not end in bankruptcy but some of the simulations
ended up lower than the only other non-bankrupt simulation of 10 bet intervals for re-
calculating of stakes (profile 8, green).
59
3.8.1 Applying the Staking Plan to a Real Type 5 Betting System
I will apply this staking plan to a strategy researched myself using Betaminic.com. This one
is called “Colossus 17 Dog Draw” and is analysed further in my book Winning Sports Betting
Strategies. This strategy backs the draw in certain leagues for fixtures meeting selection criteria.
There are 6,980 bets in this data set to use, which is excellent, because the more data a system
has, the more confidence we can have of having found a real pattern in the odds. This strategy
is my personal favorite since it has made me the most profit of all my big data researched
systems so far; 166 points since I started using it in November 2017, and it has not had three
losing months in a row since then.
It has average odds of 3.25 and we can see from the odds frequency that most of the odds fall
n the 3.0 to 3.5 bracket which places it firmly in the Type 5 bracket. There are some odds
outside that range though, but that is sometimes the nature of real systems.
60
The graph shows that this system has had 16 and 17 losses in a row once each as its biggest
losing runs. The longest winning run was 7 in a row which happened just once.
In terms of edge, the overall yield is 8.6% and we see that the higher odds of 4-6 get a much
higher yield but that is based on just 6 bets. The 10% edge in the 3-4 odds range is based on
about 5,900 bets which is a much more meaningful data set.
61
The drawdown shows one very unlucky run right at the beginning which went to -22 but apart
from that, it only goes near -14~--15 on a few occasions thereafter.
1,000 Monte Carlo simulations done on this data with Level Stakes Profile 8 (s1-r10) all end
with better results than simple staking.
62
Level Stakes Profile 1 (s1-r0) Level Stakes Profile 8 (s1-r10)
These profit progress graphs show the peaks where doubling the bank was achieved. Profile 1
shows a long wait for the 3rd peak after a bad run. Profile 8’s re-calculating of the stake each
10 bets has the effect of shortening the peaks slightly and keeping the betting bank balance
higher than Profile 1 during the bad run.
LSP1(s1-
1,167 598 9 6,312 598 598 0.00% 0.2 84 14
r0) Col17
Profile 8 made a 9.4% increase on Profile 1’s basic level stakes plan, but bearing in mind there
was a 4.9% chance of bankruptcy revealed for Profile 1 in 1,000 Monte Carlo Simulations, but
0% for Profile 8. When applied to real data, Profile 8 made a 23.6% increase in profits. Profile
1 only had a 0.2% chance of bankruptcy, but the chance was still there. Profile 8 had no
chance of bankruptcy. 9.4% and 23.6% are quite different, and this may be due to the fact
Colossus 17 Dog Daw has an edge of about 9% compared to the 5% used in the artificial data.
The good point is that the staking plan produces a positive improvement in results.
Interestingly, this real strategy suggests that it can double its bank every 11 months on average.
This is just with level stakes.
63
3.9 Level Stakes Betting on Type 6 High Odds Systems
This artificial data set has 10,000 bets that all have odds of 7.50 exactly and an edge of 5%.
As average odds get higher, the win rate gets lower and the losing runs get longer. This
means that even basic level stakes of 1% in a standard 100 point bank can lead to bankruptcy
in 77% of simulations. To get safety from simple level stakes, we would have to lower the
stake level to 0.25% of a 100 point bank. That really lowers the profit levels that might be
made. Keeping staking levels at 1% and re-calculating that stake level after every 10 bets
avoids bankruptcy since it lowers the stake during long bad runs, it then achieves better profits
in the long term than betting 0.25%. Here profile 8 is the most suitable.
I will then test this staking plan on a strategy from Betaminic called “Pro 8 Underestimated
Underdog v2” which is taken from my book Big Data Betting. It is based on backing the
underdog in football in certain conditions. This leads to higher odds and this strategy has
average odds of 5.21 and has 1,412 past bets in its data set for analysis. It has had a very good
run so far and has an edge of 19%.
64
The strategy actually has odds spread over a wide range, but its average odds of 5.21 puts it in
the Type 6 bracket.
65
The length of losing runs far outstrips winning runs. It has 65 losing runs of between 10 and
21 losses in a row compared to its single longest winning run of 4 in a row. Its high odds
bets lead to much longer losing runs.
The drawdown has a few deep drops at the start but goes on to have much better results later
on.
Its edge broken down into odds range shows greater success at higher and lower odds levels.
66
Level Stakes Profile 8 s1-r10 applied to Betaminic’s Pro 8 Underestimated Underdog v2
1,000 Monte Carlo Simultions show profits concentrating around the 330-360 point range.
Ave. Range Ave. Ave.
Ave. "Let It Low High Bnkrpt
Staking Bets Ave. as % Bets Mnths
Total Ride" Profit Profit Chnce
Plan to Dbl ROI from per to
Profit Bal. Range Range %
Bank Ave. Mnth Dble
LSP8(s1-
409 345 20 1,093 315 391 11.01% 0 18 23
r10) Pro8
LSP1(s1-
517 273 20 663 273 273 0.00% 0 18 29
r0) Pro8
LSP21(s0
25-r0) 2,046 69 20 161 69 69 0.00% 0 18 114
Pro8
Profile 1 applied to this actual strategy showed no chances of bankruptcy. This is because of
the higher edge giving it more wins than the 5% edge of the artificial data previously analysed.
Even so, Profile 8 gives 26% higher profits than Profile 1 and 500% higher profits than
Profile 21. This is higher than the 266% increase that Profile 8 had over Profile 21 with the
artificial dara. Again I put this down to the higher edge in this real strategy.
67
3.10 Level Stakes Betting on Type 7 Very High Odds Systems
This artificial data set has 10,000 bets that all have odds of 15 exactly and an edge of 5%.
Again, the much higher average odds lead to much lower level stakes settings needed to avoid
bankruptcy. A number of new profiles are needed to analyse lower staking levels. In the end
0.25% starting stakes of a standard 100 point bank are recommended. Various re-calculation
intervals are tried, but all lead to the possibility of lower profits than just keeping the same
staking level. Profile 21, constantly betting 0.25% of the starting bank, is the best staking plan
setting here.
I do not have any betting strategies myself with such high odds, so for this test I have used
past betting data from a horse-racing tipster called “Boris”. In the long run he is in profit, but
it has some big drawdowns on the way that can destroy betting banks if not managed
correctly. The data set has 4,628 bets with an edge of 6.5% and average odds of 13.
68
The odds frequency shows 1,612 of 4,628 bets in the Type 7 range. This is a good example of
a real Type 7 betting system.
The drawdowns are pretty steep. You can see how a simple 100 point bank betting 1 point per
bet would soon struggle with this kind of drawdown.
69
Analysis of his edge shows he should just give up on lower odds bets and stick to higher odds
tips where he has a 23% edge.
As can be expected with such high average odds, the losing runs far outstrip the winning runs.
The longest losing run was 32 losses in a row compared to 5 wins in a row. That would be a
tough losing run for even the most steeled professional gambler.
70
Ave. Range Ave.
Ave. "Let It Low High Bnkrpt Ave.
Staking Bets to Ave. as % Bets
Total Ride" Profit Profit Chnce Mnths
Plan Dbl ROI from per
Profit Bal. Range Range % to Dble
Bank Ave. Mnth
LSP21(s02
6,428 72 7 165 72 72 0.00% 0 165 39
5-r0) HRB
LSP23(s01
10,518 44 7 136 44 44 0.00% 0 165 64
5-r0) HRB
LSP1(s1-
1,613 287 7 731 287 287 0.00% 84.8 165 10
r0) HRB
LSP8(s1-
4,285 108 4 211 -41 410 208.80% 0 165 26
r10) HRB
Profile 21 also comes out on top when applied to this real system. Profile 1 shows a 84%
chance of bankruptcy. Profile 8, which has done well for most strategies before this, also
avoids bankruptcy with this horse racing system but the range of results includes some
simulations ending with -41. This doesn’t register as bankrupt since -41 effectively means a
starting bank of 100 would finish on 59 points. That is not technically bankrupt, but ending a
loss. This discounts Profile 8 from use. So the winner is still Profile 21, betting 0.25% of the
starting bank. A conclusion to be drawn here is that high odds strategies are not an
efficient way to use a 100 point bank with level stakes.
One more type of betting system I want to analyse is the mixed system. What if we want a
general staking plan that can be applied to any kind of betting system regardless of
odds. For this I have mixed the artificial data of Types 1-6 together. I have left out Type 7
because I feel the odds are too high and such systems are not worth using. So this data set has
60,000 bets with average odds of 3 and an edge of 5%. (10,000 bet each of odds 1.25, 1.75, 2,
2.25, 3.5 and 7.5)
71
LSP2(s1-r100) 1,709 3,511 5 3.7 trn 3199 3749 7.83% 0
LSP10(s1-r200) N/A -15,053 5 N/A -2E+07 3898 N/A 0.1
LSP11(s1-r500) N/A N/A 5 N/A -94bn 8172 N/A 8.7
LSP12(s1-r1000) N/A N/A 5 N/A -146bn N/A N/A 34.1
LSP20(s0.5-r0) 4,093 1,466 5 2.5m 1466 1466 0.00% 0
The results show some numbers that are just too big to show or my software was unable to
display, but we can see the pattern emerging. Betting 1% of the starting bank with short
intervals of stake recalculation is the best policy. If you don’t want to recalculate stakes, for
example if you have too many bets going on over different periods in different places, then
0.5% of the starting bank as an unchanging stake level is recommended. Profile 8 and 9 show
similar results with Profile 9 having a slightly higher average total profit, but I will recommend
profile 8 as the better staking plan setting since it has a smaller range of results variance
(3.21% versus 5.58%). So betting 1% of your betting bank and recalculating your stake every
10 bets could be described as the best level stakes setting.
72
3.11.1 Applying the Staking Plan to a Real Type 8 Mixed Betting System
Here I have mixed the data from 6 betting strategies from Betaminic: Colossus 03 Free
Scoring Favourites, Pro 1 Away Day Blues, Colossus 05 Against the Trend Overs, Betaminic 5
Over 2.5 Against the Trend, Colossus 17 Dog Draw and Pro 8 Underestimated Underdog v2.
This data set has 15,331 bets with a win rate of 45.8% in it. It has average odds of 2.79 and an
edge of 9.4%.
All 3 profiles show no bankruptcies in 10,000 Monte Carlo simulations. Profile 8 comes out
on top with an average profit of 1,918 points which took 6 years. In monthly terms we can see
that with 199 bets per month on average, this combined system takes 4 months on average to
double the bank.
73
Profile 8 Profit Progress Graph (Peaks show when betting bank was doubled.)
Looking at the profit progress graph, the longest number of bets between 2 peaks (doubling
the bank) was 1,599 bets (bet No. 9,520 to 11,119). At a rate of about 200 bets per month,
that would have been an 8 month wait. Some peaks are very close together, but the longer
gaps seem to be about 6-8 months. The shortest gap was just 172 bets, less than a month of
bets, but that happened only once.
In this we can see the benefits of using multiple strategies together and having an
expectation of how long it will take us to double a 100 point betting bank. This is the key
question all betting systems should answer. How long will it take me to double my
money? (*If the system being used maintains its edge.)
74
4 Level Staking Plans Summary
In conclusion from Chapter 3 we have found the best staking plans for level stakes in
different betting systems depending on their average odds.
Type System Description Recommended Ave. Bets to Average "Let It Ride"
Staking Plan Double Bank Total Profit Balance
Type 1 Low Odds Level Stakes 483 2,069 169,165,038
1.01 ~ 1.49 (1.25) Stake 3% of bank
Recalculate every 100 bets
LSP16(s3-r100)
Type 2 Odds-On Level Stakes 1,484 674 10,689
1.50 ~ 1.99 (1.75) Stake 1% of bank
Recalculate every 200 bets
LSP10(s1-r200)
Type 3 Evens Level Stakes 1,529 654 9,305
1.90 ~ 2.10 (2) Stake 1% of bank
Recalculate every 200 bets
LSP10(s1-r200)
Type 4 Over-Evens Level Stakes 1,570 637 8,271
2.00 ~ 2.49 (2.25) Stake 1% of bank
Recalculate every 100 bets
LSP2(s1-r100)
Type 5 Medium Odds Level Stakes 1,828 547 4,432
2.50 ~ 4.49 (3.5) Stake 1% of bank
Recalculate every 10 bets
LSP8(s1-r10)
Type 6 High Odds Level Stakes 3,155 317 900
4.50 ~9.99 (7.5) Stake 1% of bank
Recalculate every 10 bets
LSP8(s1-r10)
Type 7 Very High Odds Level Stakes 8,000 125 238
10.00 ~ 30.00 (15) Stake 0.25% of bank
No recalculation
LSP23(s0.15-r0)
Type 8 Mixed Odds Level Stakes 1,709 3,510 3.6trn
1.01 ~ 10.00 (3) Stake 1% of bank
Recalculate every 10 bets
LSP8(s1-r10)
If we add up the separate individual average total profits of Type 1-6 we get 4,898 points. This
is 1,388 points higher than Type 8’s mixed odds profit of 3,510 points. This suggests that
applying different staking plans to different odds ranges can produce more profit and
get more out of our betting systems. We can also see from this analysis that a 5% edge on a
system with lower odds is better than a 5% edge on a system with higher odds since we can
bet more of our bank on it each time.
All of these staking plans improve on simple level stakes, betting the same stake each time
without change, except for Type 7. Type 7 highlights the risks of following betting
systems with high odds. It is difficult to use betting capital efficiently with them.
75
Ave.
Ave. Low High Bnkrpt
Simple Level Stakes Bets to Ave. Let It Ride Range as %
Total Profit Profit Chance
Staking Plans Dble ROI Balance from Ave.
Profit Range Range %*
Bank
LS-1 (s1) AD1.25 2,000 500 5 3,200 500 500 0.00% 0
LS-1 (s1) AD1.75 2,000 500 5 3,200 500 500 0.00% 0
LS-1 (s1) AD2 2,000 500 5 3,200 500 500 0.00% 0
LS-1 (s1) AD2.25 2,000 500 5 3,200 500 500 0.00% 0
LS-1 (s1) AD3.5 2,000 500 5 3,200 500 500 0.00% 0
LS-2 (s0.5) AD7.5 4,000 250 5 566 250 250 0.00% 0
LS-3 (s0.25) AD15 8,000 125 5 238 125 125 0.00% 0
LS-2 (s0.5) ADmx 395 2,531 5 4,159,768,772 2531 2531 0.00% 0
In practical terms, the staking plans had this effect on the betting strategies.
Ave. Ave.
Ave. "Let It
Betting System Recommended Bets to Mnths
T Total Ride"
Description Staking Plan Double To Dble.
Profit Balance
Bank Bank
1 Colossus 03 Free Level Stakes
Scoring Favourites Stake 3% of bank
316 17 454 2,326
Recalculate every 100 bets
LSP16(s3-r100)
2 Pro 1 Away Day Level Stakes
Blues Stake 1% of bank
1,032 43 192 378
Recalculate every 200 bets
LSP10(s1-r200)
3 Colossus 05 Against Level Stakes
the Trend Overs Stake 1% of bank
1,128 33 250 566
Recalculate every 200 bets
LSP10(s1-r200)
4 Betaminic 5 Over 2.5 Level Stakes
Against the Trend Stake 1% of bank
365 41 193 381
Recalculate every 100 bets
LSP2(s1-r100)
5 Colossus 17 Dog Level Stakes
Draw Stake 1% of bank
932 11 749 17,977
Recalculate every 10 bets
LSP8(s1-r10)
6 Pro 8 Level Stakes
Underestimated Stake 1% of bank
409 23 345 1,093
Underdog v2 Recalculate every 10 bets
LSP8(s1-r10)
7 Horse Racing Tipster Level Stakes
“Boris” Stake 0.25% of bank
6,428 39 72 165
No recalculation
LSP23(s0.15-r0)
8 6 Betaminic Level Stakes
Strategies Mixed Stake 1% of bank
799 4 1,918 59m
Recalculate every 10 bets
LSP8(s1-r10)
76
Colossus 03 Free Scoring Favourites is the most efficient needing only 316 bets to double its
bank, but since it does not have so many picks each month it would take 17 months to double
the bank with it. The 6 Betaminic Strategies Mixed has the best results, needing only 4 months
on average to double its bank if it maintains its edge.
LS-2 (s0.5) Mix6 1,909 803 9 26,138 803 803 0.00% 0 199 10
Compared to simple level stakes, all of the new plans improve results except for Type 7’s
HRB betting system. This yet again proves the tricky nature of very high odds betting systems.
77
5 Percentage Staking
Rating
Overview A good, solid, bank-protecting staking system when used with the right
settings.
Concept: Bet a certain percentage of your bank on each bet.
Other names: Limit percentage, pure percentage, proportional staking, plateau
philosophy, bankroll percentage plan.
Variables: Percentage of bank to bet, ratchet staking, minimum stake levels.
Percentage staking is very simple. You bet a fixed percentage of your bank on each bet.
As the bank gets bigger, you bet more, but as it gets smaller you bet less. It is also possible to
add the option of ratchet staking to this plan. Ratchet staking is when you never decrease
your stake. If your bank increases in size from some wins, then the stake increases, but if
losses take the bank size down again, with ratchet staking you do not lower the stake. This
brings increased risks of bankruptcy. I discourage the use of ratchet betting. The main
advantage of percentage stakes is that it protects your betting bank. If you have a bad run, the
stake size will decrease along with your bank. One problem is that for some systems, after a
losing run, a winning run will come and you end up betting less on those winners than if you
were using level stakes. You have protected your bank, but miss out on profits. Also, if there
is a particularly bad run, you can end up with stakes so small that it would take a long time to
get back up to the original bank size and it is effectively a bankrupt system. Betting 0.02% of
the original betting bank is pointless. There are more efficient ways to use your betting capital.
Also, most betting sites have a minimum bet requirement. Some people recommend using a
minimum bet to avoid betting too small, others prefer pure percentage staking with no
minimum levels. I will analyse both types of percentage staking. For this reason, in this
analysis for percentage with minimum stake levels “PS” profiles, when the bank reaches 10%
of its original level, the stakes will not go lower so that at minimum they are calculated as a
percentage of 10% of the starting bank, because of this the results show some simulations
going bankrupt. This is a good way to weed out ineffective uses of this staking plan. I will also
analyse without a minimum level “P” profiles or Pure Percentage Staking.
Positives Negatives
Protects the betting bank. (Except if ratchet You bet less on future winners after a losing
staking is used.) run.
As your bank increases, your stakes increase You bet more on future losers after a winning
so you can capitalize on a good system. run.
If a system is failing or losing its edge, it will If the bank gets low, it can take a long time to
automatically bet less on that system. get back up to previous levels.
It is easy to use.
It is considered a low risk staking plan.
78
5.1 Analysis of Percentage Stakes
Ratchet betting goes bankrupt even at low percentage betting levels of 3-5% (PS 8,9,10)
proving their increased risk levels. Percentage betting with minimum limits makes more profit
(PS 13) than ratchet staking. With low average odds of 1.25 and a high win rate, percentage
79
betting allows up to 8% to be bet each time. Pure Percentage betting makes the most
profit though P-57 and P-72. P-57 doesn’t go bankrupt and makes 501 points profit at its
lowest level, but if I was choosing a staking plan, then P-72 has a very narrow range of profit
levels from 7,234 to 7,612. I would much prefer to use a staking plan where I have a good idea
of how much of an improvement it would give me. P-57 is more of a gamble. You don’t know
if you will get the high end of 8,501 or the low end of 501. Even with that said, P-72 uses 15%
of the bank on each bet. For some people, that is over the 5% limit people often suggest as a
red line for staking, but the point of this analysis is to show that different odds bands mean
different staking settings can be used. In this case 15% pure percentage staking with no
ratchet staking or minimum stake level is best.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit as % from Chnce %
Plan Dble ROI Balance
Profit Range Range Ave. *
Bank
P-64(s6) AD1.75 400 2,501 5 3^9 401 2851 48.98% 0
PS5 (s5-NR) AD1.75 424 2,361 5 1.2^9 2300 2730 9.11% 11.7
P-59(s5) AD1.75 427 2,344 5 1^9 2250 2705 9.71% 0
P-62(s9) AD1.75 456 2,191 5 394,064,403 -99 3473 81.52% 0
PS4 (s4-NR) AD1.75 481 2,080 5 182,567,685 2048 2238 4.57% 1.4
P-63(s4) AD1.75 481 2,079 5 181,306,598 2043 2223 4.33% 0
PS3 (s3-NR) AD1.75 584 1,711 5 14,145,660 1695 1760 1.90% 0.1
PS2 (s2-NR) AD1.75 808 1,237 5 529,348 1231 1267 1.46% 0
PS6 (s1-R) AD1.75 1,044 958 5 76,536 942 969 1.41% 0.8
PS1 (s1-NR) AD1.75 1,506 664 5 9,973 661 667 0.45% 0
As the odds increase, so the percentage that can be bet each time decreases. Ratchet betting
goes bankrupt even on 1% stakes for average odds of 1.75 (PS6). I think that highlights
the risks of it very well. Even percentage betting with minimum stakes can only manage
2% stakes (PS2). P-59(s5) does best. So for this second odds bracket of 1.50-2.00 odds,
staking 5% works best. Immediately we can see how different odds ranges need different
staking plans. The odds themselves include the expected win rate excluding your edge.
Any edge you have will increase the win rate from the odds you bet on. So if you base your
staking plan on the inherent win rate in the odds you bet on, then you are able to bet more of
your bank safely on lower odds and should bet less of your bank on the higher odds. A
common staking plan for a wide range of odds would be inefficient. So we need to find
the best staking plans for each odds bracket.
80
5.1.3 Type 3 Systems: Evens Range 1.90-2.10
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range as %
Total Profit Profit Chnce %
Plan Dble ROI Balance from Ave.
Profit Range Range *
Bank
P-59(s5) AD2 522 1,915 5 58,173,396 1828 2413 15.27% 0
P-64(s6) AD2 529 1,891 5 49,258,050 -99 2547 69.96% 0
PS2 (s2-NR) AD2 855 1,170 5 332,699 1154 1186 1.37% 0
PS1 (s1-NR) AD2 1,546 647 5 8,865 645 672 2.09% 0
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range as %
Total Profit Profit Chnce %
Plan Dble ROI Balance from Ave.
Profit Range Range *
Bank
P-63(s4) AD2.25 653 1,531 5 4,062,274 801 1905 36.05% 0
P-59(s5) AD2.25 684 1,462 5 2,518,016 -99 2241 80.03% 0
P-67(s3) AD2.25 711 1,407 5 1,719,856 1350 1616 9.45% 0
PS2 (s2-NR) AD2.25 904 1,106 5 213,497 1090 1156 2.98% 0
P-65(s2) AD2.25 904 1,106 5 213,497 1088 1228 6.33% 0
PS1 (s1-NR) AD2.25 1,587 630 5 7,879 629 665 2.86% 0
P-66(s1) AD2.25 1,587 630 5 7,879 629 665 2.86% 0
Type 4 also does best with 4% stakes but is interesting to see that it makes less profit than
Type 3 systems and has a wider range of results. This suggests that percentage betting does
not work so well for systems with higher average odds and we should look for other
staking plans to be efficient with these ranges.
81
Type 5 continues the trend of reduced returns and can only cope with 2% stakes. Pure
percentage staking manages a profit of 779 but percentage staking with minimum stake sizes
leads to bankruptcy. If you had refilled the account after that to continue, it ends up with an
average profit of 792 points which is only 13 points higher than pure percentage betting. This
underlines that pure percentage betting is better than percentage betting with minimum stake
levels.
Percentage betting is not able to improve on the simple level stake profit of 250 points.
Percentage betting is not able to improve on the simple level stake profit of 125 points. The
trend from Type 6 Systems continues and we are down to 0.5% of the bank per bet just to
avoid bankruptcy. And even P-58 does not beat simple stakes (125 points) at its lowest profit
range of 82.
82
Interestingly, the mixed system allows for 1% staking. This is probably one reason why so
many people think it is a good staking plan to bet 1% each time since it will work in
general for most systems. It is nice to see that sensible rule of thumb proved. It also
increases simple stakes profit (2,531 points) by 38%.
In conclusion, it is clear to see that percentage staking works best for lower odds systems.
Type 1 systems can make 7,304 average total profit and that number steadily declines to just
154 points for Type 7 systems.
83
This is important because we want to be able to apply these staking plans to real betting
systems with the expectation of similar or better results.
As we update the recommended staking plan table, percentage staking improves on level
staking for all odds brackets except 6 and 7.
84
6 Target Profit Staking
Rating
Overview A simple but effective risk sensitive staking plan.
Concept: Bet an odds adjusted stake to get a target profit.
Other names: Fixed Staking, Target Staking, Fixed Profit Staking, Outcome Staking
Variables: Calculate the target profit as a percentage of the current balance or keep it
constant.
Target Profit Staking’s most important characteristic is that it adjusts the stake based on the
odds of each bet. The idea is that the profit made on each bet will be the same, not the
stake being the same. This makes a lot of sense since higher odds have higher risk and we
should reduce stakes for that and lower odds have more chance of winning and so can have
more risked on them. It is one step more complicated than level stakes or percentage betting
since we need to calculate the stake based on the odds. It requires that extra step for each bet.
I view this method of staking as being in the same group as Kelly staking. Kelly staking is
based on adjusting your stake so that you bet more on events that you have a strong edge on
and less on events where you have a weak edge. Kelly is extremely complicated since you need
to know your edge and your edge needs to stay constant, but the correct idea of it is to adjust
your stakes according to the risk you are taking and the potential benefits. Target Profit
Staking does this to an extent by very simply using the odds themselves as defining the risk
inherent in the bet. The higher the risk, the less we bet. The lower the risk, the more we can
bet. You can think of Target Profit Staking as a kind of Simple Kelly Staking if you want.
But one aspect to be very careful of when using Target Profit Stakes, which Kelly Staking also
has in some circumstances, is that if you have a bet with very low odds, it will require a
correspondingly high stake to achieve the target profit. For example if the starting bank is 100
and the target profit is 5% of the bank and a bet with odds of 1.05 comes up, then that would
require the entire 100 point bank to be bet on one event to reach the 5% target. So it is
important to use different staking plan settings for different odds brackets. For this reason, in
the research, real systems with low odds bets such as 1.05 get flagged as bankrupt since they
use more than 100% of the remaining bank to place the bet. Even if it wins, for that one bet,
the bank has gone below zero, and anyway, it is never good to bet all of your bank on one
bet, even if it is Barcelona at home to a bottom of the league side or such like. Even if the
odds are 1.10, it would need 50% of the bank being staked on one event to get a 5% target
profit. For this reason, despite the good results of Target Profit Staking in all brackets, I
cannot recommend it for betting systems that have odds-on bets. Especially odds under
1.50 lead to large stakes being used.
85
Example of Target Profit Staking with Percentage Staking Option with comments
Bet Balance Odds Target 5% Stake Result P/L Balance
1 100 3.25 5 2.22 Win 5 105
We start with a betting bank balance of 100 and our target is to win 5% of the betting bank each bet. The odds of the bet are
3.25, so we stake 2.22 units to win 5 units. The bet wins so the bank increases to 105.
2 105.00 5.10 5.25 1.28 Win 5.25 110.3
The bank is now 105 units, so 5% of that is a target of 5.25 units. Odds of 5.10 mean we risk less and stake 1.28 units to win
the target. The bet wins and the bank goes up to 110.3 units.
3 110.25 4.20 5.51 1.72 Lose -1.72 108.5
The bank is now 110.25, so 5% of that is a target of 5.51 units. Odds of 4.20 mean we need to stake 1.72 units. The bet
loses, so the balance goes down to 108.5 units.
4 108.53 2.50 5.43 3.62 Lose -3.62 104.9
5 104.91 1.70 5.25 7.49 Lose -7.49 97.42
The bank is down to 104.91 after losing bet No. 9. Our odds are lower at 1.70, this means we need to stake more to achieve
our target of 5.25 units. We stake 7.49 units. But this bet loses and the bank is down to 97.42.
6 97.42 4.60 4.87 1.35 Win 4.87 102.3
7 102.29 1.25 5.11 20.46 Win 5.11 107.4
Bet No. 6 wins and the bank is up to 102.29. The target is 5.11 units, but the odds for Bet No. 7 are very low, just 1.25 This
means we need a very big stake of 20.46 units to reach our target. That is nearly 20% of the bank.
8 107.41 11.00 5.37 0.54 Lose -0.54 106.9
The odds are high at 11 on this bet, but the good point of Target Profit Staking is that the stake is reduced due to the high
odds. This risk sensitive staking plan is excellent for protecting the bank at higher odds bets. We only need to stake 0.54
units. The high risk bet loses, but we only lose a small 0.54 stake. The stake reflects the risks and rewards.
9 106.87 1.05 5.34 106.87 Win 5.34 112.2
The odds are very low at 1.05 for this bet, but a bad point of Target Profit Staking is that the stake becomes very high due to
the low odds. We need to bet 106.87 units to achieve our target. That is 100% of the bank. It is never a good idea to bet
your entire bank on one bet, no matter how certain the bet seems. This is a major reason why Target Profit Staking is not
okay for low odds betting systems.
10 112.21 3.50 5.61 2.24 Win 5.61 117.8
Positives Negatives
Adjusts stake according to the odds, betting Low odds bets can require high stakes to
less on high odds and more on lower odds. achieve the target profit.
Relatively simple to use. (Simpler than Kelly Requires an extra step of bet calculation.
Staking)
You don’t need to know your betting history
or edge to use it.
Combined with percentage staking, can
protect the bank on losing runs.
86
6.1 Analysis of Target Profit Stakes
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range as %
Total Profit Profit Chnce %
Plan Dble ROI Balance from Ave.
Profit Range Range *
Bank
TP12 (s5-P) AD1.25 127 7,867 5 48^24 1001 8501 47.67% 0
TP15 (s4-P) AD1.25 133 7,512 5 4^24 7448 7905 3.04% 0
TP14 (s6-P) AD1.25 142 7,033 5 148^21 85 8733 61.48% 0
TP3 (s3-NP) AD1.25 167 6,000 5 115^18 6000 6000 0.00% 90.9
TP13 (s7-P) AD1.25 219 4,562 5 5^15 -99 8034 89.14% 0
TP2 (s2-NP) AD1.25 250 4,000 5 109^12 4000 4000 0.00% 22.9
TP5 (s1-P) AD1.25 378 2,646 5 9^9 2639 2675 0.68% 0
TP1 (s1-NP) AD1.25 500 2,000 5 104,857,600 2000 2000 0.00% 0.1
TP4 (s0.5-NP) AD1.25 1,000 1,000 5 102,400 1000 1000 0.00% 0
First we can see very clearly that not recalculating the target stake according to the current
betting bank size quickly leads to bankruptcy. Instead, adjusting the target profit to be a
percentage of the current bank basically introduces percentage betting into this staking plan.
Percentage betting again shows its ability to help the betting bank weather the storms
of losing runs. However, this time we are using percentage betting in connection with
betting for a fixed profit and not betting with a fixed stake. The higher percentages of
5%, 6% and 7% targets lead to extreme lows in the bank that prove hard to come back from.
5% does have the highest average total profit, but its range of results spans from 1,001 to
8,501, but a target of 4% has a results range of only 7,448 to 7,905. This is a much more stable
and reliable level. This leaves 4% as the best setting for type 1 systems.
Ave.
Ave. Low High
Staking Bets to Ave. Let It Ride Range as % Bnkrpt
Total Profit Profit
Plan Dble ROI Balance from Ave. Chnce %
Profit Range Range
Bank
TP12 (s5-P) AD1.75 392 2,551 5 4.7B 2424 3216 15.52% 0
TP8 (s6-P) AD1.75 399 2,504 5 3.4B 101 3301 63.90% 0
TP15 (s4-P) AD1.75 415 2,409 5 1.7B 2349 2709 7.47% 0
TP7 (s3-P) AD1.75 481 2,077 5 178M 2042 2227 4.45% 0
TP6 (s2-P) AD1.75 639 1,564 5 5.1M 1546 1641 3.04% 0
TP5 (s1-P) AD1.75 1,152 868 5 41,015 861 871 0.58% 0
TP4 (s0.5-NP) AD1.75 3,125 320 5 919 320 320 0.00% 0
87
With higher odds, Target Profit Staking allows for higher target percentage of 5% as best. The
odds are higher, so the stakes are lower to reach the target profit. 6% proves too much and
sends some simulations down to lows that don’t go anywhere. All of the profitable plans are
based on including percentage staking. With Profile TP4 we see how much profit we can get
without using percentage staking, we have to reduce stakes down to 0.5% of the starting bank
to avoid bankruptcy and then our profit level is only 320 points which is lower than simple
level stakes of 500 points profit.
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range as %
Total Profit Profit Chnce %
Plan Dble ROI Balance from Ave.
Profit Range Range *
Bank
TP12 (s5-P) AD2.25 654 1,528 5 3,978,674 127 2017 61.85% 0
TP14 (s6-P) AD2.25 676 1,480 5 2,852,620 -99 2301 81.08% 0
TP15 (s4-P) AD2.25 693 1,444 5 2,222,661 1276 1716 15.24% 0
TP13 (s7-P) AD2.25 782 1,279 5 708,229 -99 2201 89.91% 0
TP7 (s3-P) AD2.25 805 1,243 5 551,827 1216 1416 8.05% 0
TP5 (s1-P) AD2.25 1,934 517 5 3,600 514 522 0.77% 0
Type 4 takes a step down and 4% is the best setting. Other settings do not go bankrupt but
lead to lows that end with the bank at 1% of its starting level.
88
6.1.5 Type 5 Systems: Medium Odds Range 2.50-4.50
We finally have a staking plan that can handle the high odds of Type 7 systems and get a level
of profit similar to other systems. 6% target profit seems best.
89
For mixed systems, the order and odds of bets becomes more important and there is a balance
between risk and reward to be taken. 4% targets has a higher average total profit of 14,018 but
on a few simulations it did end up with a low profit range of 5,701. 3% profit targets seems to
guarantee a profit around 12,135 with little variation. So it is a choice for the user which to go
for. I would say, if your system has more low odds in it then 3% is better, but if it is more high
odds heavy then 4% would be ok. In those cases running past data through the TSM analyser
and Monte Carlo simulations can really help.
Simple Simple
Ave. Bets Ave. Low High
Staking Ave. Range as % Levels Level
to Dble Total Profit Profit
Plan ROI from Ave. Stakes Stakes
Bank Profit Range Range
Profit Increase
TP15 (s4-P) AD1.25 133 7,512 5 7448 7905 3.04% 500 1402%
TP12 (s5-P) AD1.75 392 2,551 5 2424 3216 15.52% 500 410%
TP12 (s5-P) AD2 522 1,914 5 1828 2312 12.64% 500 282%
TP15 (s4-P) AD2.25 693 1,444 5 1276 1716 15.24% 500 188%
TP12 (s5-P) AD3.5 1,290 775 5 665 1110 28.71% 500 55%
TP12 (s5-P) AD7.5 3,247 308 5 260 519 42.05% 250 23%
TP14 (s6-P) AD15 3,135 319 5 260 611 55.02% 125 155%
TP7 (s3-P) ADmx 494 12,135 5 12075 12411 1.38% 2,531 379%
90
Here we can see that TP12 (s5-P) Pro1 shows a 78% chance of bankruptcy. This comes from
a bet on PSG to win at odds of 1.05. These low odds requite a large stake to reach the target
profit of 5%. The balance was 37.94, and 5% of that is 1.897 which was rounded up to 1.9. A
bet of 38 was needed to achieve a target profit of 1.9. It is this bet, even though it wins, that
flags this system as bankrupt since it takes the balance under 0 for that one bet and technically
there would not have been enough money to place the bet without using extra funds. If this
bet had lost, then the system would have lost its entire remaining bank in one bet. This
highlights the risks of target profit betting with low odds. In the 11.34% of simulations
that were not bankrupt the PSG 1.05 bet still uses all of the bank but has a few pennies to
spare by luck of rounding down in a calculation and being over 100 points in its current
cumulative balance at that moment. If we use TP15 (s4-P) Pro1 then this bankruptcy issue it
avoided, but the PSG bet still uses a large chunk of the betting bank to place its bet.
For TP14 (s6-P) HRB it actually makes a loss. After researching many different settings, all
staking plans ended with losses. This is because this system made its profits on the higher
odds bets. It had a better edge there, in fact all its lower odds actually had a negative edge, in
other words, they were loss making. The fixed odds staking system ended up reducing the
stakes on those higher odds bets and increasing the stakes on the lower odds bets which made
losses. Because of this, there is no way this system can make profits with Target Profit Staking.
It compounded the losses made by the lower odds bets. The point here is that if the edge is in
the higher odds bets and not the lower odds bets, then Target Profit will not be so effective.
This is another reason why it is important to log your bets so that you can know your
own betting history and test it in staking software to find where your edge lies.
91
6.4 Recommended Staking Plans Leaderboard
In the artificial data results, target profit staking beats percentage betting in odds-on systems
types 1, 2 and 3. But due to the high stakes required from low odds bets, I prefer to use
percentage betting as the recommended staking plans for low odds systems at the moment.
Target Profit Staking enters the leaderboard for Types 5, 6, 7 and 8.
92
7 Retirement Staking Plan
Rating
Overview A capital inefficient loss recovery plan that does not beat simple level stakes.
Concept: To mix target profit staking and loss recovery while also keeping stakes low
to protect the betting bank.
Other names: None
Variables: Initial stake as a percentage of the betting bank
Average winning odds (historical)
The Retirement Staking Plan tries to take the odds adjusted efficiency concept of Target Profit
Staking and merge it with a bank protecting Divisor while also adding in an element of loss
recovery by adding losses to the Target. It does a good job of protecting the bank, but as a
result the bank grows much more slowly than other plans.
Target Profit Staking started with the idea that if you are betting on a “2 to 1” odds bet,
then instead of betting a fixed stake, you adjust your stake to get a fixed profit.
Retirement Staking takes this idea, but instead of having a fixed target, it is variable. It
calculates the target based on the average winning odds. It then also increases the target by
adding losses to it. It then reduces the size of the stake by creating a divisor for the target
which is double the average wining odds. It also increases the divisor after a certain number of
losses to limit the stake size and try to prevent it escalating too quickly. After a win, the Target
is reduced by the profit. The Divisor is reduced to a level on a par with a previous Target in
the series which is similar to the new reduced Target. After an increase of 200 points, the
Target is permanently increased by 10 to reflect the large size of the bank. The Divisor is
calculated using the fractional odds figure (reflecting the profit increase, not total increase).
93
Example Bet Odds: 3.25
Target 6
Divisor = stake 6 =1
We bet £1 at odds of 3.25 to win £2.25 profit
Initial Bet
Stake x Divisor = Target
Subsequent Bets
Stake = Target* / Divisor**
*(Target = current series P/L + initial target)
**(Divisor = increases by one for each loss that is greater than the initial divisor, after a win,
the divisor is lowered based on the current target so that the divisor matches the previous
divisor of a similar target level.)
Reset series when the current target is lower than the initial target. (So in theory, one win from
the first bet would reset the series.) When Bank increases by 200 points, the target is increased
by 10 points. This is a fixed increase and never reduced.
This staking plan is quite complicated, but it becomes much easier to understand when we
look at a series of 20 bets and see it in practical application.
Current
Bet Bet Current
Series Divisor Target Stake Result P/L Balance
No. Odds Series P/L
Bet No.
1 1 6 6 1 3.25 Win +2.25 +2.25 102.25
2 1 6 6 1 5.1 Win +4.1 +4.1 106.35
3 1 6 6 1 4.2 Lose -1 -1 105.35
4 2 6 7 1.17 2.7 Lose -1.17 -2.17 104.18
5 3 6 8.17 1.36 4.5 Lose -1.36 -3.53 102.82
6 4 6 9.53 1.59 3.6 Lose -1.59 -5.12 101.23
7 5 6 11.12 1.85 7.5 Lose -1.85 -6.97 99.38
8 6 6 12.97 2.16 4.3 Lose -2.16 -9.13 97.22
9 7 7 15.13 2.16 5.3 Lose -2.16 -11.29 95.06
10 8 8 17.29 2.16 3.3 Lose -2.16 -13.45 92.9
11 9 9 19.45 2.16 2.9 Win 4.1 -9.35 97
12 10 7 15.35 2.19 3.1 Win 4.6 -4.75 101.6
13 11 6 10.75 1.79 4.6 Lose -1.79 -6.54 99.81
14 12 7 12.54 1.79 7.3 Lose -1.79 -8.33 98.02
15 13 8 14.33 1.79 1.6 Win 1.07 -7.26 99.09
16 14 6 13.26 2.21 1.9 Win 1.99 -5.27 101.08
17 15 6 11.27 1.88 3.5 Win 4.7 -0.57 105.78
18 16 6 6.57 1.1 2.4 Lose -1.1 -1.67 104.68
19 17 7 7.67 1.1 4.2 Win 3.52 +1.85 108.2
20 1 6 6 1 3.7 Lose -1 -1 107.2
Jump forward 400 bets to see what happens when the bank balance goes over 300 points.
421 1 6 16 2.67 4.0 Win 8.01 +8.01 315.5
94
Here is the same bet series with explanatory notes.
Current Current
Bet Bet
Series Divisor Target Stake Result P/L Series Balance
No. Odds
Bet No. P/L
1 1 6 6 1* 3.25 Win +2.25 +2.25** 102.25
*The initial Stake is 1% of a 100 point betting bank.
**The bet wins, the new target would be 6 - 2.25 = 3.75, which is lower than the initial target of 6, so the series is reset.
2 1 6 6 1* 5.1 Win +4.1 +4.1** 106.35
*It is an option to recalculate the initial stake of a new series based on the current betting bank balance. In this example
we are not increasing the initial stake to keep things simple, but in the results to follow, linking to the current balance and
increasing the initial stake to 1.06 shows better performance.
**The bet wins, the new target would be 6 – 4.1 = 1.9, which is lower than the initial target of 6, so the series is reset.
3 1 6 6 1* 4.2 Lose -1 -1 105.35
*The initial Stake is 1% of a 100 point betting bank.
4 2 6 7* 1.17** 2.7 Lose -1.17 -2.17 104.18
*Bet No. 3’s loss is added to the current target. 6 + 1 = 7
**The next Stake is the Target divided by the Divisor 7 / 6 = 1.17
5 3 6 8.17* 1.36** 4.5 Lose -1.36 -3.53 102.82
*Bet No. 4’s loss is added to the current target. 7 + 1.17 = 8.17
**The next Stake is the Target divided by the Divisor 8.17 / 6 = 1.36
6 4 6 9.53* 1.59** 3.6 Lose -1.59 -5.12 101.23
*Bet No. 5’s loss is added to the current target. 8.17 + 1.36 = 9.53
**The next Stake is the Target divided by the Divisor 9.53 / 6 = 1.59
7 5 6 11.12* 1.85** 7.5 Lose -1.85 -6.97 99.38
*Bet No. 6’s loss is added to the current target. 9.53 + 1.59 = 11.12
**The next Stake is the Target divided by the Divisor 11.12 / 6 = 1.85
8 6 6 12.97* 2.16** 4.3 Lose -2.16 -9.13 97.22
*Bet No. 7’s loss is added to the current target. 11.12 + 1.85 = 12.97
**The next Stake is the Target divided by the Divisor 12.97 / 6 = 2.16
9 7 7* 15.13** 2.16*** 5.3 Lose -2.16 -11.29 95.06
*The initial divisor was 6. This is the 7 losing bet in this series which is higher than 6. So 1 is added to the divisor. 6 + 1
th
=7
**Bet No. 8’s loss is added to the current target. 12.97 + 2.16 = 15.13
***The next Stake is the Target divided by the Divisor The Divisor has increased from 6 to 7 so the stakes are kept under
control. In fact, the stake remains the same as bet 8. 15.13 / 7 = 2.16
10 8 8* 17.29** 2.16*** 3.3 Lose -2.16 -13.45 92.9
*The current divisor is 7. This is the 8 losing bet in this series which is higher than 6. So 1 is added to the divisor. 7 + 1
th
=8
**Bet No. 9’s loss is added to the current target. 15.13 + 2.16 = 17.29
***The next Stake is the Target divided by the Divisor The Divisor has increased from 7 to 8 so the stakes are kept under
control. In fact, the stake remains the same as bet 9. 17.29 / 8 = 2.16
11 9 9* 19.45** 2.16*** 2.9 Win 4.1 -9.35 97
*The current divisor is 8. This is the 9th losing bet in this series which is higher than 6. So 1 is added to the divisor. 8 + 1
=9
**Bet No. 10’s loss is added to the current target. 17.29 + 2.16 = 19.45
***The next Stake is the Target divided by the Divisor The Divisor has increased from 8 to 9 so the stakes are kept under
control. In fact, the stake remains the same as bet 10. 19.45 / 9 = 2.16
12 10 7* 15.35** 2.19*** 3.1 Win 4.6 -4.75 101.6
*The current divisor is 9. Bet No. 11 won so the +4.1 profit from that is subtracted from the Target which becomes 15.35.
We look back in the series to find when the target was at a similar level. Bet No. 9 (The 7 th in this series.) has a target of
15.13. That is close to this target. The Divisor at that level was 7. So we lower the Divisor from 9 to 7.
**Bet No. 11’s profit is subtracted from the current target. 19.45 – 4.10 = 15.35
***The next Stake is the Target divided by the Divisor The Target has been lowered and the Divisor has been reduced
95
from 9 to 7 so the stakes are kept in proportion. 15.35 / 7 = 2.19
13 11 6* 10.75** 1.79*** 4.6 Lose -1.79 -6.54 99.81
*The current divisor is 7. Bet No. 12 won so the +4.6 profit from that is subtracted from the Target which becomes 10.75.
We look back in the series to find when the target was at a similar level. Bets No. 6 and 7 (The 5th and 6th in this series.)
had targets of 9.53 and 11.12 with Bet No. 7’s 11.12 being nearer to 10.75. Bet No. 7’s Divisor was 6. So we lower the
Divisor from 7 to 6.
**Bet No. 12’s profit is subtracted from the current target. 15.35 – 4.60 = 10.75
***The next Stake is the Target divided by the Divisor The Target has been lowered and the Divisor has been reduced
from 7 to 6 so the stakes are kept in proportion. 10.75 / 6 = 1.79
14 12 7* 12.54** 1.79*** 7.3 Lose -1.79 -8.33 98.02
*The current divisor is 6. This is the 10th losing bet in this series which is higher than the original Divisor which was 6. So
1 is added to the divisor. 6 + 1 = 7
(Note that we do not add 4, we only add 1. It is not the difference between the number of losing bets and the initial
Divisor, just the fact that the number of losing bets is higher than the initial Divisor that triggers 1 being added to the
current Divisor. Basically we add 1 to the Divisor for each loss after the 6 th loss when the initial Divisor is 6. If the initial
Divisor was 11, then we would add 1 for each loss after the 11th loss. This is the Recovery part of the staking plan.)
**Bet No. 13’s loss is added to the current target. 10.75 + 1.79 = 12.54
***The next Stake is the Target divided by the Divisor The Divisor has increased from 6 to 7 so the stakes are kept under
control. In fact, the stake remains the same as bet 13. 12.54 / 7 = 1.79
15 13 8* 14.33** 1.79*** 1.6 Win 1.07 -7.26 99.09
*The current divisor is 7. This is the 11 losing bet in this series which is higher than 6. So 1 is added to the divisor. 7 + 1
th
=8
**Bet No. 14’s loss is added to the current target. 12.54 + 1.79 = 14.33
***The next Stake is the Target divided by the Divisor The Divisor has increased from 7 to 8 so the stakes are kept under
control. In fact, the stake remains the same as bet 13 and 14. 14.33 / 8 = 1.79
16 14 6* 13.26** 2.21*** 1.9 Win 1.99 -5.27 101.08
*The current divisor is 8. Bet No. 15 won so the +1.07 profit from that is subtracted from the Target which becomes
13.26. We look back to the beginning of the series to find when the target was at a similar level. Bets No. 8 and 9 (The 7 th
and 8th in this series.) had targets of 12.97 and 15.13 with Bet No. 8’s 12.97 being nearer to 13.26. Bet No. 8’s Divisor was
6. So we lower the Divisor from 8 to 6.
(Note that Bet No. 14 and 15 have targets 12.54 and 14.33 with Divisors of 7 and 8, but we do not use these Targets, but
go back to the beginning of the current series and find a similar Target from there.)
**Bet No. 15’s profit is subtracted from the current target. 14.33 – 1.07 = 13.26
***The next Stake is the Target divided by the Divisor The Target has been lowered and the Divisor has been reduced
from 8 to 6 so the stakes are kept in proportion. 13.26 / 6 = 2.21
17 15 6* 11.27** 1.88*** 3.5 Win 4.7 -0.57 105.78
*The current divisor is 6. Bet No. 14 won so the +1.99 profit from that is subtracted from the Target which becomes
11.26. We look back to the beginning of the series to find when the target was at a similar level. Bet No. 7 (The 6th in this
series.) had a target of 11.12. Bet No. 7’s Divisor was 6. So we keep the Divisor at 6.
**Bet No. 16’s profit is subtracted from the current target. 13.26 – 1.99 = 11.27
***The next Stake is the Target divided by the Divisor The Target has been lowered but the Divisor remains the same.
11.27 / 6 = 1.88
18 16 6* 6.57** 1.1*** 2.4 Lose -1.1 -1.67 104.68
*The current divisor is 6. Bet No. 17 won so the +4.7 profit from that is subtracted from the Target which becomes 6.57.
We look back to the beginning of the series to find when the target was at a similar level. Bet No. 4 (The 2nd in this
series.) had a target of 7. Bet No. 4’s Divisor was 6. So we keep the Divisor at 6.
**Bet No. 17’s profit is subtracted from the current target. 11.27 – 4.7 = 6.57
***The next Stake is the Target divided by the Divisor The Target has been lowered but the Divisor remains the same.
6.57 / 6 = 1.1
19 17 7* 7.67** 1.1*** 4.2 Win 3.52 +1.85**** 108.2
*The current divisor is 6. This is the 12th losing bet in this series which is higher than 6. So 1 is added to the divisor. 6 + 1
=7
**Bet No. 18’s loss is added to the current target. 6.57 + 1.1 = 7.67
***The next Stake is the Target divided by the Divisor The Divisor has increased from 6 to 7 so the stakes are kept under
control. In fact, the stake remains the same as bet 18. 7.67 / 7 = 1.1
96
****The Current Series Profit is finally positive again. This means that the next target will be 4.15 which is lower than the
initial target, so the series has been completed and the staking plan is reset.
20 1* 6 6 1** 3.7 Lose -1 -1 107.2
*The staking plan has been reset, so a new series begins again in a similar way to bet No.3.
**The initial Stake is 1% of a 100 point betting bank.
Jump forward 400 bets to see what happens when the bank balance goes over 300 points.
421 1* 6** 16*** 2.67**** 4.0 Win 8.01 +8.01 315.5
*The staking plan has been reset, so a new series begins again.
**The initial Divisor is still 6. (Average Winning Fractional Odds x 2. i.e. 3 x 2 = 6)
***The bank balance has increased by over 200 points, so 10 is added to the initial target permanently.
****The initial Stake is 1% of a 100 point betting bank. But the Stake is the Target divided by the Divisor The Divisor is
6. The Target is 16. 16/6 = 2.67 In this way, after a significant increase in the bank size, then and only then, is the base
stake level increased via the Target.
We can see the stakes only went up to 2.21 at its highest. If we did not have the Divisor
keeping stakes under control, then the recovery side of the staking plan would have become
too risky. Note how the actual odds of the current bet do not affect the stake, only the average
winning odds, which set the initial Target and Divisor. In this way the stakes follow a more
regular pattern and you will not get caught out by 1.05 odds bet needing a 100% stake to
achieve a target profit of 5% which can occur with Target Profit Staking. Retirement Staking
tries to take the edges of the various other plans it draws from.
Positives Negatives
Protects the bank Bank grows slowly
Increases stakes in a controlled manner to Complicated to calculate
recover losses
Requires historical data to find your average
winning odds
97
7.1 Analysis of Retirement Staking Plan
Even with very low odds bets, the recovery part of the staking plan leads to bigger stakes that
give increased chances of bankruptcy. With stake of just 1% of the bank, we are able to avoid
bankruptcy, but it only leads to a slight improvement on simple level stakes. It takes a lot of
effort and calculation to get a very small improvement on profits. Percentage staking and
target profit staking easily beat this staking plan.
As the odds increase, we have to reduce the stakes to 0.5% to avoid bankruptcy. After doing
that the profits are lower than simple level stakes.
We now have to reduce odds to 0.15% just to avoid bankruptcy in some of our Monte Carlo
simulations. Some people might say this staking plan works well if you have a big bank to start
with, but that is nonsense. If it is beaten on a 100 point bank by other staking plans, then it is
not going to beat them with a bigger bank. It is an inefficient staking plan.
99
7.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
The very high odds are the worst performers, needing stakes of just 0.05% to avoid
bankruptcy.
With all the brackets mixed together, it still only manages stakes of 0.15% without going
bankrupt.
The summary from this is that Retirement Staking is just not efficient. In order to stay out of
bankruptcy and maintain normal stake amounts, you would need a large betting bank. This is
proven by the fact that when limited to a 100 point bank, it can only manage very low stake
levels. Target Profit Staking and Percentage Staking are both much more efficient at using
betting capital. This Retirement Staking plan should not be used at all.
100
One useful result of the research is that Type 1-6’s aggregated total profit comes to 14,514
while the Mixed Odds profit is 14,018. This shows that it is more efficient to use different
staking plans for different odds ranges than using a one size fits all staking plan. Even
with a bad staking plan, it still makes sense to use different settings for different odds
brackets. This will become the basis of Whitaker Staking at the end of this book.
Unsurprisingly, the same poor results from the artificial data are reflected when applying the
staking plans to real betting systems. The Recommended Staking Plans Leaderboard remains
unchanged.
101
8 1326 Staking Plan
Rating
Overview A terrible, illogical staking plan that can only work if lucky.
Concept: To take advantage of a high win rate.
Other names: None
Variables: Link stake to current bank total or not.
Ratchet staking
This staking plan increases the bet after a win, then reduces and then increases again. It makes
no sense. It is suggesting that over time, after a winning bet, there is more chance of another
winner, so you should bet three times your bet, then if you win, bet just twice your stake since
there is slightly less chance of a winner after two wins but still more chance than on your first
bet. Then if you win that, we suddenly bet 6 times your normal stake since it says there is
much more chance of getting a fourth winner in a row than three in a row. If we are saying
that there is more chance of a winner after a winner, then it would make more sense for our
progression system to go 1,2,3,4 and not 1,3,2,6. It makes no sense. Also, if you have a system
with a high win rate, then betting regressively 4,3,2,1 (reducing the stake after each win) would
make more sense since there is less chance of 4 winners in a row than two winners in a row.
To prove this, I will also test those systems. I predict that “1234 progressive staking” (that
increases the bet after a loss) will work better for systems with low hit rates than 1326 staking
and “4321 regressive staking” (that decreases the stake after a win) will work better for
systems with high hit rates than 1326 staking. I include the 1326 system here to prove it
should not be used. I imagine that the 1326 system was invented by someone backing horses
at various odds and it just happened that the 1326 got lucky at hitting some winners on the 3
and 6 point bets. But it gives a talking point to prove the nature of systems based on their
average odds which reflect their win rate. It will also give us the starting point for considering
Fibonacci and Martingale staking plans which are progressive staking plans on a much more
aggressive and dangerous level than simple 1234 progressive staking.
102
6 110.29 4.60 1 1.10 Win 3.97 114.26
7 114.26 1.25 3 3.43 Win 0.86 115.11
Bet No.6 won, so we move up one level in the 1326 series and bet 3 times the normal stake on the next bet. The bank
was 114.26 so we are betting 3% of the bank on Bet No. 7.
8 115.11 4.15 2 2.30 Win 7.25 122.37
Bet No. 7 also won, so we move up another level in the 1326 series and bet 2 times the normal stake on the next bet.
The bank was 115.11 so we are betting 2% of the bank on the Bet No. 8.
9 122.37 11.00 6 7.34 Lose -7.34 115.03
Bet No. 8 also won, so we move up another level in the 1326 series and bet 6 times the normal stake on the next bet.
The bank was 122.37 so we are betting 6% of the bank on the Bet No. 9. A weak point of this betting style is it does not
take the odds into account, so we are betting 6% of our bank on odds of 11.
10 115.03 1.05 1 1.15 Win 0.06 115.08
Bet No. 9 lost so we go back to the start of the 1326 series and bet the normal stake on Bet No. 10 which is 1% of the
bank. Again this system does not take into account the odds and we end up putting only 1% of our bank on a 1.15 odds
bet.
Positives Negatives
Takes advantage of luck in the short term Illogical
Fun to use Not profitable in the long term
Simple Does not take into account the odds
At low odds 1326 staking’s best settings are betting 4% of the bank each bet. This is without
ratchet staking and using percentage betting. Normal percentage betting allows up to 15% of
the bank to be bet, but 1326 bets much lower with much worse results.
103
8.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
104
8.1.5 Type 5 Systems: Medium Odds Range 2.50-4.50
105
The poor results continue with a mixed system and the high win rates of Type 1 and Type 2
are not enough to lift the staking plan into winnings that improve on simple level stakes.
In the end, only Types 1 and 2 show an improvement on simple level stakes. Their high hit
rates mean the higher stakes on the 2nd, 3rd and 4th bets often lead to big wins. I can imagine it
must be quite a rush to put 6 times your normal bet on the next game after you have just won
three in a row. For that reason, I would say the only redeeming feature of this staking plan is
its ‘fun’ factor. But I don’t think anyone reading this book is in need of that kind of fun. So
this staking is firmly and conclusively rated as a very bad one.
106
8.3 Applying to Real Betting Systems
Ave. Range Ave. Simple
Ave. "Let It Low High Brpt Ave.
Staking Bets Ave. as % Bets Level
Total Ride" Prof. Prof. chnce Mnths
Plan to Dbl ROI from per Stakes
Profit Bal. Rnge Rnge % to Dble
Bank Ave. Mnth Increase
1326S-7
(s4-NR- 114 1,257 8 608,061 593 1943 53.70% 0 19 6 1042%
LC) Col3
1326S-4
(s2-NR- 317 625 8 7,611 3 1153 92.00% 0 24 13 334%
LC) Pro1
1326S-2
(s1-NR- 614 459 7 2,408 148 799 70.92% 0 34 18 140%
LC) Col5
1326S-2
(s1-NR- 0 173%
182 388 21 1,472 201 585 49.48% 9 20
LC)
BTMO2.5
1326S-9
(s0.5-NR- 1,154 605 9 6,626 307 907 49.59% 0 84 14 1%
LC) Col17
1326S-11
(s0.15- 0 -13%
1,722 82 19 177 27 127 60.98% 18 96
NR-LC)
Pro8
1326S-12
257.14 0 71%
(s0.1-NR- 11,019 42 6 134 -38 178 165 67
%
LC) HRB
1326S-9
22,351,3 121%
(s0.5-NR- 863 1,777 10 1221 2277 29.71% 0 199 4
12
LC) Mix6
Interestingly, when applied to real betting systems, three of them show profits. I can imagine
that it was this kind of result that brought this staking plan into existence. It was a lucky run of
winners that made it look good. But from extensive testing, we know that it is inefficient and
we can improve the profits of a good betting system more efficiently with other staking plans.
107
9 1234 Staking Plan
Rating
Overview Increase stakes after a loser because we are “due” a winner. Only good for
high win rates over 50%.
Concept: To take advantage of consistent win rates
Other names: None
Variables: Link to cumulative total or not.
1234 is the logical opposite of 1326 staking. 1326 staking increases after a winner, but 1234
staking increases after a loser. The simple idea is that after a loser, we are more likely to get a
winner. If we are following a big data based betting system or a tipster with a long history and
a consistent win rate, we can have the opinion that we are due a win after a loss. In table
games such as roulette there is the Gambler’s Fallacy, which is the idea that when red has
come up in the last 10 throws then there must be a higher chance of black coming up on the
next throw. This is incorrect because the throws are unconnected. It does not matter if red
come up 1,000 times in a row, there would be no effect on the probability of black coming up
on the next throw. In sports betting, however, the Gambler’s Fallacy is true to a certain extent,
but not in others. If you backed the home win for a mid-table team, eventually they would get
a home win. It is extremely unlikely that they would lose 100 home games in a row. This is
because the results are connected. When a team won their previous 10 games, we can
definitely imagine there is a good chance they will win their next if the opposing team is from
the same league and of a similar level to the previous teams they played. This is why there is
value in sports betting. If you can find an edge from your own sports knowledge or find a
pattern of value in the betting odds, then there is money to be made. The big data systems
from Betaminic have found such value patterns in the odds and when used with staking plans
like 1234 that take advantage of their consistency, then higher profits can be made.
108
Bet No. 6 won, so we go back to the original stake of 1%. (If Bet No. 6 had lost, we would still go back to the original
stake level of 1% because the 1234 series ends on the 4th bet. This acts as a limiter and stops the stake increasing too high.
8 114.88 4.15 1 1.15 Win 3.62 118.50
9 118.50 11.00 1 1.19 Lose -1.19 117.31
10 117.31 1.05 2 2.35 Win 0.12 117.43
Positives Negatives
Very simple to use. Requires a consistent betting system
Easy to follow because it taps into the Chasing losses is not a good habit to have.
“chasing losses” urge
Has a limit to stake increases at the 4th bet. Not good for systems with most odds over
evens.
Good for systems with most odds under
evens.
Ave. Range
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride as %
Total Profit Profit Chnce
Plan Dble ROI Balance from
Profit Range Range %*
Bank Ave.
1234 S-9 (s8-LC) AD1.25 199 5,025 6 133^15 201 6277 60.46% 0
1234 S-8 (s7-LC) AD1.25 212 4,721 5 16^15 3484 5584 22.24% 0
1234 S10 (s9-LC) AD1.25 212 4,715 5 15^15 -99 6957 74.83% 13.1
1234 S-7 (s6-LC) AD1.25 236 4,231 5 545^12 3085 4985 22.45% 0
1234 S-6 (s5-LC) AD1.25 270 3,699 5 13^12 2975 4244 17.15% 0
1234 S-5 (s4-LC) AD1.25 326 3,070 6 174B 2621 3484 14.06% 0
1234 S-4 (s3-LC) AD1.25 421 2,378 6 1.4B 1963 2683 15.14% 0
1234 S-3 (s2-LC) AD1.25 615 1,626 6 7,847,796 1407 1821 12.73% 0
1234 S-2 (s1-LC) AD1.25 1,203 831 6 31,737 728 916 11.31% 0
1234 S-1 (s1-NLC) AD1.25 1,681 595 6 6,182 515 650 11.34% 0
Staking Plan Name Key:
1234 S-8 = Percentage Staking Plan Profile No. 8
s = percentage of bank staked
LC= Link stake to cumulative balance (percentage staking) NLC= Not link stake to cumulative
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
1234 Staking with 7% of the bank and adjusting the stake as the bank changes (percentage
staking option) produces a very respectable result of 4,721 points average profit. (1234 S-8
(s7-LC)) 1234 Staking without the percentage option 1234 S-1 (s1-NLC) can only manage 1%
109
of the bank being staked. Anything higher than that leads to bankruptcy. Again, having an
element of percentage staking in a plan gives it much better bank protection and avoids
bankruptcy.
Range
Ave. Bets Ave. Low High Bnkrpt
Staking Ave. Let It Ride as %
to Dble Total Profit Profit Chnce
Plan ROI Balance from
Bank Profit Range Range %*
Ave.
1234 S-4 (s3-LC) AD1.75 543 1,841 5 34,830,701 362 2912 69.26% 0
1234 S-3 (s2-LC) AD1.75 614 1,628 5 7,957,348 883 2255 42.14% 0
1234 S-5 (s4-LC) AD1.75 648 1,543 5 4,414,615 -99 3251 108.55% 0
1234 S-2 (s1-LC) AD1.75 1,025 976 5 86,707 645 1295 33.30% 0
For the higher odds of range of AD1.75, the size of stake reduces to 2%.
Similar to 1326 staking, as the odds get higher, the staking plan no longer beats simple level
stakes betting. While it does get average profits higher than simple level stakes, some of the
Monte Carlo Simulations end up lower than the simple level stakes profit of 500. For this
reason, the staking method becomes inefficient from this point.
110
9.1.5 Type 5 Systems: Medium Odds Range 2.50-4.50
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
1234 S-2 (s1-LC) AD3.5 1,650 606 5 6,672 -99 1311 116.34% 0
1234 S-11 (s0.5-LC) AD3.5 1,946 514 5 3,526 88 872 76.26% 0
1234 S-12 (s0.25-LC) AD3.5 3,175 315 5 888 123 519 62.86% 0
The inefficiencies get worse and the stakes are forced lower.
Some of the low profit ranges start to show negative numbers. This highlights the dangers
of increasing stakes after losing a bet to recover losses. Even with a limited 1234 increase,
it becomes an inefficient use of the betting capital available.
111
9.2 1234 Staking Plan Summary
Simple Simple
Ave. Bets Ave. Low High Range
Staking Ave. Levels Level
to Dble Total Profit Profit from
Plan ROI Stakes Stakes
Bank Profit Range Range Ave. %
Profit Increase
1234 S-8 (s7-LC) AD1.25 212 4,721 5 3484 5584 22.24% 500 844%
1234 S-3 (s2-LC) AD1.75 614 1,628 5 883 2255 42.14% 500 225%
1234 S-2 (s1-LC) AD2 1,029 972 5 469 1381 46.91% 500 94%
1234 S-2 (s1-LC) AD2.25 1,059 942 5 393 1422 54.50% 500 88%
1234 S-11 (s0.5-LC) AD3.5 1,946 514 5 88 872 76.26% 500 3%
1234 S-13 (s0.1-LC) AD7.5 7,353 136 5 0 250 91.91% 250 -83%
1234 S-14 (s0.05-LC) AD15 15,152 66 5 -23 153 133.33% 125 -89%
1234 S-11 (s0.5-LC) ADmx 1,983 3,025 5 2218 3786 25.92% 2,531 19%
We can see clearly that this progressive staking plan that increases the stakes after a loss only
works for lower odds with high hit rates. For odds over evens it becomes inefficient. To avoid
going bankrupt, you have to reduce stake sizes and then you might end up making less profit
than just using simple level stakes.
None of the 1234 Staking Plans improve on Percentage Staking or Target Profit Staking, so
the recommended staking plan table from the end of Chapter 6 remains the same.
112
10 4321 Staking Plan
Rating
Overview A simple but effective staking plan that reduce stakes after a winner to take
advantage of consistent betting systems.
Concept: To bet less on losers by reducing stakes after winners.
Other names: None
Variables: Link stakes to cumulative total (percentage staking option)
The 4321 staking plan bets 4 times the base stake on the first bet, if it wins, it bets 3 times the
base stake on the next bet, if that also wins, it bets 2 times the base stake on the next bet. If
that also wins, the stake is reduced to the base stake until a loser comes. In this way you are
trying to bet less on bets where you are due a loser. If a loser is hit at any point, the series
resets to the beginning and the next stake is 4 times the base stake.
113
Positives Negatives
More conservative than increasing stakes It goes against some people’s betting
after a loss. mentality to bet less after you win, so it
needs strong discipline to follow.
More effective than 1234 staking.
114
10.1.4 Type 4 Systems: Over-Evens Range 2.00-2.50
4321 Staking is still able to improve on simple level stakes, showing its improvement on 1326
Staking.
The stakes are pushed to 0.25% just to stay in profit at the lowest ranges.
115
10.1.8 Type 8 Systems: Types 1-6 Mixed
For the mixed data with a wide range of odds, the stakes are pushed down to 0.5%, but that
still equals an opening bet of 2% that gradually goes down to 0.5% if there are 3 wins in a row.
Lower odds systems are able to make profit, but higher odds are inefficient.
116
10.3 Applying to Real Betting Systems
Surprisingly, almost all the real systems show improvement on simple level stakes with 4321
Staking. It definitely shows that reducing stakes after a win is better than increasing stakes
after a loss. This idea is taken further with Phantom Bet Staking, a plan analysed later in this
book.
117
11 1234 Winners Staking Plan
Rating
Overview A terrible system if used in practice, but created and analysed here for
comparison to 1326 Staking.
Concept: To increase bets after a win to take advantage of a winning streak.
Other names: None
Variables: Link the stake to the cumulative total or not (percentage staking option)
1234 Winners Staking is a new staking plan I have created for the purposes of showing how
illogical 1326 Staking is and to help compare 3 different concepts in staking. 1234 Staking
increases stakes after a loser (progressive or recovery staking), 4321 Staking reduces stakes
after a winner (regressive staking). 1234 Winners increases stakes after a winner (progressive
staking, but not recovery staking) which is similar to 1326 Staking, but 1234 Winners does it in
a more logical progression of 1,2,3,4 rather than 1,3,2,6.
118
Positives Negatives
Makes more sense than 1326 You bet more after a winner and could end
up losing all your winnings unless you often
win 3 or 4 bets in a row.
Takes advantage of lucky streaks. Does not take into account the odds being
bet on.
Most staking plans work with low odds systems and 1234 Winners is no exception, but it is all
downhill after this.
Even with 1.75 odds, 1% of staking is the most allowed, but this low staking amount does not
beat simple level stakes.
119
11.1.3 Type 3 Systems: Evens Range 1.90-2.10
Even with relatively low odds of 2.25, all of the staking plans show some low profit ranges in
negative numbers. The odds come down to 0.25% of the bank and profit levels are much
lower than the 500 points of simple level stakes.
We are not even at the highest odds range, but for the first time the average total profit
numbers turn negative.
120
11.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
For mixed odds a starting stake of 0.5% is possible, but the average total profit is well below
the 2,531 simple level stakes.
Simple Simple
Ave. Bets Ave. Low High Range
Staking Ave. Levels Level
to Dble Total Profit Profit from
Plan ROI Stakes Stakes
Bank Profit Range Range Ave. %
Profit Increase
1234W-S7 (s5-LC) AD1.25 250 4,000 4 109^12 2770 5420 500 700%
1234W-S3 (s1-LC) AD1.75 1,908 524 3 3,779 203 889 500 4.8%
1234W-S10 (s0.5-LC) AD2 4,237 236 3 513 38 442 500 -53%
1234W-S10 (s0.5-LC) AD2.25 5,525 181 3 351 -7 341 500 -74%
1234W-S11 (s0.25-LC) AD3.5 27,778 36 1 128 -48 152 500 -93%
1234W-S11 (s0.25-LC) AD7.5 -500,000 -2 0 99 -58 92 250 -100%
1234W-S12 (s0.1-LC) AD15 -166,667 -6 0 96 -31 59 125 -101%
1234W-S10 (s0.5-LC) ADmx 6,826 879 3 44,264 1 1520 2,531 -65%
1234 Winners Staking doesn’t even come close to affecting the top staking plans. For the
systems with high hit rates it has some success, but for lower hit rates it cannot even make a
profit.
121
11.3 Applying to Real Betting Systems
Interestingly, the real systems have better results than the artificial data sets. The random
nature of the event fixtures and odds in the real systems seem to do better with the 1234
Winners Staking.
122
11.4 1326, 1234, 4321 and 1234W comparison
4321 Staking shows itself to be the best compared to 1326, 1234 or 1234W staking. 1326
and 1234W bet more after a winner, 4321 Staking bets less after a winner and 1234 bets
more after a loser. 4321 Staking shows itself to be a much better strategy than the others. But
none of the three improve on the current recommended staking methods of Percentage and
Target Profit Staking. My main point here was to show that the 1326 staking plan is
meaningless because it has no theory behind it. There is no real betting system where there is
more chance of 4 winners in a row than 3 winners in a row. If there was more chance of a
winner after a winner, it should be 1236, not 1326. That is just getting lucky with the order of
bets, the horse racing list for the day or the fixture list. These four staking plans also show
that progressive staking (increasing stakes) after a winner or loser is not an efficient
way to bet because it forces you to reduce staking levels to avoid going bankrupt. In
other words, to allow more space for higher staking levels, the base level of staking is lowered
and this means you bet less in general and make less profit. By using a better staking plan, the
betting capital can be used more effectively. A lot of betting communities and websites often
say you should not chase losses by increasing stakes after a loss and this research proves that
advice to be very good indeed. It is better to have a staking plan that uses the right
123
amound of your betting bank on each bet and not connecting previous wins or losses
directly into the size of the next bet. However, including percentage betting as part of that
staking plan is good because it takes into account your overall performance up to now rather
than just your previous bet when deciding the next stake size. Recovery staking is not a good
idea, but percentage staking is.
124
12 Fibonacci Staking Plan
Rating
Overview A dangerous, aggressive progressive recovery staking plan.
Concept: To recover losses by increasing stakes after losses following the Fibonacci
sequence.
Other names: None
Variables: Initial stake as percentage of bank or fixed starting stake level
Reset the sequence after a win or not
Going down two places after a win or resetting after a win.
Sequence used: 1-2-3-5-8-13-21-34 or 1-1-2-3-5-8-13-21-34.
How long the sequence used is: e.g. 8 levels, 5 levels, etc.
Fibonacci Staking is a more aggressive version of 1234 Staking and since it has a longer series
and higher stake levels. But the idea is the same, you bet more after a loss since we are “due” a
winner. As this is a progressive (recovery) staking plan, I want to show just how poor it
is and to discourage people from using it. The Fibonacci Sequence itself is created by
adding the previous two numbers in the sequence to get the next number. i.e. 5-8-13 is
5+8=13
Example of Fibonacci Staking to 8 Levels in the Series and Resetting after a Win
Bet Balance Odds Fibonacci Multiple Stake Result P/L Balance
1 100 3.25 1 1.00 Win 2.25 102.25
2 102.25 5.10 1 1.02 Win 4.19 106.44
3 106.44 4.20 1 1.06 Lose -1.06 105.38
4 105.38 2.50 2 2.11 Lose -2.11 103.27
Bet No. 3 lost, so we go up one level in the series to bet 2 times the base stake on Bet No. 4.
5 103.27 1.70 3 3.10 Lose -3.10 100.17
Bet No. 4 lost, so we go up one level in the series to bet 3 times the base stake on Bet No. 5.
6 100.17 4.60 5 5.01 Win 18.03 118.20
Bet No. 5 lost, so we go up one level in the series to bet 5 times the base stake on Bet No. 6.
7 118.20 1.25 1 1.18 Win 0.30 118.50
Bet No. 6 won, so the series is reset and the next stake is at level 1.
8 118.50 4.15 1 1.18 Win 3.73 122.23
9 122.23 11.00 1 1.22 Lose -1.22 121.01
10 121.01 1.05 2 2.42 Lose -2.42 118.59
11 118.59 2.50 3 3.56 Lose -3.56 115.03
12 115.03 1.70 5 5.75 Lose -5.75 109.28
13 109.28 4.60 8 8.74 Lose -8.74 100.54
14 100.54 1.25 13 13.07 Lose -13.07 87.47
15 87.47 4.15 21 18.37 Lose -18.37 69.10
16 69.10 7.00 34 23.49 Lose -23.49 45.61
Bets No. 9 to 15 all lost and now on Bet No. 16 we bet 34 times the base stake.
17 45.61 2.40 1 0.46 Win 0.64 46.25
125
Bet No. 16 lost, but we reached the 8th level in the series and thus reset back to the start for the next bet. So we accept
our losses and start again.
18 46.25 4.15 1 0.46 Lose -0.46 45.79
19 45.79 1.56 2 0.92 Lose -0.92 44.87
20 44.87 1.40 3 1.35 Win 0.54 45.41
Positives Negatives
Sounds Italian. Stakes can quickly rise.
Very aggressive.
Almost any staking plan can improve on level stakes at low odds levels and even Fibonacci
can get 603 points instead of 500. But after this, all other odds brackets and results are
bankrupt or inefficient.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave.% %*
Bank
Fib11 (s2-sq2-R) AD1.75 743 1,346 6 1,126,844 460 1810 50.15% 98.4
Fib1 (s1-sq1-NR) AD1.75 946 1,057 6 152,015 400 1600 56.76% 88.1
Fib3 (s1-sq1-R) AD1.75 1,074 931 6 63,473 303 13253 695.49% 71.9
Fib2 (s1-sq2-NR) AD1.75 1,328 753 6 18,482 278 1049 51.20% 59.2
Fib4 (s1-sq2-R) AD1.75 1,490 671 6 10,469 200 900 52.16% 35.1
Fib7 (s0.5-sq1-NR) AD1.75 1,838 544 6 4,341 224 763 49.54% 13.8
Fib9 (s0.5-sq1-R) AD1.75 2,075 482 6 2,825 159 709 57.05% 5.9
Fib8 (s0.5-sq2-NR) AD1.75 2,494 401 6 1,611 150 550 49.88% 4
Fib12 (s0.5-sq2-R) AD1.75 2,755 363 6 1,238 154 490 46.28% 0.7
Fib10 (s0.25-sq1-NR) AD1.75 3,597 278 6 687 121 397 49.64% 0.1
126
Fib14 (s0.25-sq2-R) AD1.75 5,464 183 6 356 79 229 40.98% 0
Fib13 (s0.1-sq1-NR) AD1.75 7,874 127 7 241 42 177 53.15% 0
Fib15 (s0.1-sq2-R) AD1.75 10,638 94 8 192 53 113 31.91% 0
Immediately we see the stakes lowered to 0.25% of the bank just to avoid bankruptcy and that
is at relatively low 1.75 betting odds. Since the base stake is so low, the average profits of 183
are lower than simple level stakes. This again shows how staking plans that increase stakes
after a loss actually produce less overall profit because the initial stakes have to be low
to avoid bankrupting the 100 point betting bank.
The stake is forced down to a near meaningless 0.01% and even at that level the low profit
range comes out negative.
With mixed systems the stake is allowed up to 0.05%, but the profits are very poor.
128
12.2 Fibonacci Staking Plan Summary
Simple Simple
Ave. Bets Ave. Low High Range
Staking Ave. Levels Level
to Dble Total Profit Profit from
Plan ROI Stakes Stakes
Bank Profit Range Range Ave. %
Profit Increase
Fib1 (s1-sq1-NR) AD1.25 1,658 603 6 6,534 523 667 500 20%
Fib14 (s0.25-sq2-R) AD1.75 5,464 183 6 356 79 229 500 -63%
Fib15 (s0.1-sq2-R) AD2 12,346 81 6 175 2 137 500 -84%
Fib15 (s0.1-sq2-R) AD2.25 12,195 82 5 177 -18 174 500 -84%
Fib17 (s0.01-sq2-R) AD3.5 142,857 7 3 105 -20 34 500 -98.5%
Fib17 (s0.01-sq2-R) AD7.5 142,857 7 2 105 -20 33 250 -97%
Fib17 (s0.01-sq2-R) AD15 32,258 31 5 124 -31 95 125 -75%
Fib15 (s0.1-sq2-R) ADmx 10,101 594 6 6,139 320 848 2,531 -76%
Fibonacci shows it can only work with some high strike rate systems (Type 1), but in terms of
efficient use of betting capital, it lags far behind other staking plans. In other words, there are
better ways to use a 100 point betting bank than with Fibonacci staking plans. Fibonacci is
awful and should not be used.
With the real systems also, Fibonacci staking comes nowhere near the performance of Target
Profit Staking, 1234 Staking or 4321 Regressive Staking.
129
13 Martingale Staking
Rating
Overview A terrible, dangerous and ruinous staking method that should never be used.
Concept: Increase stakes after each loss in order to win back the losses and win the
initial bet when a winner comes.
Other names: None
Variables: Percentage Staking option
Martingale staking comes from the idea that after a loss, you should increase your bet to win
back your losses and make win the profit you aimed for in the first bet. This staking idea was
originally created for table games in casinos, but due to the Gambler’s Fallacy (thinking that
unrelated events are related), the house margin, statistically unsurprising losing runs needing a
large bankroll and table staking limits, the staking plan did not work for that. Some people,
however, felt that since sports betting events have some connection, Martingale can work. A
point being that, for example, if you back a Premier league team to win, and keep increasing
bets on the same team’s next games if they lose until that win comes, you will eventually get a
win, since no team ever goes an entire season without a win. Derby came nearest in 2007/8
season with only 11 points, the lowest in any Premier League season, and in that season they
got 8 draws and only one solitary win against Newcastle. So in sports, the events are
connected more than table games are. Martingale is most commonly used for evens bets since
it makes the doubling up simpler, so for the example below I have used all odds of 2.
130
Positives Negatives
None. Stakes get high very quickly.
You will lose all your bank doing this.
Very inefficient use of capital.
Bookmaker stake limits can stop bets being placed
when bets get too high
Even with a starting staking of 0.01% of the bank and a very high win rate of 83%, we still get
13% of 1,000 simulations ending in bankruptcy. There is no way to make a reasonable
working staking plan from Martingale. If it cannot cope at this level, it will not cope with
others. Any staking settings higher than 0.01% of the bank just have higher bankruptcy levels.
13.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce %
Plan Dble ROI Balance
Profit Range Range Ave. % *
Bank
M-1 (s0.01) AD1.75 23,256 43 10 135 28 49 24.42% 24.2
As the odds increase, the win rates go down and the chance of bankruptcy increases.
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range from
Total Profit Profit Chnce %
Plan Dble ROI Balance Ave. %
Profit Range Range *
Bank
M-1 (s0.01) AD2 18,868 53 10 144 52 53 0.94% 31.3
M-2 (s0.02) AD2 9,524 105 10 207 100 106 2.86% 42.6
Still all bankrupt.
131
13.1.4 Type 4 Systems: Over-Evens Range 2.00-2.50
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range from
Total Profit Profit Chnce %
Plan Dble ROI Balance Ave. %
Profit Range Range *
Bank
M-1 (s0.01) AD2.25 16,667 60 11 152 -26 72 81.67% 28.9
Even at low stakes of 0.01% the bank goes bankrupt.
132
13.2 Martingale Stakes Plan Summary
The results conclusively prove that Martingale is much worse than simple level stakes. If you
have a system that makes profit on level stakes, this plan will not make it better. Trying to
make a loss making system make money with Martingale staking is just gambling,
and not in a good way. Never, ever use Martingale Staking.
133
14 Parlay Staking
Rating
Overview A better way to make multiple bets.
Concept: Let your winnings ride over a number of bets.
Other names: Let It Ride, Double or Bust, Staggered Multiples
Variables: How many bets to let the bet ride on.
Link the stake to the cumulative total (percentage staking)
Parlay Staking is where we add the profits of a win to the next bet for a specified series of
bets. The longer the series, the less likely it is to make a profit. If we bet on a series of 3 bets,
it is in effect a multiple bet over 3 legs, but we manually bet on the 3 legs instead of making
one multiple bet at the beginning. In general, multiple betting is not an efficient way to
bet, but if you do like it, then Parlay Staking is a better way of Multiple Betting
because if a leg lets you down, then you don’t waste the legs and can begin a new
multiple. For those people who like to pick 3 or 4 “banker bets”, then this is a better way to
do that, since if one leg loses, then you can still have a chance with the other legs and find a
new fixture to replace the lost one. For example, if you have Man Utd, Chelsea and Man City
all at home on a Saturday 3pm kick off, it is better to make them all the first leg of 3 separate
parlays that you can continue on the Saturday 5:15pm kick off for the English Premier or
maybe Italian or Spanish kick-offs later on Saturday afternoon/evening. The third legs can be
filled with Sunday Italian, Spanish and English Premier games. In this way, as there always is,
the one game that fails may be made up by the others.
134
Bet No. 7 won, so the profit of 1.15 from Bet No.7 is added to Bet No. 7’s Stake of 4.60 and a stake of 5.75 is placed on
Bet No. 8.
9 120.85 2.33 3 23.86 Win 31.73 152.59
Bet No. 8 won, so the profit of 18.11 from Bet No.8 is added to Bet No. 8’s Stake of 5.75 and a stake of 23.86 is placed on
Bet No. 9.
10 152.59 1.05 0 1.00 Win 0.05 152.64
Bet No. 9 wins, but that is the 3rd Parlay and so we reset to the start again. The stake returns to the base of 1.00.
Positives Negatives
An improvement on simple multiples betting Risks losing your profits quickly unless you
as you can start a new multiple if the first leg have regular winning runs.
loses instead of wasting other legs on a lost
bet.
Good for betting systems with odds under Not good for betting systems with most odds
2.00. over 2.00.
Sounds like the way pirates might bet.
135
Par37 (s8-r4-LC) 171 5,853 5 41^18 5188 6484 11.07% 0
Par53 (s11-r8-LC) 176 5,668 5 11^18 1 8451 74.54% 0
Par12 (s9-r3-NLC) 179 5,578 6 6^18 5285 5861 5.16% 77.4
Par36 (s7-r4-LC) 187 5,359 5 1^18 4781 5881 10.26% 0
Par11 (s8-r3-NLC) 202 4,948 5 78^15 4673 5212 5.45% 60
Par33 (s6-r3-LC) 224 4,460 5 2/5^15 4096 4746 7.29% 0
Par10 (s7-r3-NLC) 231 4,329 5 1^15 4092 4524 4.99% 33
Par35 (s7-r3-LC) 231 4,328 5 1^15 4090 4530 5.08% 32.6
Par34 (s5-r4-LC) 241 4,152 5 315^12 3799 4475 8.14% 0
Par31 (s5-r3-LC) 260 3,847 6 38^12 3619 4109 6.37% 0
Par9 (s6-r3-NLC) 267 3,740 6 18^12 2358 3928 20.99% 13.5
Par23 (s5-r2-LC) 282 3,541 5 4^12 3398 3680 3.98% 0
Par22 (s5-r4-NLC) 292 3,424 5 2^12 3136 3636 7.30% 9.9
Par26 (s6-r2-LC) 297 3,363 6 1.3^12 3256 3466 3.12% 6.3
Par29 (s4-r3-LC) 315 3,176 5 363^9 2964 3405 6.94% 0
Par8 (s5-r3-NLC) 324 3,084 5 192^9 2929 3229 4.86% 3
Par17 (s4-r2-LC) 344 2,909 5 57^9 2771 3006 4.04% 0
Par21 (s4-r4-NLC) 365 2,740 5 17^9 2546 2930 7.01% 1.7
Par7 (s4-r3-NLC) 405 2,471 5 2.7^3 2347 2577 4.65% 0.7
Par28 (s3-r3-LC) 407 2,455 5 2.4^3 2315 2597 5.74% 0
Par30 (s3-r5-NLC) 438 2,285 6 756M 2075 2480 8.86% 0.1
Par16 (s3-r2-LC) 447 2,236 5 538M 2148 2312 3.67% 0
Par3 (s3-r2-NLC) 475 2,107 6 220M 1458 2658 28.48% 99.9
Par20 (s3-r4-NLC) 483 2,070 6 170M 1934 2228 7.10% 0.3
Par6 (s3-r3-NLC) 538 1,860 6 39M 1760 1948 5.05% 0.1
Par27 (s2-r3-LC) 593 1,686 6 11.8M 1602 1790 5.58% 0
Par15 (s2-r2-LC) 655 1,527 6 3,951,191 1475 1571 3.14% 0
Par25 (s2-r5-NLC) 659 1,518 6 3,712,234 1338 1632 9.68% 0
Par18 (s2-r4-NLC) 730 1,370 5 1,330,794 1270 1458 6.86% 0
Par5 (s2-r3-NLC) 817 1,224 5 483,735 1162 1288 5.15% 0
Par2 (s2-r2-NLC) 913 1,095 5 197,824 1063 1123 2.74% 0
Par38 (s1-r5-LC) 954 1,048 5 142,822 930 1130 9.54% 0
Par32 (s1-r4-LC) 1,052 951 5 72,911 871 1021 7.89% 0
136
Par19 (s1-r3-LC) 1,159 863 6 39,618 818 914 5.56% 0
Par14 (s1-r2-LC) 1,282 780 5 22,286 751 808 3.65% 0
Par24 (s1-r5-NLC) 1,319 758 5 19,134 684 819 8.91% 0
Par13 (s1-r4-NLC) 1,460 685 5 11,536 640 720 5.84% 0
Par4 (s1-r3-NLC) 1,634 612 5 6,955 582 639 4.66% 0
Par1 (s1-r2-NLC) 1,825 548 5 4,463 528 564 3.28% 0
Staking Plan Name Key:
Par8 = Parlay Staking Plan Profile No. 8
s = percentage of bank staked
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
I did not expect good results from a staking plan that let its winnings ride over a number of
bets, but very surprisingly Parlay Staking’s 7,014 points came very close to beating
Percentage Staking’s 7,304 points profit on the current staking plan leaderboard. To see if
it could beat percentage staking I tried a large number of variations and all that research can
be seen here. Looking at the table below, Percentage Staking has a narrower range of
results, while Parlay Staking has a wider range and its highest profit is higher than
percentage staking. Percentage staking keeps its top spot, though. But this clearly shows that
staggering bets on low odds multiples is a better way to bet rather than a normal multiple bet
where you bet on all the fixtures in one bet at the start. It is interesting that for Type 1 low
odds systems it is able to make a good profit having even 5 bets in its series. That means many
times it wins 5 bets in a row and makes a good profit. The results also show that using
percentage staking option as part of Parlay Staking is much safer and protects the bank.
137
Par31 (s5-r3-LC) AD1.75 538 1,858 5 39,186,576 -99 3301 91.50% 0
Par56 (s3-r4-LC) AD1.75 548 1,826 5 31,391,186 -99 3351 94.47% 0
Par55 (s2-r4-LC) AD1.75 579 1,727 5 15,804,764 811 2279 42.50% 0
Par27 (s2-r3-LC) AD1.75 596 1,677 5 11,175,656 1037 2262 36.52% 0
Par32 (s1-r4-LC) AD1.75 885 1,130 5 252,138 620 1628 44.60% 0
Par19 (s1-r3-LC) AD1.75 1,008 992 5 96,876 675 1275 30.24% 0
Parlay Staking still shows a wide range of results with ranges from 30-44% in its results.
Unsurprisingly, for low win rate systems letting winnings ride onto the next bet is not at all
profitable.
139
14.2 Parlay Staking Plan Summary
Parlay Staking shows itself to be good for systems with odds-on bets. As the odds go over
evens, this method stops being efficient.
The Type 1 Colossus 3 betting system goes bankrupt 1 time in 1,000 simulations, but because
of that, it means the staking system was too aggressive. Instead of betting up to 5 legs in a
Parlay, betting 4 legs shows no bankruptcy in any simulations.
140
15 Rolling Doubles
Rating
Overview A slightly complex way of letting winnings ride.
Concept: To let half the returns of a winning bet carryover to the next bet.
Other names: Rollover staking
Variables: Link the stake to the cumulative bank (percentage staking option)
Similar to Parlay Betting, this staking plan lets the winning from one bet run onto the next bet.
Positives Negatives
Stakes only increase after a winner. A little complex to calculate how much to
carry over to the next bet.
Only half of a winning bet is carried over, If the win rate is low, it throws away winnings.
not all of it.
141
15.1 Analysis of Percentage Stakes
Rolling Doubles gets a very good result from the low odds range and shows similar results to
the Parlay staking plan. If we look at the comparison table below, Parlay staking with 11% of
the bank over a series of 5 bets does slightly better than Rolling Doubles which does best with
10% of the bank being bet and linking the stake to the cumulative balance. Rolling Doubles
shows a slightly narrower profit range than Parlay staking. Neither beat percentage betting,
which is a lot simpler to manage than either of the other two staking plans. But it shows why
Rolling Doubles is a respectable staking method since it obviously can produce great
improvements on level staking.
Ave. Bets Ave. Let It Low High Range Bnkrpt
Staking Ave.
to Dble Total Ride Profit Profit as % Chnce
Plan ROI
Bank Profit Balance Range Range from Ave. %*
P-72(s15) AD1.25 137 7,304 5 971^21 7234 7612 2.59% 0
Par47 (s11-r5-LC) AD1.25 143 7,014 5 130^21 5501 8051 18.18% 0
RD-4 (s10-LC) AD1.25 149 6,716 5 16^21 6029 7409 10.27% 0
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce %
Plan Dble ROI Balance
Profit Range Range Ave. % *
Bank
RD-6 (s4-LC) AD1.75 426 2,350 5 1,186,328,320 601 2953 50.04% 0
RD-5 (s3-LC) AD1.75 472 2,119 5 239,235,491 1591 2541 22.42% 0
RD-2 (s1-LC) AD1.75 1,048 954 5 74,443 796 1086 15.20% 0
Rolling Doubles still shows respectable results.
142
15.1.3 Type 3 Systems: Evens Range 1.90-2.10
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range from
Total Profit Profit Chnce %
Plan Dble ROI Balance Ave. %
Profit Range Range *
Bank
RD-5 (s3-LC) AD2 577 1,734 5 16,590,525 1119 2341 35.24% 0
RD-7 (s2-LC) AD2 675 1,482 5 2,892,441 1098 1882 26.45% 0
RD-2 (s1-LC) AD2 1,101 908 5 54,119 694 1117 23.29% 0
For Type 3 systems Rolling Doubles allows 3% of the bank to be staked on each bet.
Similar to Parlay Staking, as odds get higher, the staking plan turns inefficient.
Here the low profit ranges both show -99, effectively bankrupt, even with low stakes of
0.25%.
143
15.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
144
15.3 Applying to Real Betting Systems
Ave. Rang Ave. Ave.
Ave. "Let It Low High Level
Staking Bets Ave. from Brpt Bets Mths
Total Ride" Prof. Prof. Stakes
Plan Profile to Dbl ROI Ave. % per to
Profit Bal. Rnge Rnge Increase
Bank % Mnth Dble
RD-4 (s10- 7,902,38 1379%
88 1,627 8 1341 1929 18.07% 0 19 5
LC) Col3 2
RD-5 (s3-LC) 374%
290 683 8 11,377 448 881 31.70% 0 24 12
Pro1
RD-5 (s3-LC) 327%
345 816 7 28,603 459 1134 41.36% 0 34 10
Col5
RD-7 (s2-LC) 474%
123 572 21 5,271 437 719 24.65% 0 9 14
BTMO2.5
RD-8 (s0.5- -3%
1,210 577 9 5,457 331 771 38.13% 0 84 14
LC) Col17
RD-8 (s0.5- 97%
527 268 19 641 90 490 74.63% 0 18 29
LC) Pro8
RD-9 (s0.25- 11%
5,785 80 5 174 -92 1042 708 % 0 165 35
LC) HRB
RD-2 (s1-LC) 13,991,7 236%
567 2,706 10 2136 3116 18.11% 0 199 3
Mix6 38,594
The results from the artificial data are played out in the real system results, but the mixed
system manages to get some decent results. Even so, Rolling Doubles remains a staking plan
that should only be considered for betting systems with odds under evens.
145
16 Square Root Staking
Rating
Overview A controlled way to increase stake size when in profit.
Concept: When in profit, add the square root of the profit to the stake.
Other names: None
Variables: None
This staking plan seeks to answer the question level stakers have which is “When and by how
much should I increase my stakes?” Square root staking only increases stakes when the
betting bank is in profit. It then adds the square root of the profit to the stake. This has the
effect of increasing stakes at a slow and fixed pace. The concept has a lot of good points
about it. It does not increase when losing. It does not increase in response to specific
wins, but only the overall profit level. It increases by a fraction of the profits. It is like a
kind of percentage betting, but only with the profits. The stakes do increase faster than with
percentage staking though.
146
Positives Negatives
Only increases stakes when in profit. Stakes size increases faster at lower profit levels.
Does not increase stakes when in loss. Stakes do not reflect the odds or risk level of a bet.
Increases stakes by a fraction of profits. Progressive staking plans force down the base
staking level and reduce overall profits.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce %
Plan Dble ROI Balance
Profit Range Range Ave. % *
Bank
Sq2 (s2) AD1.75 574 1,742 5 17,536,484 1098 2298 34.44% 4.2
Sq1 (s1) AD1.75 897 1,115 5 227,240 597 1597 44.84% 0
Square Root Staking just manages to stay green for Type 2 betting systems.
Ave.
Ave. Low High Bnkrpt
Staking Bets to Ave. Let It Ride Range from
Total Profit Profit Chnce %
Plan Dble ROI Balance Ave. %
Profit Range Range *
Bank
Sq1 (s1) AD2.25 1,170 855 5 37,481 267 1411 66.90% 0.3
Sq6 (s0.5) AD2.25 2,160 463 5 2,476 -3 1009 109.29% 0
Sq7 (s0.25) AD2.25 3,891 257 5 594 -8 792 155.64% 0
For the smallest stake level I can test, 0.01%, the system just manages to show profit without
going bankrupt.
148
16.1.8 Type 8 Systems: Types 1-6 Mixed
Secure Staking adjusts the stake size depending on the odds. This is one of a number of
staking plans that seek to react to the odds as a risk signifier and bet accordingly. Other
such staking plans being Kelly and Target Profit staking. This is one of the best staking
plans out there if you know the right settings for each bracket. The difficult aspect here
is to know how many odds brackets to use and how much of the bank to bet on each odds
bracket. In this book I have used 7 odds brackets for my research, so I will simply transfer
those odds brackets to this Secure Staking method. It is possible to use either level stakes or
percentage stakes on each bracket. From the research so far, percentage staking is the most
efficient use of betting capital, so I will put the settings from Percentage Staking into this plan.
150
Positives Negatives
Varies the stake according to the risk of the You need to choose the right settings to make
bet which is signaled by the odds. the plan effective.
Can use percentage or level stakes betting Percentage staking is not efficient for higher
with each bracket. odds bets.
Flexible brackets and staking levels according
to personal risk levels.
Much simpler than Kelly Staking (an odds
sensitive staking plan)
You don’t need to know your betting history
to use it.
151
17.1.5 Type 5 Systems: Medium Odds Range 2.50-4.50
Ave. Bets Ave. Low High Bnkrpt
Staking Ave. Let It Ride Range from
to Dble Total Profit Profit Chnce %
Plan ROI Balance Ave. %
Bank Profit Range Range *
Sec1 AD3.5 1,284 779 5 22,132 625 1105 30.81% 0
For the mixed Type 8 system Secure Staking spectacularly improves on Percentage Staking. It
gets a huge 13,766 points profit. This really proves the advantage of using different
staking settings on different odds brackets.
152
17.3 Applying to Real Betting Systems
It is when we apply the Secure Staking Plan to real betting systems that we can see the effects
of this method much better.
Depending on the betting system, using Percentage Staking, Target Profit Staking or Secure
Staking can be better. It is for this reason that it is important to know your betting history
and know your edge at different odds brackets. If you have a better edge at higher odds,
then Secure Staking will reduce your profits compared to Target Profit Staking. If you have a
153
consistent edge across all odds brackets or you have a better edge at lower odds, then Secure
Staking will improve your results. That is why it is good keep a record of your bets and even
better, to run them through staking software like TSM and see where you edge is strongest
and to know for sure which staking system to use.
I chose these odds brackets at the beginning of my research not knowing what the best
staking plan was. I chose them to reflect a variety of betting styles that exist, low priced
favorites, Asian-handicap betting, draw betting, long-shot horse races, etc. For this reason it is
entirely possible that more brackets with different percentages may produce better results.
There will probably be a smoother progression from 15% in Bracket 1 down to 0.5% in
Bracket 6 that can be found with more brackets of narrower odds ranges. But for the
purposes of comparable research, I calculated with these brackets. Another important point to
note is that Percentage Staking for the higher odds bracket of 4.50 upwards did not produce
better results than level stakes. But I cannot use level stakes settings for those brackets using
the Secure Staking plan while keeping percentage staking on lower odds. For this reason I
used the least worst settings from Percentage Staking for that highest odds bracket. As we
know from the research so far, Percentage Staking is best for odds under 2.00 and Target
Profit Staking is better for odds over 2.00. For this reason I have created Whitaker Staking
which is shown in Chapter 35 that will apply the best individual staking plans to each bracket
instead of just the best percentage staking plan settings. This will create the best and most
efficient staking plan there has ever been and ultimately the purpose of this book.
154
17.4 Recommended Staking Plans Leaderboard
Secure Staking replaces Target Profit Staking as the recommended staking plan for mixed
odds betting systems. But the other staking plans remain the same.
155
18 Kelly Staking
Rating
Overview A famous but complicated method that tries to combine the historical edge
with the bet odds to get a risk adjusted stake, but with a dangerous flaw.
Concept: To bet more on odds that the bettor has a higher edge with.
Other names: None
Variables: The Divisor
The historical average winning rate (expected hit rate)
Link to the cumulative total or not
Kelly Staking tries to take historical betting data to work out your edge based on your average
win rate. It then makes you bet more on events where the odds are above your average win
rate and not bet on events where the odds are under your average win rate. This ends up
leading to low odds bets being ignored and high odds bets having very large amounts of the
bank being staked on them unless a Divisor is introduced to lower the staking levels.
((fractional odds * expected win rate) – (expected loss rate)) / fractional odds = Kelly Recommended Stake %
or
(((decimal odds – 1) * expected win rate) – (1-expected win rate)) / (decimal odds – 1) = Kelly Recommended Stake %
It is quite complex, but is much easier to understand when we look at a worked through
example.
Example of Kelly Staking where Expected Hit Rate is 45.8% and Divisor 1
Bet Odds Kelly Recommended Stake % Stake Result Profit Balance
1 3.66 25.42% 25.42 Win 67.62 167.62
Kelly Equation (((3.66-1)*0.458) – (1-0.458)) / (3.66-1) = 0.2542
The expected hit rate of 45.8% equates to anything over odds of 2.18 being value. Thus it will bet increasingly
large amounts on odds that are above this level. For odds of 3.66, Kelly recommends 25% of the bank being
bet. This is too much to bet on one bet that has odds of 3.66 and an implied probability of 27.3% of winning.
This is where people may increase the Divisor to reduce the stake. For example a Divisor of 2 would halve the
recommended Kelly stake and a Divisor of 10 would cut the recommended Kelly stake to 10% of its original.
The problem here is knowing what the right divisor is and this all based on firstly knowing what your win rate
is.
2 1.3 -134.87% 0 Win 0 167.62
Kelly Equation (((1.3-1)*0.458) – (1-0.458)) / (1.3-1) = -134.87
Here the odds are 1.3. which are under the value odds line 2.18 (45.8% win rate converted to odds). This ends
up with the Kelly recommended stake percentage coming out as negative. This equates to nothing being bet
on this event. This basically means any odds under 2.18 will be ignored. We miss out on a lot of profit by
cutting out any bets with odds under our long term win rate.
156
3 1.27 -154.94% 0 Win 0 167.62
4 7.8 37.83% 37.83 Loss -37.83 129.79
5 2.95 18.01% 18.01 Loss -18.01 111.78
6 2.15 -1.33% 0 Win 0 111.78
7 3.14 20.47% 20.47 Loss -20.47 91.31
8 1.95 -11.25% 0 Loss 0 91.31
9 14.44 41.77% 41.77 Loss -41.77 49.54
Kelly Equation (((14.44-1)*0.458) – (1-0.458)) / (14.44-1) = 0.4177
The extremely high odds of 14.14 in this event come out as a recommendation to bet 41.77% of the bank. This
is an terribly large portion of the bank to bet on one event. If the odds were 1.15 I might understand, but with
odds of 14.44, the chance of winning are actually around 7%. Nobody in their right mind should bet 41% of
their bank on an 7% chance of winning. The Kelly method perceives that you usually win 45.8% of your bets, so
this 14.44 odds bet is Christmas come early if you have a 45.8% chance of winning this bet. But the fact is this
bet is probably not a 45.8% chance of winning, unless you have a ridiculously high edge, which is unlikely. This
is the downfall of Kelly and rules it out for me as being a safe staking method.
10 3.4 23.22% 23.22 Win 55.73 105.27
Positives Negatives
Tries to adjust the stake to correspond to the Any odds under the expected win rate get
potential profit. ignored.
Tries to use historical betting data to work out The stake gets increased as the odds increase
your edge and bet accordingly. above the expected win rate. This is a major
flaw in the method.
Sometimes recommends very high portions
of the bank to be risked on one bet.
If the average wining odds of the Kelly settings are low, then many bets will be staked. A
slight Divisor of 1.2 is necessary. The success here is because the expected win rate is high.
157
18.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Kel5 (D1-EW60-LC) AD1.75 394 2,540 5 4,427,533,847 705 2913 43.46% 0
Kel6 (D2-EW60-LC) AD1.75 543 1,843 5 35,316,920 1821 2017 5.32% 0
Kel4 (D2-EW60-NLC) AD1.75 595 1,681 6 11,489,847 1681 1681 0.00% 24.6
Since the expected win rate equates to a 5% edge, the Kelly Staking plan works very well with
is artificial data set.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Kel8 (D1-EW52.5-LC) AD2 521 1,921 5 60,643,773 1830 2402 14.89% 0
Kel9 (D2-EW52.5-LC) AD2 723 1,383 5 1,456,280 1357 1447 3.25% 0
Kel7 (D2-EW52.5-NLC) AD2 800 1,250 5 579,262 1250 1250 0.00% 15.8
Again the artificial data set works well with the Kelly staking.
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
Kel15 (D1-EW46.7-LC) AD2.25 649 1,540 5 4,323,764 101 2117 65.45% 0
Kel26 (D2-EW46.7-LC) AD2.25 896 1,116 5 228,820 1098 1230 5.91% 0
All the odds are 2.25 and the expected win rate equates to a 5% ROI so the Kelly Method
does well here.
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
Kel16 (D1-EW30-LC) AD3.5 1,294 773 5 21,231 572 1111 34.86% 0
Kel27 (D2-EW30-LC) AD3.5 1,825 548 5 4,463 530 718 17.15% 0
Good results continue.
158
18.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
Kel17 (D1-EW14-LC) AD7.5 3,268 306 5 834 259 518 42.32% 0
Kel28 (D2-EW14-LC) AD7.5 4,808 208 5 423 202 256 12.98% 0
Since the odds are constant at 7.5, just above the win rate, the flaw of Kelly has not yet been
exposed.
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
Kel18 (D1-EW7-LC) AD15 6,849 146 5 275 115 235 41.10% 0
Kel29 (D2-EW7-LC) AD15 10,101 99 5 199 89 149 30.30% 0
Even with the artificial data, Kelly cannot make a working plan for very high odds. This
highlights the difficulty of longshot betting systems.
For mixed odds systems the flaws of Kelly start to appear. All the profiles except one lead to
lows of virtual bankruptcy. The higher odds lead to foolishly high stakes and low odds under
the 49.7% expected win rate are ignored. Effectively we bet more on high odds and nothing
on low odds.
159
18.2 Kelly Staking Plan Summary
For the artificial data, Kelly seems like any other staking plan, good with low odds and poor
with high odds.
Kelly beats Percentage Staking for Type 1-3 with the artificial data, but when applied to real
betting systems it bankrupts Type 1 and 2. It does not do well for the other systems either,
160
but for HRB it does improve on Target Staking’s performance when applied to real betting
systems. Most worryingly, for the mixed system of a wide range of odds it returns an awful
return less than level stakes results and in some simulations loses nearly the entire bank. Kelly
Staking is a method based on a good idea which is to tailor your stakes to the odds and also to
your edge so that you bet more on the events where you are more likely to make a big profit
and to bet less on those events you have a smaller edge in. The problem with this is that it
relies a lot on historical data, knowing your edge and expecting your edge to remain constant.
The fact is that in the real world all three of those aspects are very hard to control. It is a
method with a good idea, but the results are too unpredictable and hard to maintain. For these
reasons I will not replace Type 1 and 2’s recommended staking plan with Kelly and stay with
Percentage staking. If we cannot easily transpose the results for the artificial data to actual
betting system, then it is not a reliable method. Kelly Staking also faces similar problems to
Target Profit Staking in that when the historical average winning odds are very low, the Kelly
Staking plan can end up staking a very large portion of the bank on one bet and that is never
very good. So the Recommended Staking Leaderboard remains unchanged from Chapter 17.
Kelly’s major flaw is that it bets too much on higher odds and nothing on odds below
the historical average win rate.
161
19 LP28 Staking (Long Priced Series 28)
Rating
Overview Introduces an element of recovery to percentage staking.
Concept: To increase stakes very slowly after losses following a fixed series.
Other names: None.
Variables: Base stake level
Link base stake level to cumulative total or not
This plan is a form of progressive recovery staking since it increases stakes after losses, but
while Fibonacci does this with quick, big jumps, LP28 does this with very slow, small
increases. In fact, for the first 10 losses it does not increase the stake size at all, then it moves
up to double the stake for the next 6 losses, then triple stakes over the next 4 losses, quadruple
stakes for the next 3 losses, then 2 bets at x5, 2 bets at x6 and 1 bet at x7. That means to reach
the end of the series and bet 7 times the original stake, there would have had to be 27 losses
beforehand. While it sounds like a very simple and pedestrian method, it produces
surprisingly good results.
The LP28 series looks like this:
1111111111222222333344455667
An easier way to think of it is 10-6-4-3-2-2-1, which is the number of losses until an increase
in stakes occurs. In theory, if you lose 28 bets in a row, it comes to the end of the series and
we reset back to the start. We do not keep staking 7 or more points on bets. If we link the
stake to the cumulative total (percentage staking option) then a multiple of 7 becomes 7% of
the current bank. It is essentially percentage staking with an element of recovery
introduced into it. If a bet wins, the series is reset.
162
20 78.45 4.60 3 2.35 Lose -2.35 76.09
21 76.09 1.25 4 3.04 Lose -3.04 73.05
After 20 losing bets, Bet No. 21 increases the stake to 4% of the bank.
22 73.05 4.15 4 2.92 Lose -2.92 70.13
23 70.13 7.00 4 2.81 Lose -2.81 67.32
24 67.32 2.40 5 3.37 Lose -3.37 63.96
After 23 losing bets, Bet No. 24 increases the stake to 5% of the bank.
25 63.96 4.15 5 3.20 Lose -3.20 60.76
26 60.76 1.56 6 3.65 Lose -3.65 57.11
After 25 losing bets, Bet No. 26 increases the stake to 6% of the bank.
27 57.11 1.40 6 3.43 Lose -3.43 53.69
28 53.69 2.30 7 3.76 Lose -3.76 49.93
After 27 losing bets, Bet No. 28 increases the stake to 7% of the bank.
29 49.93 4.10 1 0.50 Lose -0.50 49.43
After 28 losing bets, Bet No. 29 resets back to the start of the LP28 series and the stake goes back to 1% of the bank.
*If a bet wins, the series is reset to the start. So I have only shown the losing bet series example.
Positives Negatives
Introduces an element of recovery to Progressive recovery staking methods
percentage staking. increase stakes and the risks placed on
individual bets.
Useful if your betting system is based on a For higher odds levels it is not good to be
long term trend and when underperforming placing 4-7% of the bank on one bet.
can be reasonably expected to be due an
upswing in results. For example with betting
systems based on big data trends.
Much less aggressive increases than other loss
recovery staking methods
Good for focussed systems that remain
within a specific band of odds.
163
LP28-S9 (s11-LC) AD1.25 164 6,098 5 227^21 6057 6232 1.43% 0
LP28-S8 (s10-LC) AD1.25 175 5,707 5 15^18 5673 5918 2.15% 0
LP28-S7 (s9-LC) AD1.25 189 5,280 5 784^15 5249 5425 1.67% 0
LP28-S4 (s8-LC) AD1.25 208 4,818 5 31^15 4794 7932 32.57% 0
LP28-S2 (s7-LC) AD1.25 231 4,322 5 1^15 4308 4364 0.65% 0
LP28-S3 (s4-NLC) AD1.25 500 2,000 5 104,857,600 2000 2000 0.00% 0
LP28-S1 (s3-NLC) AD1.25 667 1,500 5 3,276,800 1500 1500 0.00% 0
Staking Plan Name Key:
LP28-8 = Long Priced Series 28 Staking Plan Profile No. 8 s = percentage of bank staked
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
It is important to note that if 16% is the base staking level, then after 10 losses, when the
LP28 multiple goes up to 2, we will be betting 32% of our bank on one bet. If it gets to 16
losses, a multiple of 3 leads to 48% of the bank being staked on one bet. But with the artificial
data base set AD1.25 the longest losing run was 6, and the E.L.S. (Expected losing sequence
based on its odds) was 5. This suggests that in the 1,000 Monte Carlo Simulations there were
some cases thrown up when the losing run went on longer than 10 losses, possibly 16 losses.
It is on those cases when higher stakes were placed on winning bets and those few results
have increased the profits (7,516) above Pure Percentage (7,304 points) at this odds level.
164
Here again LP28-S11 (1,919 points) just beats Percentage Staking (1,915 points). This is
because the artificial data of AD2 has average odds of 2.00 with a 5% stake so it has a win rate
of 52.5%. If it goes on a long losing run, it will definitely come back with wins since it is
created that way. For a real betting system, this would also only be true if the bettors edge is
constant, so that if there is a long losing run, we can be sure that the wins will come back. As
mentioned, when following a betting system based on fixed parameters and selection criteria
we can be more confident that after a long run of losses there will be wins to bring the
winning average back up to its mean. This is one reason why LP28 may work better with big
data based betting systems than with a tipster.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
LP28-S12 (s4-LC) AD2.25 651 1,535 5 4,176,480 519 2001 53.19% 0
LP28-S13 (s3-LC) AD2.25 713 1,403 5 1,672,826 1109 1709 21.38% 0
LP28 (1,535 points) slightly improves on Percentage Staking (1,528 points) at this odds level.
LP28 (1,031 points) makes a significant improvement on Target Profit Staking (775 points) at
this level. It is interesting that in the previous Types 1-4 that linking the stake to the
cumulative profit was the best (LC), but here not linking produces the best result (NLC).
165
19.1.6 Type 6 Systems: High Odds Range 5.00-10.00
For mixed odds LP28 returns to the rule of thumb of 1% of the betting bank. Interestingly it
has a very narrow range of results, only 2.73% which is surprisingly constant.
166
19.2 LP28 Staking Plan Summary
LP28 gets great results for all types of betting systems except those with very high odds ranges
and higher ELS ranges.
Applying the settings to real betting systems mirrors the artificial results closely. That is a
positive sign when looking for a good staking plan.
167
19.4 Recommended Staking Plans Leaderboard
LP28 Staking replaces Types 1, 2, 3, 4 and 5. With such a conservative, long and slow
paced increase, it adds a very light hint of recovery to percentage staking with great results that
slightly beat pure percentage staking.
168
20 D’alembert Staking
Rating
Overview A simple and poor recovery method. The stake increases are too aggressive
making it inefficient for almost all odds levels.
Concept: To increase stakes after a loss and reduce after a win to follow a trend.
Other names: None
Variables: Maximum bet before a reset
Link stake to cumulative total or not
The D’alembert Staking method is very simple. When you lose, you add 1 point to the stake.
When you win, you take away a point from the stake. But stakes cannot go below 1. In theory
stakes could rise infinitely so a maximum limit needs to be placed on the stake increases. Even
by introducing percentage staking to protect the bank there is no real benefit to using this
staking plan.
Positives Negatives
Can take advantage of consistent betting Not sensitive to odds and the implied risk
systems by betting more when under the of each bet.
trend line and returning to normal stakes
when over the trend line.
A maximum stake limit can be set. Progressive recovery staking is very risky as
stakes increase quickly
Doesn’t allow for reducing stakes when over
the trend line, only for increasing when
under.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Dal13 (s5-r5-LC) AD1.25 264 3,781 5 24^12 2674 4514 24.33% 0.7
Dal12 (s4-r5-LC) AD1.25 317 3,155 5 314^9 2515 3595 17.12% 0
Dal10 (s3-r5-LC) AD1.25 409 2,445 5 2,291,835,245 1982 2798 16.69% 0
Dal11 (s3-r5-NLC) AD1.25 544 1,837 6 33,878,254 1628 2012 10.45% 1
Dal7 (s2-r10-NLC) AD1.25 808 1,238 6 533,030 1096 1350 10.26% 0.1
Dal8 (s2-r5-NLC) AD1.25 816 1,225 5 487,099 1076 1346 11.02% 0
Dal3 (s1-r10-LC) AD1.25 1,153 867 6 40,731 731 961 13.26% 0
Dal6 (s1-r50-LC) AD1.25 1,155 866 6 40,450 752 952 11.55% 0
Dal4 (s1-r5-LC) AD1.25 1,171 854 5 37,222 741 941 11.71% 0
Dal1 (s1-r10-NLC) AD1.25 1,618 618 6 7,250 542 680 11.17% 0
Dal5 (s1-r50-NLC) AD1.25 1,618 618 6 7,250 541 682 11.41% 0
Dal2 (s1-r5-NLC) AD1.25 1,631 613 6 7,003 528 672 11.75% 0
Staking Plan Name Key:
Dal8 = D’alembert Staking Plan Profile No. 8 s = percentage of bank staked r = reset level (max stake limit)
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
Most staking plans work with high win rate systems, but D’alambert only achieves half of the
profits LP28 does.
170
20.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave. Bets Ave. Let It Low High Range Bnkrpt
Staking Ave.
to Dble Total Ride Profit Profit from Chnce
Plan ROI
Bank Profit Balance Range Range Ave. % %*
Dal23 (s2-r10-LC) AD1.75 362 2,764 0 20^9 -73788 24730 1782 % 92.6
Dal3 (s1-r10-LC) AD1.75 730 1,369 5 1^6 465 2215 63.92% 0
Dal1 (s1-r10-NLC) AD1.75 781 1,281 6 718,115 768 1748 38.25% 51.2
Dal21 (s0.5-r20-NLC) AD1.75 1,332 751 6 18,228 447 937 32.62% 20
Dal22 (s0.5-r20-LC) AD1.75 NA -1,077 6 0 -171^3 3807 -8142 % 53.2
As soon as odds increase slightly, the plan descends into bankruptcy and inefficiency.
171
20.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Dal42 (s0.05-r30-LC) AD7.5 3,922 255 5 586 -99 761 168.63% 0
Dal41 (s0.05-r50-LC) AD7.5 4,926 203 4 408 -99 913 249.26% 0
Dal43 (s0.05-r15-LC) AD7.5 5,208 192 5 378 -24 408 112.50% 0
Dal46 (s0.01-r15-LC) AD7.5 23,256 43 5 135 3 83 93.02% 0
For higher odds levels the low profit range turns negative and to stay positive the stakes go as
low as 0.01%.
For mixed things Profile 44 returns to profit, but only a third of Secure Staking.
172
Dal3 (s1-r10-LC) AD1.75 730 1,369 5 465 2215 63.92% 500 173%
Dal35 (s0.5-r12-LC) AD2 961 1,041 5 328 1624 62.25% 500 104%
Dal44 (s0.5-r7-LC) AD2.25 1,264 791 5 360 1160 50.57% 500 58%
Dal45 (s0.25-r6-LC) AD3.5 2,519 397 5 165 606 55.54% 500 -20%
Dal46 (s0.01-r15-LC) AD7.5 23,256 43 5 3 83 93.02% 250 -82%
Dal49 (s0.01-r7-LC) AD15 55,556 18 5 -2 43 125% 125 -85%
Dal44 (s0.5-r7-LC) ADmx 1,522 3,943 5 2540 5190 33.60% 2,531 55%
None of the D’alembert staking plans come close to affecting the leaderboard of best staking
plans.
Applied to real betting systems the results match expect for the Type 2 Pro1 which goes
bankrupt. D’alambert is inefficient and not a good staking plan to use.
173
21 Professional Staking
Rating
Overview This is an aggressive recovery staking plan that increases stakes not just to
win back losses but to also achieve a target profit.
Concept: To increase stakes until a target profit level is reached.
Other names: None.
Variables: The target to win on each bet
The Divisor to divide the Target + Losses by
The stop loss level (fixed level or dynamic % of current bank)
Use only integer value of the stake or include the decimal part, too.
The idea of this plan is to win a target profit each bet, for example 1 point each bet. If we lose
a bet, we add the loss to the stake of the next bet and we also add the unachieved target from
the lost bet to the next stake. So the stake increases as losses happen and the number of bets
goes on. This could increase to high levels so a stop loss is necessary.
An important characteristic of this staking plan is that even if you win a bet, you might
increase the stakes of the next bet. For example, if the target is 1 point, but the first bet is on
odds under evens, then even by winning the target has not been reached. The Target of 1
point is then added to Bet 2 which also has a target of 1 point, so the target is now 2 points.
Btu we reduce the remaining target by the profit we got in the first bet. With each bet we add
another point onto the target. This would continue forever if we keep betting on odds-on
games. If we have a divisor of 2 or 3, then betting on odds-on events will take a long time to
reach the target even if we win many bets. It sounds complicated, but it is much easier to
understand by looking at a worked through example that shows a winning run and a losing
run. There is an option to use only the integer part of the stake and drop the decimal part.
This is because some bookmakers get nervous when people bet with decimals and they flag
those users as possibly being professional bettors using a staking plan and they may limit or
ban those users in the fear of making losses on them.
The stake is calculated as the target plus losses divided by the divisor. The divisor’s purpose is
to lower the stake levels and stop them rising too quickly.
174
Example of Professional Staking with 1% stakes (Target of 1 point per bet), Divisor of
2, Reset at 20% of bank.
Target Series
Bet (Remaining) Result Stake Odds Profit Balance Profit Reset
1 1 (1) Win 0.5 1.99 0.5 100.5 0.5
Bet No. 1’s target is 1 point. The stake is (target + losses / divisor) so (1 + 0) / 2 = 0.5.
2 2 (1.5) Win 0.75 1.99 0.74 101.24 1.24
Bet No.2’s target is 2 points, but we already have 0.5 points profit from Bet No. 1’s win, so the remaining target is 1.5
points. 1.5 / the divisor of 2 gives a stake of 0.75.
3 3 (1.76) Win 0.88 2.39 1.22 102.46 2.46
Bet No.3’s target is 3 points, but we already have 1.24 points profit from Bet No. 1 and 2’s win, so the remaining target is
1.76 points. 1.76 / the divisor of 2 gives a stake of 0.88.
4 4 (1.54) Win 0.77 3.57 1.98 104.44 4.44 Reset
Bet No.4’s target is 4 points, but we already have 2.46 points profit from Bet No. 1, 2 and 3’s win, so the remaining target is
1.54 points. 1.54 / the divisor of 2 gives a stake of 0.77. The bet wins with odds of 3.57 and we reach the target of 4 points.
5 1 (1) Win 0.5 3.43 1.22 105.66 1.22 Reset
Bet No. 4 reached its target, so Bet No. 5 is the beginning of a new series with the base target of 1 point. The bet wins with
odds of 3.43 and so the target is reached in one bet.
6 1 (1) Win 0.5 3.38 1.19 106.85 1.19 Reset
Bet No. 5 reached its target, so Bet No. 6 is the beginning of a new series with the base target of 1 point. The bet wins with
odds of 3.38 and so the target is reached in one bet.
7 1 (1) Win 0.5 1.95 0.48 107.33 0.48
8 2 (1.52) Loss 0.76 1.74 -0.76 106.57 -0.28
9 3 (3.28) Loss 1.64 2.08 -1.64 104.93 -1.92
Bet No. 8 lost and the current series is now in a loss. This loss of -0.28 is added to the target of 3 to get a remaining target of
3.28.
10 4 (5.92) Loss 2.96 1.94 -2.96 101.97 -4.88
Bet No. 9 lost and the current series is now in a loss. This loss of -1.92 is added to the target of 4 (because this is the 4th bet
in the current series) to get a remaining target of 5.92
11 5 (9.88) Loss 4.94 2.37 -4.94 97.03 -9.82
12 6 (15.82) Loss 7.91 2.02 -7.91 89.12 -17.73 Reset
Bet No. 11 lost and the current series is now in a loss. This loss of –9.82 is added to the target of 6 (because this is the 6th
bet in the current series) to get a remaining target of 15.82. This is divided by the Divisor of 2 to get a stake of 7.91 which
loses. The total loss is now at -17.73. The next bet will add the target of 7 to the loss and get a new remaining target of
24.73. This figure is over the stop loss of 20% of the bank and so the series is reset. The next bet will begin again from 1.
13 1 (1) Loss 0.5 2.5 -0.5 88.62 -0.5
Positives Negatives
Focusses on the returns so profits can be Complicated and many settings to consider.
more predictable.
Works well with consistent betting systems. Stakes increase quickly.
High odds and a low divisor can lead to Progressive recovery staking plans are
hitting the target quickly. inefficient.
Low odds and a high divisor needs more
time and bets to reach the target.
Adds to the stake side and bets without
adjusting to the odds.
175
21.1 Analysis of Professional Staking
There are a myriad of settings to test with Professional Staking so there are more entries in
this chapter.
177
21.1.4 Type 4 Systems: Over-Evens Range 2.00-2.50
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Pro-S46 (S1-D1-DSL15-
853 1,172 5 337,343 115 2142 86.48% 92.8
Dec-NLC) AD2.25
Pro-S67 (S2-D2.5-DSL15-
878 1,139 5 268,369 -99 1901 87.80% 0
Dec-LC) AD2.25
Pro-S65 (S1-D2.5-DSL15-
1,009 991 5 96,207 453 1503 52.98% 1.4
Dec-NLC) AD2.25
Pro-S64 (S1-D2.5-DSL15-
1,049 953 5 73,929 281 1581 68.21% 0
Dec-LC) AD2.25
Pro-S52 (S0.5-D1-DSL15-
1,236 809 5 27,248 -33 1847 116.19% 61.4
Dec-NLC) AD2.25
Pro-S68 (S0.5-D2.5-DSL15-
1,312 762 5 19,672 288 1277 64.90% 0
Dec-NLC) AD2.25
Pro-S55 (S0.5-D1-DSL15-
1,475 678 5 10,990 -99 1651 129.06% 0
Dec-LC) AD2.25
Pro-S44 (S1-D1-DSL15-
1,754 570 4 5,198 -99 2041 187.72% 0
Dec-LC) AD2.25
After odds go over evens, the staking plans drops into inefficiency and bankruptcy.
178
Pro-S2 (S1-D3.15-FSL10-
1,745 573 6 5,308 195 900 61.52% 41
Dec-NLC) AD3.5
Pro-S21 (S1-D5-FSL10-
2,222 450 6 2,263 187 657 52.22% 10.8
Dec-NLC) AD3.5
Pro-S15 (S0.5-D3.15-
2,421 413 5 1,751 -4 806 98.06% 0.2
DSL10-Dec-NLC) AD3.5
Pro-S25 (S0.5-D5-DSL10-
2,747 364 5 1,247 143 623 65.93% 3.5
Dec-NLC) AD3.5
Pro-S22 (S1-D7.5-FSL10-
3,096 323 5 938 168 490 49.85% 1.6
Dec-NLC) AD3.5
Pro-S20 (S0.25-D5-DSL10-
3,195 313 5 875 -11 613 99.68% 0
Dec-LC) AD3.5
Pro-S18 (S0.25-D5-DSL10-
3,268 306 5 834 19 559 88.24% 0
Dec-NLC) AD3.5
Pro-S17 (S0.25-D3.15-
3,300 303 5 817 -22 664 113.20% 0
DSL10-Dec-NLC) AD3.5
Pro-S19 (S0.25-D3.15-
3,333 300 5 800 -34 666 116.67% 0
DSL10-Dec-LC) AD3.5
Pro-S23 (S1-D10-FSL10-
3,891 257 6 594 122 392 52.53% 0.1
Dec-NLC) AD3.5
Pro-S24 (S1-D12.5-FSL10-
4,808 208 6 423 62 302 57.69% 0
Dec-NLC) AD3.5
Pro-S14 (S0.02-D3.15-NSL- - 73611.11
12,346 81 10 175 183 63.7
Dec-NLC) AD3.5 119067 %
Pro-S16 (S0.02-D3.15-
18,519 54 5 145 -62 161 206.48% 0
DSL10-Dec-NLC) AD3.5
All the results at this level result in low profit ranges in negative numbers.
180
Pro-S78 (S0.25-D2.4-
1,908 3,144 5 291,328,561,000 1531 4781 51.69% 30.2
DSL15-Dec-NLC) ADmx
Pro-S84 (S1-D5-DSL10-
1,908 3,145 5 293,354,911,389 2182 3995 28.82% 0
Dec-LC) ADmx
Pro-S83 (S1-D5-DSL10-
1,935 3,101 5 216,242,057,801 2410 3914 24.25% 0.2
Dec-NLC) ADmx
Pro-S74 (S0.5-D2.4-DSL15-
2,102 2,855 5 39,301,251,141 -99 5751 102.45% 0
Dec-LC) ADmx
Pro-S82 (S0.5-D5-DSL10-
2,233 2,687 5 12,265,238,102 1909 3428 28.27% 0
Dec-NLC) ADmx
Pro-S79 (S0.25-D2.4-
2,350 2,553 5 4,845,022,400 -98 4608 92.17% 0
DSL15-Dec-LC) ADmx
Pro-S18 (S0.25-D5-DSL10-
2,715 2,210 5 449,534,372 1507 2857 30.54% 0
Dec-NLC) ADmx
Pro-S80 (S0,1-D1-DSL15-
2,751 2,181 5 367,675,088 88 4320 97.02% 99.9
Dec-NLC) ADmx
Pro-S75 (S0.25-D1-DSL15-
13,100 458 3 2,392 -98 2854 322.27% 0
Dec-LC) ADmx
For mixed odds Professional achieves only a third of the profit of Secure Staking.
None of the Professional Staking plans affect the leaderboard which stays the same as at the
end of Chapter 19.
181
21.3 Applying to Real Betting Systems
Ave. Rang Ave.
Staking Ave. "Let It Low High Ave. Level
Bets Ave. from Brpt Bets
Plan Total Ride" Prof. Prof. Mnths Stakes
to Dbl ROI Ave. % per
Profile Profit Bal. Rnge Rnge to Dble Increase
Bank % Mnth
Pro-S39 (S4-
D1-DSL15- 125 1,151 8 291,645 805 1491 29.80% 0 19 7 946%
Dec-LC) Col3
Pro-S62 (S5-
D1.5-DSL15- 260 763 8 19,809 375 1230 56.03% 0 24 11 429%
Dec-LC) Pro1
Pro-S66 (S6-
D2-DSL15- 343 823 7 30,025 304 1154 51.64% 0 34 10 330%
Dec-LC) Col5
Pro-S64 (S1-
D2.5-DSL15-
Dec-LC) 208 339 21 1,048 155 505 51.62% 0 9 23 138%
BTMO2.5
Pro-S18
(S0.25-D5-
DSL10-Dec- 1,763 396 9 1,556 168 609 55.68% 0 84 21 -33%
NLC) Col17
Pro-S95
(S0.5-D13-
DSL10-Dec- 1,167 121 20 231 26 202 72.73% 0 18 65 -11%
NLC) Pro8
Pro-S73 (S1-
D28-DSL15- 182.43 2%
6,254 74 6 167 -49 221 0 165 38
Dec-NLC) %
HRB
Pro-S107 (S4-
D5-DSL10- 977,049, 189%
660 2,322 10 1944 2630 14.77% 0 199 3
Dec-LC) 348
Mix6
In a positive sign for my research method, the staking plans artificial data results are largely
mirrored when applied to real betting systems. Odds on systems and mixed odds do okay, but
others are still inefficient.
182
22 Coup Master Staking
Rating
Overview A good staking method for Types 1-4 that focuses on increasing the bank
by a certain percentage each day.
Concept: To break a daily target into stages or “coups”.
Other names: None.
Variables: Daily Target as percentage of bank
Divisor based on your average winning odds
Number of Coups per day to divide the target into
Stop loss limit
The Coup Master aims to increase the betting bank by a set percentage each day. This target is
then broken down into smaller legs or “coups”. Two coups are recommended but it is
possible to set more. On top of this we adjust the stake based on the remaining target. Wins
are deducted from the remaining target and losses are added to it. A minimum stake level is
set as the same as the current coup target. This avoids the stakes going too low when in
profit. Since this staking plan involves progressive recovery staking, a stop loss is also needed
to stop stakes getting too high. The stake is also divided by a Divisor which is based on the
betting systems average winning odds. This is to make sure that each bet has a chance of
clearing the Coups. For this reason this staking method can become complicated because it
relies on knowing your average winning odds. It also needs tracking carefully to know where
you are in the coups and targets. The stake is the remaining target divided by the Divisor,
unless that number is lower than the Coup target in which case the coup target is the stake.
183
the Divisor. 5.5 / 1.4 is 3.92 but the minimum stake is also the target, so instead of staking 3.92 we stake 5.5. The bet
wins and the profit is 2.75 points. 2.75 of Coup 1’s target remains.
4 1.5 5.5 5.5 Win 2.75 Landed ! 5.5 115.5 0
The target is still 10% of a 110 point bank. Bet No.3 achieved half of Coup 1’s target of 5.5 and now Bet No. 4 tries to
get closer to the target. The stake is the remaining target divided by the Divisor. 5.5 / 1.4 is 3.92 but the minimum stake is
also the target, so instead of staking 3.92 we stake 5. The bet wins and the profit is 2.75 points. This hits the target and
Coup 1 is “Landed”.
5 1.5 5.5 5.5 Win 2.75 2 2.75 118.25
6 1.5 5.5 5.5 Win 2.75 Landed ! 6.05 121 0
7 1.95 6.05 6.05 Loss -6.05 1 12.1 114.95 0
Bets No. 5 and 6 won and achieved Coup 2 and reached the 10% target. We start a new day.
The target is 10% of a 121 point bank. That is 12.1 points. We have 2 coups, so the target is divided in to 2 targets of 6.05
points each. Bet No. 7 is trying to achieve the target for Coup 1 of 6.05 points. The stake is the remaining target divided
by the Divisor. 6.05 / 1.4 is 4.32 but the minimum stake is also the target, so instead of staking 4.32 we stake 6.05. The
bet loses and the loss is 6.05 points. That loss is added to the remaining target and becomes 12.1.
8 1.74 6.05 8.64 Loss -8.64 1 20.74 106.31 0
The target is still 10% of a 121 point bank. That is 12.1 points. We have 2 coups, so the target is divided in to 2 targets of
6.05 points each. Bet No. 7 lost and the remaining target is now 12.1. The stake is the remaining target divided by the
Divisor. 12.1 / 1.4 is 8.64. This shows how after losses occur, stakes increase to recover. The bet loses and the of-8.64 is
added to the remaining target and becomes 20.74.
9 2.08 6.05 14.81 Loss -14.81 Stop Loss 4.57 91.5 0
The target is still 10% of a 121 point bank. That is 12.1 points. We have 2 coups, so the target is divided in to 2 targets of
6.05 points each. Bets No. 7 and 8 lost and the remaining target is now 20.74. The stake is the remaining target divided by
the Divisor. 20.74 / 1.4 is 14.81. This shows how after losses occur, stakes increase to recover. The bet loses and the of -
14.81 is added to the remaining target and becomes 35.55. That would mean the next stake would be 35.55 / 1.4 = 25.39
which is over our 20% stop loss trigger for stake levels. So the series is abandoned and we start again from a new Coup 1
on the next bet.
10 1.94 4.57 4.57 Win 4.3 1 0.27 95.8
The balance at the end of Bet No. 9 was 91.5. so the new target is 10% of that, which is 9.15. The target for each of 2
coups becomes 4.57.
11 2.37 4.57 4.57 Win 6.26 Landed ! -1.42 102.06 5.99
Bet No. 10 and 11 win and reach the target for Coup 1. There is also a surplus of 5.99 leftover after the target is reached.
This is carried over to the next Coup and the remaining target of 4.57 becomes -1.42.
12 2.02 4.57 4.57 Loss -4.57 2 3.15 97.49 0
Bet No. 12 loses and the loss of 4.57 adds to the remaining target of -1.42 to become 3.15.
13 2.5 4.57 4.57 Loss -4.57 2 7.72 92.92 0
Bet No. 13 loses and the loss of 4.57 adds to the remaining target of 3.15 to become 7.72.
14 3.42 4.57 5.51 Loss -5.51 2 13.23 87.41 0
. The stake is the remaining target divided by the Divisor. 7.72 / 1.4 is 5.51.
15 3.3 4.57 9.45 Loss -9.45 Stop Loss 3.9 77.96 0
The stake is the remaining target divided by the Divisor. 13.23 / 1.4 is 9.45. Bet No. 15 loses and that loss takes the
remaining target to 22.68 which divided by the Divisor is 16.2 which is over the 20% stop loss of our current 77.96 bank.
16 1.95 3.9 3.9 Loss -3.9 1 7.8 74.06 0
. Having hit the Stop Loss in Bet No. 15 we start a new Coup 1 based on the 77.96 bank balance.
Positives Negatives
Works well for lower odds Types 1-4. Doesn’t work for higher odds Types 5-7.
Need to know your betting history and average
winning odds to decide the Divisor.
Complex and needs a record of bets to be kept.
184
22.1 Analysis of Coup Master Staking
186
22.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
CM-44 (t1-c2-d16.5-SL5) AD7.5 4,184 239 5 524 -75 597 140.59% 0
CM-45 (t1-c2-d6.5-SL10) AD7.5 4,630 216 4 447 -99 701 185.19% 0
CM-46 (t2-c2-d6.5-SL10) AD7.5 4,975 201 3 403 -99 9001 2263% 0.2
CM-50 (t0.5-c2-d6.5-SL5) AD7.5 6,452 155 5 293 -79 411 158.06% 0
CM-57 (t0.1-c2-d6.5-SL5) AD7.5 24,390 41 5 133 -77 107 224.39% 0
Similar to Type 6, all low profit ranges are negative, even with low starting stakes.
Even for mixed systems, the results are almost all on the verge of bankruptcy.
187
22.2 Percentage Stakes Plan Summary
For Type 2 betting systems Coup Master (2,497 points) comes very close to knocking LP28
(2,498 points) off its spot. Type 3 is also very close with Coup Master on 1,918 points and
LP28 on 1,919 points. Type 1 and 4 also do very well, but as the odds get higher, types 5-8
become inefficient.
Coup Master Staking performs better with the real betting systems than the artificial data sets.
But the higher level odds still show inefficiencies and some Monte Carlo Simulations ending
lower than level stakes. At lower odds brackets this staking plan may have some use though.
188
23 Labouchere Staking
Rating
Overview A recovery method that focusses on stakes used instead of odds and
profit, which leads to rapidly rising stakes and bankruptcy.
Concept: Start with a staking series of progressive numbers. Add lost stakes to the
staking series. Cancel out stakes when won.
Other names: Cancellation System
Variables: Numbers used in the series
Maximum stake limit
The Labouchere staking method was originally created for 50-50 table games such as the red-
black bet at casinos, which is one reason why it may not be so worried about focusing on the
stakes instead of the odds. It starts with a series of numbers which will be the stakes we bet,
for example 1-2-3-4-5-6. The stake is the first number added to the last number, in this case 1
+ 6. If the bet wins, the numbers are cancelled from the series, for example 2-3-4-5. But if the
bet loses, the lost stake of 7 is added to the end of the series, for example 1-2-3-4-5-6-7. This
means the next bet is 7 +1. The stake has gone up 1 point. The series ends after we cancel out
all of the stakes in the series. When using this method in sports betting the odds vary and then
this plan does worse. Another way to think of the problem of a stake focused plan is that we
are focusing on spending money without regard for what we spend it on and what we want to
get from spending that money. We are not bureaucrats trying to spend our entire budget just
for the sake of it. It would be much better to focus on risks and outcomes than on the
input.
189
7 3-4-5-7-10-13 16 1.95 Loss -16 80.2
Bet No.6 loses, so the lost stake of 13 is added to the end of the series.
The new stake is the first plus the last number in the current series. 3 + 13 = 16. An increase of 3.
8 3-4-5-7-10-13-16 19 1.74 Win 14.06 94.26
Bet No.7 loses, so the lost stake of 16 is added to the end of the series.
The new stake is the first plus the last number in the current series. 3 + 16 = 19. An increase of 3.
9 4-5-7-10-13 17 2.08 Win 18.36 112.62
Bet No. 8 wins, so the first and last numbers in the series are cancelled.
The new stake is the first plus the last number in the current series. 4 + 13 = 17. A decrease of 2.
10 5-7-10 15 1.94 Win 14.1 126.72
Bet No. 9 wins, so the first and last numbers in the series are cancelled.
The new stake is the first plus the last number in the current series. 5 + 10 = 15. A decrease of 2.
11 7 7 2.37 Win 9.59 136.31
Bet No. 10 wins, so the first and last numbers in the series are cancelled.
The new stake is the first plus the last number in the current series, but there is only one number left 7. A decrease of 8.
This bet wins and the series ends. The next bet will begin a new series.
Positives Negatives
Sounds French. Stake focused and ignores odds and profit targets. You could end up
betting a large amount on high odds.
Stakes rise very quickly.
To avoid bankruptcy initial stakes have to be low, leading to low profits.
Increases in stake are larger than the decreases.
Slightly complex and requires notes to know where you are in the series.
Labouchere’s aggressive recovery system results in bankruptcy and rapid stake rises which
need lower base stake levels. This push down to 0.1% of stakes leads to lower profits than
level stakes.
No good news.
191
23.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
LaB112(s0.01-L1-H6-M21) AD7.5 15,873 63 6 155 31 96 51.59% 0
The range of results is also very wide at 51%, another sign of a poor plan.
192
23.3 Applying to Real Betting Systems
Ave. Rang Ave. Ave.
Ave. "Let It Low High Level
Staking Bets Ave. from Brpt Bets Mths
Total Ride" Prof. Prof. Stakes
Plan Profile to Dbl ROI Ave. % per to
Profit Bal. Rnge Rnge Increase
Bank % Mnth Dble
LaB9 (L1-H3-M6)
381 376 8 1,355 308 408 13.30% 0 19 20 1508%
Col3
LaB11 (s0.1-L1-
1,479 134 8 253 93 174 30.22% 0 24 62 429%
H6-M21) Pro1
LaB11 (s0.1-L1-
1,476 191 7 376 127 256 33.77% 0 34 43 381%
H6-M21) Col5
LaB112(s0.01-L1-
H6-M21) 5,036 14 21 110 10 18 28.57% 0 9 560 413%
BTMO2.5
LaB112(s0.01-L1-
10,265 68 9 160 52 84 23.53% 0 84 122 25%
H6-M21) Col17
LaB112(s0.01-L1-
4,413 32 20 125 19 43 37.50% 0 18 245 33%
H6-M21 Pro8
LaB12 (s0.01-L1-
13,223 35 7 127 4 68 91.43% 0 165 80 59%
H6-M21) HRB
LaB12 (s0.01-L1-
9,828 156 10 295 133 174 13.14% 0 199 49 138%
H6-M21) Mix6
The poor results of the artificial data are mirrored in the real betting systems.
193
24 Reverse Labouchere Staking
Rating
Overview A stakes focused plan that seeks to take advantage of winning runs.
Concept: Start with a staking series of progressive numbers. Add winning stakes to
the staking series. Cancel out stakes when lost.
Other names: Reverse Cancellation System
Variables: Numbers used in the series
Maximum stake limit
It is often a useful thing to try the reverse of a method to see if it performs better. Both
versions of Labouchere use a series of numbers which will be our stakes, for example 1-2-3-4-
5-6. The next stake is the first and last number added together, in this case 1 + 6 = 7. Normal
Labouchere adds lost stakes to the end of the series and cancels out the first and last numbers
if it wins. It continues until we win enough to cancel out all the stakes. Reverse Labouchere
does the opposite, if a bet wins, we add the stake to the end of the series. If the bet loses, we
cancel the numbers at the beginning and end of the series. In this way we keep a certain
control on losses.
194
9 5-6 11 2.08 Loss -11 98.06
Bet No. 8 loses, so the first and last numbers in the series are cancelled.
The new stake is the first plus the last number in the current series. 5 + 6 = 11. No change.
10 1-2-3-4-5-6 7 1.94 Loss -7 91.06
Bet No. 9 loses, so the first and last numbers in the series are cancelled.
The series has been cancelled out so a new series is begun.
11 2-3-4-5 7 2.37 Win 9.59 100.65
Positives Negatives
Works better than normal Labouchere for Normal Labouchere is still better for Type 1
Types 2, 3 and 4. (1.50-2.50 odds range) (1.01-1.49 odds range) than Reverse
Labouchere.
More conservative than normal Labouchere. Neither Labouchere plan works for Type 5
and above.
Increasing stakes force the base stake down
and overall results are poor in comparison to
other staking plans.
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
RLab2 (s0.1-L1-H6-M21) AD1.25 1,605 623 5 7,506 574 679 8.43% 0
Staking Plan Name Key:
RLab8 = Reverse Labouchere Staking Plan Profile No. 8 s = percentage of bank staked
L1 = lowest number in series 1 H7 = highest number in series 7 M10 = Maximum number of units to bet 10
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
Higher stakes lead to bankruptcy and the best settings to remain efficient with a 100 point
bank have a starting stake level of 0.1%. That leads to an average total profit that is only
slightly above level stakes betting.
The average total profit is only 3 points above simple level stakes.
The aggressive recovery nature of the plan suppresses stakes to 0.01% and very low profit
levels.
197
25 Bookies Bank Staking
Rating
Overview A recovery plan that is less aggressive than others.
Concept: To stake a percentage of the Bookies Bank, not own bank when losing.
Other names: None.
Variables: Percentage of bank to bet initially
Link ongoing stake to current total or not
The concept behind this staking plan is very good and it is surprising that it does not get
better results. Often the problem with recovery plans is that the stakes increase too quickly
because the stakes are focusing on the current losses. In Bookies Bank, instead of trying to
win back a percentage of our losses, we add our starting bank to the losses and stake a
percentage of that larger number. This greatly reduces the size of the stakes. When we
are in profit, we revert to our original stake or a percentage of the current bank, depending on
our settings. Unfortunately, as is the case with other recovery plans (except LP28), because we
have to increase stakes later on in recovery mode, the initial stakes are pushed lower, which
leads to a lower starting point and overall lower profits.
198
12 2.02 5 Win 5.1 106.6 -5.1 94.9
The Bookie is not in profit, so the next stake is 5% of our original 100 point bank, not our current 101.5 bank. 100 * 0.5= 5
(If we were using the percentage staking option, the next stake would be 5% of the current balance after Bet No. 11, 101.5.)
13 2.5 5 Loss -5 101.6 5 105
Positives Negatives
Increases recovery stake size slowly. Stakes increase, pushing down initial stake levels.
Odds and risk not taken into account.
A little complex and needs record keeping to know
what the next stake should be.
Not efficient for higher odds bands.
Bookies Bank shows positive results with 2% stakes. But the resulting profits are much lower
than other staking plans.
199
25.1.3 Type 3 Systems: Evens Range 1.90-2.10
Ave. Bets Ave. Low High Range Bnkrpt
Staking Ave. Let It Ride
to Dble Total Profit Profit from Chnce
Plan ROI Balance
Bank Profit Range Range Ave. % %*
BB1 (s1-NLC) AD2 1,828 547 6 4,432 532 550 1.65% 0.3
BB2 (s1-LC) AD2 1,267 789 6 23,721 752 792 2.53% 1.3
BB7 (s0.5-NLC) AD2 3,817 262 6 615 257 263 1.15% 0
BB11 (s0.5-LC) AD2 2,688 372 5 1,318 362 385 3.09% 0
200
25.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
Ave. Bets Ave. Low High Range Bnkrpt
Staking Ave. Let It Ride
to Dble Total Profit Profit from Chnce
Plan ROI Balance
Bank Profit Range Range Ave. % %*
BB9 (s0.25-NLC) AD15 6,061 165 6 314 66 166 30.30% 13.6
BB10 (s0.1-LC) AD15 12,821 78 6 172 48 88 25.64% 0
BB14 (s0.1-NLC) AD15 17,857 56 6 147 47 61 12.50% 0
Bookies bank is worth considering for low odds, high strike rate systems. Bookies Bank seems
to have a novel approach to recovery staking, but in the end the results for higher odds are
inefficient, which is probably why someone created Bookies Bank Version 2 (See Chapter 27.)
201
25.3 Applying to Real Betting Systems
Ave. Rang Ave.
Staking Ave. "Let It Low High Ave. Level
Bets Ave. from Brpt Bets
Plan Total Ride" Prof. Prof. Mnths Stakes
to Dbl ROI Ave. % per
Profile Profit Bal. Rnge Rnge to Dble Increase
Bank % Mnth
BB6 (s2-LC)
446 321 8 925 315 321 0.93% 0 19 23 191%
Col3
BB1 (s1-
1,304 152 8 287 147 153 1.97% 0 24 54 5%
NLC) Pro1
BB11 (s0.5-
2,104 134 7 253 129 135 2.24% 0 34 62 -29%
LC) Col5
BB11 (s0.5-
LC) 684 103 21 204 102 104 0.97% 0 9 76 -27%
BTMO2.5
BB10 (s0.1-
8,410 83 9 178 82 85 1.81% 0 84 100 -86%
LC) Col17
BB10 (s0.1-
4,413 32 20 125 31 33 3.13% 0 18 245 -76%
LC) Pro8
BB10 (s0.1-
12,856 36 6 128 18 38 27.78% 0 165 78 -50%
LC) HRB
BB10 (s0.1-
7,335 209 10 426 208 211 0.72% 0 199 37 -73%
LC) Mix6
The poor artificial results are reflected when the staking plan settings are applied to real
betting systems.
202
26 Phantom Bet Staking
Rating
Overview An interesting idea that manages to increase the ROI, but unfortunately
reduces overall profits due to a smaller bet volume.
Concept: To skip bets after a winner
Other names: None
Variables: How many bets to skip.
Link stake to cumulative total or not
Phantom staking begins with the very simple thought that “you can’t win them all”, and it is
less likely to get two winners in a row. In that case why not just skip the second bet after a
winner and save the stake. It seems like an idea with merit. The actual outcome is that we do
indeed back more winners than losers in this method, but since overall we are making fewer
bets, we make less profit than just following level stakes. One interesting result of phantom
staking is that since the bets we do make are more likely to be winners, the ROI goes up for
each bet. And of course, this all only works with the premise that the betting system being
used follows a trend or has a consistency that we can expect a certain number of winners and
losers to come. This may not work as well for a mixed betting system, but would work better
for systems attached to fixed selection criteria such as systems created from big data research.
Phantom Staking follows the idea of betting less after a winner which is a little similar to 4321
Staking from Chapter 10, which did quite well. Overall, betting less after winners seems better
than betting more after losers in many cases.
203
Positives Negatives
Increases the ROI of bets Reduces overall profits
Phantom bets produce positive results but the overall profits are reduced
204
26.1.3 Type 3 Systems: Evens Range 1.90-2.10
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
PhB-S31 (s2-Ph0-C) AD2 853 1,172 5 337,343 1153 1189 1.54% 0
PhB-S30 (s2-Ph1-C) AD2 1,294 773 5 21,231 385 1218 53.88% 0
PhB-S27 (s1-Ph0-C) AD2 1,546 647 5 8,865 642 672 2.32% 0
PhB-S23 (s1-Ph0-NC) AD2 2,000 500 5 3,200 500 500 0.00% 0
PhB-S26 (s1-Ph1-C) AD2 2,358 424 5 1,890 213 645 50.94% 0
PhB-S20 (s1-Ph1-NC) AD2 3,040 329 6 978 196 504 46.81% 0
205
26.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
PhB-S23 (s1-Ph0-NC) AD15 2,000 500 5 3,200 500 500 0.00% 1000
PhB-S20 (s1-Ph1-NC) AD15 2,146 466 5 2,528 144 720 61.80% 99.4
PhB-S27 (s1-Ph0-C) AD15 10,000 100 3 200 -99 601 350.00% 0
PhB-S26 (s1-Ph1-C) AD15 11,494 87 2 183 -99 711 465.52% 0
Phantom bets increase the yield of each bet slightly but overall the profit levels are reduced
due to reduced number of bets.
206
26.3 Applying to Real Betting Systems
Ave. Rang Ave. Ave.
Staking Ave. "Let It Low High Level
Bets Ave. from Brpt Bets Mnths
Plan Total Ride" Prof. Prof. Stakes
to Dbl ROI Ave. % per to
Profile Profit Bal. Rnge Rnge Increase
Bank % Mnth Dble
PhB-S29 (s8-
Ph1-C) AD1.25 236 607 8 6,718 268 909 52.80% 0 19 12 451%
Col3
PhB-S22 (s3-
Ph1-C) AD1.75 610 325 8(7) 951 32 582 84.62% 0 24 25 125%
Pro1
PhB-S30 (s2-
Ph1-C) AD2 921 306 7 834 85 517 70.59% 0 34 27 60%
Col5
PhB-S30 (s2-
Ph1-C) AD2.25 281 251 21(20) 570 130 376 49.00% 0 9 31 76%
BTMO2.5
PhB-S26 (s1-
Ph1-C) AD3.5 1,246 560 9 4,850 219 816 53.30% 0 84 15 -6%
Col17
PhB-S26 (s1-
Ph1-C) AD7.5 513 275 20(19) 673 107 422 57.27% 0 18 29 102%
Pro8
PhB-S26 (s1-
Ph1-C) AD15 4,207 110 4(6) 214 -96 601 316% 0 165 25 52%
HRB
PhB-S26 (s1-
Ph1-C) Mix6 1,164 1,317 10(9) 921,648 964 1622 24.98% 0 199 6 64%
Mix6
I have highlighted in yellow the places where the Phantom Staking has improved the ROI
over flat staking. The brackets show the flat staking ROI. Unfortunately, my software only
returns integers for the ROI, so while the other systems return the same ROI by integer, I
suspect they also improved the ROI if we could see the decimals. The HRB system, however
returns a worse ROI.
207
27 Bookies Bank Version 2 Staking
Rating
Overview Martingale combined with Target Profit Staking to recover losses.
Concept: Raise the stake just enough to win back all losses in the current series.
Other names: Odds adjusted Martingale, Target Martingale
Variables: Number of losing bets allowed until resetting the series
Link the stake to the cumulative bank or not
In Bookies Bank Version 1 we were staking a percentage of the Bookies Bank when we were
losing. It focused on the input side. In Bookies Bank Version 2 we are focusing on the output
side, and we focus on betting just enough to win back the losses so far in the current series
plus the winnings of our standard stake. That means that even if our bank is in overall profit,
and we have a short losing run, stakes increase to win back the losses from just the current
losing series. Since stakes can increase infinitely, we need a limit on how many losses in a row
we will try to win back before cutting our losses and accepting defeat. Depending on the reset
limit used, this can be a very aggressive and risky system just like Martingale.
208
Positives Negatives
Adjusts the stake to the odds. Still a version of Martingale and stakes can
increase infinitely unless a limit placed.
Has a reset limit to stop stakes getting Since stakes increase later, the initial stake is
too high. forced lower and thus overall profits are lower.
209
27.1.3 Type 3 Systems: Evens Range 1.90-2.10
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
BBv2-S8 (s4-R2-LC) AD2 481 2,079 5 181,306,598 1 3151 75.76% 0
BBv2-S9 (s5-R2-LC) AD2 528 1,894 5 50,293,067 -99 3351 91.08% 0
BBv2-S7 (s3-R2-LC) AD2 603 1,659 5 9,864,785 942 2314 41.35% 0
BBv2-S12 (s3-R3-LC) AD2 871 1,148 5 285,644 -99 2801 126.31% 0
210
BBv2-S15 (s0.5-R2-LC) AD7.5 3,846 260 5 606 196 348 29.23% 0
BBv2-S10 (s2-R2-LC) AD7.5 6,098 164 3 312 -99 1201 396.34% 0
Bookies bank Version 2 performs much better than the original Bookies Bank. This is because
it has an element of Target Profit Staking and adjusts stakes to the odds.
211
27.3 Applying to Real Betting Systems
Ave. Rang Ave.
Staking Ave. "Let It Low High Ave. Level
Bets Ave. from Brpt Bets
Plan Total Ride" Prof. Prof. Mnths Stakes
to Dbl ROI Ave. % per
Profile Profit Bal. Rnge Rnge to Dble Increase
Bank % Mnth
BBv2-S8
(s4-R2-LC) 187 765 8 20,085 121 1265 74.77% 0 19 10 595%
Col3
BBv2-S8
(s4-R2-LC) 318 624 6 7,558 73 1154 86.62% 0 24 13 333%
Pro1
BBv2-S7
(s3-R2-LC) 369 763 7 19,809 388 1166 50.98% 0 34 11 299%
Col5
BBv2-S10
(s2-R2-LC) 147 481 21 2,805 388 588 20.79% 0 9 16 238%
BTMO2.5
BBv2-S6
(s1-R4-LC) 673 1,037 9 132,337 588 1388 38.57% 0 84 8 73%
Col17
BBv2-S16
(s0.5-R3- 648 218 19 453 157 255 22.48% 0 18 36 60%
LC) Pro8
BBv2-S15
(s0.5-R2- 4,821 96 4 195 30 318 150% 0 165 29 33%
LC) HRB
BBv2-S4
(s1-R2-LC) 645 2,377 9 1^9 2102 2641 11.34% 0 199 3 196%
Mix6
As the results show, it is possible to get this Martingale variation under control by using limits
on the number of losses to follow.
212
28 Bookies Bank Version 2 with Phantom
Losing Bets
Rating
Overview To bet more after a winner by aggressively recovering losses we didn’t
actually have.
Concept: Insert Phantom losing bets to increase the target at the beginning of each
series.
Other names: None
Variables: Number of phantom bets after a winner
Odds of the phantom bets
Rests limit on the number of losses to recover
Link the base stake to a percentage of the current balance or not
Bookies Bank Version 2 with Phantom Bets is where after each winning bet we can choose to
insert a specified number of losing bets. We can also specify the odds of those losing bets.
The point being that we don’t actually bet real money on those phantom bets, but we increase
the series loss as if we had bet and lost on those phantom bets. What this leads to in practical
terms is that after a winner, we bet more. As we have seen in 1234 Winners Staking, that is not
a good way to bet.
213
places its own standard 1 point stake. So the total stake on Bet No. 5 is 4.
6 2 8 Loss -15 -8 93
7 2 16 Loss -31 -16 77
8 2 32 Loss -63 -32 45
9 2 4 Loss -7 -4 41
10 2 8 Win 1 8 49
Positives Negatives
Varies stakes according to the odds of a Phantom losing bets increase the size of the series
bet. deficit from the start and leads to stakes rising
faster. This pushes down the base level stake and
ultimately overall profits.
Complicated to use.
An over-complex way to bet more after a win and
then go into recovery mode.
Profile 51 shows a very wide range of results from the average, 63%. This reflects the gamble
we take by increasing stakes artificially after a win.
214
28.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
BBv2-S8 (s4-R2-LC) AD1.75 478 2,093 5 199,782,668 101 3269 75.68% 0
BBv2PH S53 (s0.5-r2-LC-Ph2) AD1.75 542 1,844 5 35,562,568 1072 2422 36.61% 0
BBv2PH S51 (s1-r2-LC-Ph1) AD1.75 588 1,700 5 13,107,200 1144 2200 31.06% 0
BBv2PH S52 (s1-r3-LC-Ph1) AD1.75 659 1,518 5 3,712,234 1 2761 90.91% 0
BBv2-S4 (s1-R2-LC) AD1.75 1,135 881 5 44,882 662 1122 26.11% 0
BBv2PH S50 (s4-r2-LC-Ph1) AD1.75 3,106 322 3 932 -99 2449 395% 0
215
BBv2PH S58 (s1-r4-LC-Ph1) AD3.5 1,511 662 5 9,836 -99 1245 101 % 0
BBv2-S6 (s1-R4-LC) AD3.5 1,520 658 5 9,567 226 1119 67.86% 0
BBv2PH S53 (s0.5-r2-LC-Ph2) AD3.5 1,678 596 5 6,225 431 761 27.68% 0
BBv2PH S56 (s0.5-r3-LC-Ph1) AD3.5 1,957 511 5 3,454 293 698 39.63% 0
216
28.2 Bookies Bank Version 2 with Phantom Losing Bets Summary
217
29 Recovery Type 1 Staking
Rating
Overview The stakes increase too much in this plan, meaning the initial stake levels
are too low and overall profit is reduced compared to other staking plans.
Concept: To recover a percentage of losses split over the next few bets
Other names: Recovery Leg Staking
Variables: Number of losses to recover
Percentage to recover
Number of legs to split the recovery cycle into
Recover losses while also betting regular stakes or just bet to recover losses
Link stake to current balance
The recovery staking plan does exactly as the name suggests but with many options and
settings. It tries to recover a percentage of a loss over a specifiable number of bets. It can do
this for multiple numbers of losses. So you could be recovering one loss over the 4 bets and if
you lose a bet during that 4 bet stretch, that starts a new “cycle” and you now have 2 lost bets
being recovered in a new 4-leg cycle of recovery bets. The recovery is done by splitting the
lost stake over the next few bets on top of the regular base stake. For example if you lose 1
point on a bet, and have set a 4-leg recovery cycle, it will bet 1.25 points on the next 4 bets.
218
9 2.08 1 Loss -1 -1 106.1
10 1.94 1.5 Loss -1.5 -2.5 104.6
Bet No. 9 lost, so the current loss of -1 point will be split into 2 legs of 0.5 stakes each. Add this to the original
stake for Bet No.10 and the stake becomes 1.5.
11 2.37 2.25 Loss -2.25 -4.75 102.35
Bet No.10 lost, so the current loss of -2.5 points will be split into 2 legs of 1.25 point stakes each. Add this to the
original stake for Bet No.11 and the stake becomes 2.25. This is now a 2nd cycle.
12 2.02 3.38 Loss -3.38 0 98.97
Bet No.11 lost, so the current loss of -4.75 points will be split into 2 legs of 2.38 point stakes each. Add this to the
original stake for Bet No.12 and the stake becomes 3.38. This is now a 3rd cycle. This bet loses, but we cannot start
a new 4th cycle because our limit is 4 losses. The deficit resets to zero and the next bet begins fresh at 1 point stakes.
13 2.5 1 Loss -1 -1 97.97
Positives Negatives
Does not try to recover all losses in one Splits the lost stake into equal stakes without
bet. considering the odds of each bet.
Has a limit on the number of losses it Rising stakes mean the initial stake levels are
recovers. lower, and so overall profits are lower.
A little complicated to use. Requires record
keeping and tracking which legs, cycles and
stakes you are betting on.
219
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
The low 1.25 odds in this type of system mean that when the recovery bets are made, they do
not win enough to win back the money in the bets.
As the odds get higher, the recovery bets have more effect and we see higher profits for Type
2 than Type 1 for the first time in this book.
220
Rec15 (s0.5-RP100-CR10-
1,416 706 5 13,344 -93 1617 121.10% 0
BN2-RIS-LC) AD2
Rec16 (s0.5-RP100-CR10-
1,493 670 5 10,397 414 864 33.58% 0
BN4-RIS-LC) AD2
Rec19 (s1-RP100-CR1-BN1-
1,786 560 5 4,850 533 583 4.46% 0
RIS-LC) AD2
Rec17 (s0.5-RP100-CR10-
2,778 360 6 1,213 310 405 13.19% 0
BN10-RIS-NLC) AD2
With the slight increase in odds, and lower hit rates, for Type 3 systems, the profits begin to
fall.
Once again we see the higher odds and the recovery aspect of this plan forcing initial stakes
lower and the profits become inefficient compared to level stakes.
221
29.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave. Bets Ave. Low High Range Bnkrpt
Staking Ave. Let It Ride
to Dble Total Profit Profit from Chnce
Plan ROI Balance
Bank Profit Range Range Ave. % %*
Rec25 (s0.25-RP100-CR10-
3,891 257 5 594 -23 565 114.40% 0
BN5-RIS-LC) AD7.5
Rec23 (s02.5-RP100-CR10-
3,937 254 5 582 -98 766 170.08% 0
BN4-RIS-LC) AD7.5
Rec26 (s0.25-RP100-CR10-
4,808 208 5 423 72 322 60.10% 0
BN10-RIS-LC) AD7.5
Rec21 (s0.1-RP100-CR10-
5,025 199 5 397 -66 571 160.05% 0
BN3-RIS-LC) AD7.5
Rec24 (s0.25-RP100-CR10-
5,181 193 5 381 109 259 38.86% 1.2
BN10-RIS-NLC) AD7.5
Rec22 (s0.1-RP100-CR10-
6,369 157 3 297 -98 982 343.95% 0
BN2-RIS-LC) AD7.5
While the artificial data sets suggested some ability to get more profit than level stakes for
lower odds ranges, when actually applied to real betting systems they do generate more profit
on average, but there are some simulations that come out with lower than level stakes results.
Which means that it is the luck of the sports calendar whether the hard work of following
such a staking plan will result in something. There are other plans that do not have that
calendar gamble, and so yet again I feel that these recovery style plans are not worth the effort
of following. I hope the mounting data in this book also helps you come to the same
conclusion.
223
30 Recovery Type 2 Staking
Rating
Overview An unwise plan that bets large amounts after a win, effectively needing two
wins in a row to recover previous losses.
Concept: To take advantage of winning runs to start the recovery of losses
Other names: Winning Run Recovery
Variables: Link stakes to the current balance or not
This is a very risky and unadvisable staking plan that I add here for the sake of completion and
to highlight again the dangers of recovery staking plans and the folly of trying to wait for a
winning run when someone’s “magic” has come back. In this plan, when a stake is lost, a
record is kept of the losses so far and when a win comes, all of the losses become the stake for
the next bet. That means if you bet 1 point per bet and lost 5 bets in a row, when a win comes,
the next bet after that will have a 5 point stake on it. It is the equivalent of saying that there is
more chance of 2 wins coming in a row, so it is a better time to bet than after a loss. In
previous chapters we have seen 1326 Staking and 1234 Winners Staking having okay results
with low odds systems that have high hit rates, but with higher odds they perform badly. This
plan is even more aggressive than them in its increases, so I expect even worse results.
Positives Negatives
None. Needs two wins in a row to recover losses.
Needs a high hit rate.
A recovery plan, so stakes increase.
Focusses on stakes without thinking of the odds.
No limit, so stakes can go very high.
Goes bankrupt in almost all simulations.
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
Rec S3 (S2-r2-LC) AD1.25 966 1,035 6 130,515 -381856 1544 18521. % 6.6
Rec S1 (s1-r2-NLC) AD1.25 1,923 520 6 3,676 406 547 13.56% 0.5
Rec S2 (s1-r2-LC) AD1.25 1,385 722 6 14,909 -3306 794 283.93% 1.9
Rec S8 (S0.5-r2-LC) AD1.25 2,755 363 6 1,238 275 375 13.77% 0.1
Rec S7 (S0.01-r2-LC) AD1.25 NA -15 -14 90 -16 -15 -3.33% 0
Rec S4 (S3-r2-LC) AD1.25 NA -698,302 6 0 -431^6 2250 -30872 % 26
Staking Plan Name Key:
Rec S8 = Recovery Version 2 Staking Plan Profile No. 8 s = percentage of bank staked
r = which bet after a win to place recovery bet on (fixed at 2nd bet)
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
Similar to Martingale, almost all the results are bankrupt. This is an awful staking plan.
225
30.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Rec S1 (s1-r2-NLC) AD1.75 1,145 873 6 42,461 -3308 1316 264.83% 99.7
Rec S9 (s0.1-r2-LC) AD1.75 12,195 82 7 177 -32613 187 20000% 35.1
Rec S7 (S0.01-r2-LC) AD1.75 47,619 21 13 116 -196 70 633% 1.9
As the numbers get too large, my staking software returns errors, but it is sure that even
starting at such low stakes as 0.01%, the bank will go bankrupt 100% of the time.
226
30.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
Rec S7 (S0.01-r2-LC) AD15 NA NA NA NA NA NA NA 100
Rec S10 (S0.01-r2-NLC) AD15 NA NA NA NA NA NA NA 100
This is a terrible staking plan and most of the results are bankrupt. Only one profile with a
high hit rate shows profit and even then it is an inefficient plan.
227
30.3 Applying to Real Betting Systems
Ave. Rang Ave. Ave.
Staking Ave. Let It Low High Level
Bets Ave. from Brpt Bets Mnths
Plan Total Ride Prof. Prof. Stakes
to Dbl ROI Ave. % per to
Profile Profit Bal. Rnge Rnge Increase
Bank % Mnth Dble
Rec S8
(S0.5-r2-LC) 0 78 8 172 9 86 49.36% 0.1 0 NA 29%
Col3
Rec S7
(S0.01-r2- 0 3 9 102 -22 29 850% 0.1 0 NA -97%
LC) Pro1
Rec S10
(S0.01-r2- 0 7 13 105 -1178 44 8728 % 2.4 0 NA -96%
NLC) Col5
Rec S10
(S0.01-r2-
0 5 28 104 -11 39 500.% 0.8 0 NA -96%
NLC)
BTMO2.5
Rec S10
109916 67535%
(S0.01-r2- 0 404,462 62 NA -262 1358.% 99.2 0 NA
NLC) Col17 06
Rec S10
103906
(S0.01-r2- 0 53^6 113 NA -15204361 980.% 91 0 NA NA
NLC) Pro8 2853
Rec S10
-3.31526 3.7961
(S0.01-r2- 0 3^12 183 NA 4872.% 100 0 NA NA
NLC) HRB E+14 7E+13
Rec S7
120374 75747%
(S0.01-r2- 0 609,055 23 NA -146006 1000.% 73.8 0 NA
LC) Mix6 85
This is one of the worst staking plan so far as all the systems go bankrupt. This staking plan
should not be used at all.
228
31 Recovery Type 3 Staking
Rating
Overview Combines Recovery Staking and Target Profit Staking.
Concept: Bet in the normal way, but add a recovery aspect using target profit staking
to recover a specifiable number of losses.
Other names: Target Profit Recovery Leg Staking
Variables: Number of losses to recover
Percentage to recover
Number of legs to split the recovery cycle into
Recover losses while also betting regular stakes or just bet to recover losses
Link stake to current balance
Recovery Type 1 Staking’s major flaw was that it simply split the deficit into equal
stakes and bet them regardless of odds on the next bets. Recovery Type 3 fixes that flaw by
splitting the deficit into equal profit targets and then using the odds of each bet to
calculate how much stake is required. It is much more capital efficient since there is no chance
of betting large amounts on risky high odds. However, it does mean more money is bet on
lower odds to make sure the target for that leg is reached. You can also choose to bet to just
recover losses or to also keep the regular stake going at the same time as being in recovery
mode. You can also flat bet or use percentage staking for the regular staking part. It has a lot
of good options to play with, but it ends up being very complicated in that respect to find out
which settings are most efficient. In the end it proves itself to be the best of the recovery
staking plans, but its profits fall just short of LP28 Staking, Pure Percentage Staking and
Target Profit Staking. But it does suggest that recovery staking is not always bad.
229
win 0.5 points.
6 2.5 2 Loss -2 -5 97.5
Bet No.5 lost, so the series deficit is -3. We start a 2 leg recovery cycle for the 2nd time. Bet No.6 is the regular stake of 1
plus a Leg 1 recovery bet of 1 @ odds of 2.5 to win 1.5 points.
7 1.25 11 Loss -11 0 86.5
Bet No.6 lost, so the series deficit is -5. We start a 2 leg recovery cycle for the 3rd time. Bet No.7 is the regular stake of 1
plus a Leg 1 recovery bet of 10 @ odds of 1.25 to win 2.5 points. Not here that the low odds of the bet mean we need a
higher stake to reach our target. This is one drawback of target profit staking with low odds bets. The bet loses which mean
we need to start a new 2 leg recovery for a 4th time, but since the limit here is 4 cycles we cannot do it and the series is
abandoned. The series deficit is rest and the next bet starts from 1 again.
8 3 1 Loss -1 -1 85.5
9 10 1.06 Loss -1.06 -2.06 84.44
Bet No.8 lost, so the series deficit is -1. Bet No.9 is the regular stake of 1 plus a Leg 1 recovery bet of 0.06 @ odds of 10 to
win 0.5 points. Note that with high odds of 10, we only need a small stake to reach our target. This is a good point of target
profit staking. A bet with odds of 10 is unlikely to win and the stake reflects the implied risk of the odds.
10 2.5 1.69 Win 2.54 -1.03 86.98
Bet No.9 lost, so the series deficit is -2.06. We start a 2 leg recovery cycle for the 2nd time. Bet No.10 is the regular stake
of 1 plus a Leg 1 recovery bet of 0.69 @ odds of 2.5 to win 1.03 points.
11 3 1.51 Loss -1.51 -2.54 85.47
Bet No. 10 wins, so Leg 1 of the recovery bet succeeded. Bet No. 11 stakes the basic stake of 1 point plus a Leg 2 recovery
bet to win 1.03 points. The odds are 3 see we need to bet 0.51 point to achieve that target profit. The total stake is the
regular stake 1.00 plus the Leg 2 recovery stake 0.51 to make 1.51 points.
12 2.1 2.15 Win 2.37 -1.27 87.84
Bet No.11 lost, so the series deficit is -2.54. We start a 2 leg recovery cycle for the 3rd time. Bet No.12 is the regular stake
of 1.00 plus a Leg 1 recovery bet of 1.15 @ odds of 2.1 to win 1.27 points.
13 3 1.63 Loss -1.63 0 86.21
Bet No. 12 wins, so Leg 1 (Cycle 3) of the recovery bet succeeded. Bet No. 13 stakes the basic stake of 1.00 point plus a
Leg 2 recovery bet to win 1.27 points. The odds are 3 see we need to bet 0.63 points to achieve that target profit. The total
stake is the regular stake 1.00 plus the Leg 2 recovery stake 0.63 to make 1.63 points. The bet loses which mean we need to
start a new 2 leg recovery for a 4th time, but since the limit here is 4 cycles we cannot do it and the series is abandoned. The
series deficit is rest and the next bet starts from 1 again.
14 3 1 Win 2 0 88.21
Positives Negatives
Adjusts stakes to bet risk signaled by odds Recovery plans increase stakes, reducing
initial stakes and overall profit.
Smaller stakes are used for higher odds Stakes can be pushed higher by low odds
reflecting the implied risk. bets when using target profit staking.
Recovery bets are spread over a number of
legs.
A cycle limit can be set to keep stakes under
control.
230
31.1 Analysis of Recovery Type 3 Staking
The profile RecT3-5 (s15-RP100-CR1-BN4-RIS-LC) comes out as best, but since the cycle is
set at 1, that actually means no recovery being done at all, since after 1 loss happens, it reaches
the limit. It ends up being just percentage staking. So the real winner is the bright green
highlighted RecT3-9 (s10-RP100-CR2-BN4-RIS-LC) which only tries to recover 1 loss over
the next four bets.
231
RecT3-17 (s5-RP100-CR1-
427 2,342 5 1,122,334,979 2300 2516 4.61% 0
BN2-RIS-LC) AD1.75
RecT3-20 (s5-RP100-CR2-
432 2,313 5 917,960,090 1001 3108 45.55% 0
BN2-RIS-LC) AD1.75
RecT3-50 (s5-RP100-CR3-
606 1,650 5 9,268,190 -99 3301 103.03% 0
BN2-RIS-LC) AD1.75
RecT3-5 (s15-RP100-CR1-
2,079 481 4 2,805 -99 2511 271.31% 0
BN4-RIS-LC) AD1.75
RecT3-14 (s5-RP100-CR1-
NA -823,654 6 0 NA 8890 -13175% 84.6
BN5-RIS-LC) AD1.75
It is worth noticing here that by adding recovery with 2 cycles (RecT3-20) to a non-recovery
plan (RecT3-17) that it produces slightly less profit (-19 points) but with a far greater range of
results from a low of 1,001 to a high of 3,108. That is a variation of 45% compared to the
non-recovery versions 4%. While Recovery Version 3 does better than the other Recovery
plans, it still has drawbacks that keep me from recommending recovery staking plans that seek
to recover losses directly from the last bets. It is much better to focus on the value of your
bets and to extract the most value from those bets with odds sensitive staking plans.
RecT3-22 is the non-recovery version of RecT3-26 which gets the best results in this range.
The average profit of the recovery version is almost half that of the non-recovery version.
This illustrates well the dangers of using recovery staking.
232
31.1.5 Type 5 Systems: Medium Odds Range 2.50-4.50
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
RecT3-26( s2-RP100-CR2-
1,285 778 5 21,979 101 1113 65.04% 0
BN3-RIS-LC) AD3.5
RecT3-29( s1-RP100-CR4-
1,287 777 5 21,827 101 1157 67.95% 0
BN1-RIS-LC) AD3.5
RecT3-22( s2-RP100-CR1-
1,289 776 5 21,677 501 1017 33.25% 0
BN3-RIS-LC) AD3.5
RecT3-28( s2-RP100-CR5-
1,379 725 5 15,222 -99 1305 96.83% 0
BN3-RIS-LC) AD3.5
RecT3-27( s1-RP100-CR1-
1,825 548 5 4,463 531 573 3.83% 0
BN3-RIS-LC) AD3.5
RecT3-13 (s5-RP100-CR1-
3,831 261 3 611 -99 1629 331.03% 0
BN4-RIS-LC) AD3.5
Recovery Plan RecT3-26 makes an average profit above level stakes but the range of results in
the simulations suggests it is not an efficient method and results can vary according to the luck
of the sports calendar.
233
RecT3-32 ( s0.5-RP100-
6,536 153 4 289 66 409 112.09% 0
CR7-BN5-RIS-LC) AD15
RecT3-34 ( s0.25-RP100-
8,065 124 5 236 101 230 52.02% 0
CR15-BN15-RIS-LC) AD15
RecT3-27( s1-RP100-CR1-
10,526 95 2 193 -99 605 370.53% 0
BN3-RIS-LC) AD15
RecT3-29( s1-RP100-CR4-
10,989 91 2 188 -99 913 556.04% 0
BN1-RIS-LC) AD15
Recovery Type 3 does very well for Type 1 and 2 betting systems. The high hit rates work well
and Profile 9 comes close to Percentage Staking but not quite. It suggests that Target Profit
elements added into recovery betting may work.
234
31.3 Applying to Real Betting Systems
Ave. Rang Ave. Ave.
Ave. "Let It Low High Level
Staking Bets Ave. from Brpt Bets Mnths
Total Ride" Prof. Prof. Stakes
Plan Profile to Dbl ROI Ave. % per to
Profit Bal. Rnge Rnge Increase
Bank % Mnth Dble
RecT3-9 (s10-
RP100-CR2- 4,034,2 1290%
94 1,530 8 1024 2024 32% 2.2 19 5
BN4-RIS-LC) 14
Col3
RecT3-20 (s5-
RP100-CR2-
BN2-RIS-LC)
265 748 5 17,853 375 1119 49% 0 24 11 419%
Pro1
RecT3-46 (s5-
RP100-CR2-
162 1,743 5 17^6 -99 2501 74% 0 34 5 812%
BN3-RIS-LC)
Col5
RecT3-26( s2-
RP100-CR2-
209 338 21 1,041 166 488 47% 0 9 23 138%
BN3-RIS-LC)
BTMO2.5
RecT3-26( s2-
RP100-CR2- 1,199,3 126%
515 1,355 9 1241 1529 10% 0 84 6
BN3-RIS-LC) 79
Col17
RecT3-29( s1-
RP100-CR4-
319 443 18 2,156 287 587 33% 0 18 18 225%
BN1-RIS-LC)
Pro8
RecT3-33 (
s0.5-RP100-
4,821 96 4 195 36 312 143% 0 165 29 33%
CR15-BN15-
RIS-LC) HRB
RecT3-25 ( s2-
RP100-CR2-
387 3,960 9 83^12 3685 4244 7% 0 199 2 393%
BN2-RIS-LC)
Mix6
Col3 goes bankrupt which highlights the dangers of using target profit staking with low odds
systems, the low odds can force high stakes which lead to a higher chance pf bankruptcy.
235
32 Stop at a Winner Staking
Rating
Overview Martingale with stop losses combined with Target Profit Staking
Concept: To win a target profit and then stop for the day. Also, increase stakes to
recover all losses until a winner is reached.
Other names: SAW, Stop Loss Martingale
Variables: Stop Loss limit
Increase stakes to recover target profit on each bet or not
Link stakes to a percentage of the current balance or not
Stop at a Winner staking (SAW) takes Martingale and tries to get it under control and more
risk efficient. It tries to get it under control by setting a target profit to be achieved each day.
The idea being that you only need to find one winner a day to reach the target. It doesn’t
sound hard, does it? Just one win. But all that does is slow down the eventual failure of the
system. It still follows the Martingale concept of increasing stakes to win back all losses until a
winner is found. SAW also tries to improve on Martingale by using Target Profit Staking and
adjusting the stake to the odds to win back the losses. Martingale just kept doubling its stakes
regardless of the odds. SAW also uses a stop loss to prevent stakes form getting too high and
chooses a point to just accept a loss and move on. Compared to normal Martingale, this is a
great improvement, but it is still a recovery system where the stakes increase rapidly. For this
reason the initial staking levels will be pushed down and the overall profits reduced compared
to other non-recovery staking plans.
236
odds of 3.38.
7 1.95 2.11 Loss -2.11 3.11 97.88
Bet No. 6 loses so the series loss is 1, To recover this and Bet No. 5’s target profit of 1, the stake is 2.11 @ odds
of 1.95.
8 1.74 5.55 Loss -5.55 8.66 92.33
Bet No. 7 loses so the series loss is 3.11, To recover this and Bet No. 5’s target profit of 1, the stake is 5.55 @
odds of 1.74. Note how odds under 2.00 need a higher stake to reach the target. The bet loses and the series
loss is now 8.66. But this is the 4th loss in a row, so the stoploss of 4 kicks in and we accept the loss. The next bet
will begin as a new series.
9 2.08 0.93 Loss -0.93 0.93 91.4
10 1.94 2.05 Win 1.93 0 93.33
11 2.37 0.73 Win 1 0 94.33
Positives Negatives
Stakes adjusted to odds implied risk level Stakes increase rapidly
Uses a stop loss Recovery staking means lower initial stakes
Goes bankrupt easily without a stop loss
Daily target slows down betting velocity and
ultimate target of doubling a bank
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
SAW-3 (T1-SL2-NLC-RTP) AD1.25 294 3,401 6 1^12 2140 4440 33% 96.2
SAW-2 (T1-SL2-LC-NRTP) AD1.25 296 3,376 6 1^12 1897 4887 44% 0
SAW-4 (T1-SL2-LC-RTP) AD1.25 305 3,281 5 752^9 1581 4962 51% 0
SAW-1 (T1-SL2-NLC-NRTP) AD1.25 322 3,106 5 223^9 2376 3895 24% 79.9
SAW-8 (T2-SL2-LC-NRTP) AD1.25 371 2,697 5 13^9 -99 6124 115% 0
SAW-9 (T2-SL2-LC-RTP) AD1.25 903 1,108 4 216,477 -100 4976 229% 5.1
SAW-5 (T0.1-SL100-NLC-NRTP) AD1.25 1,190 840 7 33,779 528 828 17% 99.5
Staking Plan Name Key:
SAW-8 = Stop At A Winner Staking Plan Profile No. 8 S = percentage of bank as target
SL = Stop Loss RTP = Recover the target profit each time as well NRTP = Only recover losses
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
237
32.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
SAW-9 (T2-SL2-LC-RTP) AD1.75 561 1,781 5 22,979,693 73 2791 76.31% 0
SAW-7 (T1-SL3-LC-RTP) AD1.75 758 1,320 5 941,014 118 2372 85.38% 0
SAW-4 (T1-SL2-LC-RTP) AD1.75 794 1,260 5 620,838 671 1815 45.40% 0
SAW-511(T0.1-SL100-LC-RTP) AD1.75 828 1,208 10 432,955 1080 1218 5.71% 100
SAW-10 (T1-SL4-LC-RTP) AD1.75 1,577 634 4 8,101 -99 2751 224% 0
238
32.1.6 Type 6 Systems: High Odds Range 5.00-10.00
Ave.
Ave. Let It Low High Range Bnkrpt
Staking Bets to Ave.
Total Ride Profit Profit from Chnce
Plan Dble ROI
Profit Balance Range Range Ave. % %*
Bank
SAW-13 (T2-SL3-LC-RTP) AD7.5 3,922 255 5 586 33 474 86.47% 0
SAW-9 (T2-SL2-LC-RTP) AD7.5 4,348 230 5 492 0 392 85.22% 0
SAW-7 (T1-SL3-LC-RTP) AD7.5 6,289 159 5 301 51 196 45.60% 0
SAW-4 (T1-SL2-LC-RTP) AD7.5 7,246 138 5 260 26 226 72.46% 0
239
SAW-9 (T2-SL2-LC-RTP)
1,669 599 5 172 1005 69.53% 500 19%
AD3.5
SAW-13 (T2-SL3-LC-RTP)
3,922 255 5 33 474 86.47% 250 2%
AD7.5
SAW-4 (T1-SL2-LC-RTP) 125
16,129 62 5 6 123 94.35% -50%
AD15
SAW-4 (T1-SL2-LC-RTP)
837 7,170 6 5502 8334 19.75% 2,531 183%
ADmx
Up-Down staking sounds very logical by increasing by certain levels after a loser and reducing
by certain levels after a winner, but it is still a form of recovery betting that is directly
connected to the previous bets wins or losses. Other staking plans either reset after a win or
try to recover a certain amount on each recovery bet, but Up-down Staking is purely stake
focused and directly connecting to the previous results. This is a recipe for disaster and the
bankruptcies in the results will attest to that.
241
Positives Negatives
Relatively simple to use. Stake focused with no adjusting to odds and risk.
When on a winning run, cannot decrease below 1.
Stakes can rise over time and bankrupt the betting bank.
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
UD-3 (s2-LU4-WD1-NLC)
112 8,962 6 95^27 8055 9005 5.30% 100
AD1.25
UD-4 (s2-LU4-WD1-LC)
198 5,048 5 157^15 2554 7160 45.62% 99.8
AD1.25
UD-1 (s1-LU4-WD1-NLC)
223 4,479 6 3^15 4115 4507 4.38% 18.8
AD1.25
UD-2 (s1-LU4-WD1-LC)
247 4,048 5 153^12 2156 5145 36.92% 59.7
AD1.25
UD-14 (s5-LU1-WD1-LC)
266 3,764 5 21^12 3171 4221 13.95% 0.2
AD1.25
UD-6 (s2-LU3-WD1-LC)
272 3,671 5 11^12 2554 4759 30.03% 32.4
AD1.25
UD-5 (s2-LU3-WD1-NLC)
273 3,667 6 10^12 2829 4077 17.02% 99.6
AD1.25
UD-16 (s4-LU1-WD1-LC)
319 3,134 6 271^9 2639 3559 14.68% 0
AD1.25
UD-9 (s2-LU1-WD1-LC)
599 1,669 5 10,572,815 1405 1845 13.18% 0
AD1.25
UD-10 (s0.25-LU4-WD1-
815 1,227 6 493,899 1079 1271 7.82% 11.8
NLC) AD1.25
UD-8 (s1-LU1-WD1-LC)
1,167 857 5 38,004 745 941 11.44% 0
AD1.25
UD-7 (s1-LU1-WD1-NLC)
1,618 618 6 7,250 505 685 14.56% 0
AD1.25
Staking Plan Name Key:
UD-8 = Up Down Staking Plan Profile No. 8 s = percentage of bank staked
LU1 = stakes up 1 point after a loss WD1 = stakes down 1 point after a win
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.25 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
As with most staking plans, low odds and their high win rate result in some settings being
profitable.
242
33.1.2 Type 2 Systems: Odds-On Range 1.50-1.99
Ave. Bets Ave. Let It Low High Range Bnkrpt
Staking Ave.
to Dble Total Ride Profit Profit from Chnce
Plan ROI
Bank Profit Balance Range Range Ave. % %*
UD-9 (s2-LU1-WD1-LC) AD1.75 493 2,028 5 127^6 844 3194 57.94% 87.5
UD-7 (s1-LU1-WD1-NLC) AD1.75 660 1,516 6 3^6 928 1828 29.68% 90.1
UD-8 (s1-LU1-WD1-LC) AD1.75 664 1,506 5 3,^6 837 2087 41.50% 17.5
UD-10 (s0.5-LU1-WD1-LC) AD1.75 1,031 970 6 83,175 629 1240 31.49% 0.8
UD-11 (s0.25-LU1-WD1-LC) AD1.75 1,550 645 6 8,743 439 739 23.26% 0
As the odds increase, the plans turn mostly to bankruptcy and one inefficient profile.
The artificial data results show a terrible staking plan with a lot of bankruptcy, which makes
the results for the real betting systems all the more surprising.
244
33.3 Applying to Real Betting Systems
Ave. Rang Ave.
Staking Ave. "Let It Low High Brpt Ave. Level
Bets Ave. from Bets
Plan Total Ride" Prof. Prof. % Mnths Stakes
to Dbl ROI Ave. per
Profile Profit Bal. Rnge Rnge to Dble Increase
Bank % Mnth
UD-16 (s4-
LU1-WD1- 197 728 8 15,542 515 899 26.37% 0 19 10 561%
LC) Col3
UD-11
(s0.25-LU1-
WD1-LC) 1,120 177 8 341 90 190 28.25% 0 24 47 22%
Pro1
UD-17
(s0.05-LU1-
WD1-LC) 1,431 197 7 392 184 214 7.61% 0 34 42 3%
Col5
UD-13
(s0.01-LU1-
WD1-LC) 493 143 21 269 143 143 0.00% 0 9 55 1%
BTMO2.5
UD-21
(s0.05-LU1-
WD3-LC) 1,167 598 9 6,312 598 598 0.00% 0.2 84 14 0
Col17
UD-23
(s0.02-LU1-
WD16-LC) 517 273 20 663 273 273 0.00% 0 18 29 100%
Pro8
UD-23
(s0.02-LU1-
WD16-LC) 1,613 287 7 731 286 290 0.70% 87.7 165 10 298%
HRB
UD-15
(s0.01-LU1- 2,432,24
1,052 1,457 10 1457 1457 0.00% 0 199 5 81%
WD2-LC) 3
Mix6
When applied to real betting systems, the results seem very positive. They do not improve on
the recommended staking plans but it is these kinds of results that can make people believe
that a staking plan works when actually it probably does not. There are much better staking
plans than simple Up-Down, so this staking plan can be avoided.
245
34 i-TSM (The Staking Machine) Staking
Rating
Overview Takes advantage of consistent betting systems by betting more when a system
is doing worse than it has done historically and betting less when it is doing
better than it has done historically. It aims to bet more when we are due a win
and bet less when we are expecting losses.
Concept: To increase stakes when win rate is below average, and/or decrease stakes
when win rate is above average.
Other names: TSM staking
Variables: How much to increase/decrease stakes when under/over the winning
average.
Link stake to current betting bank or not.
Need to know historical win rate.
Can set the percentage increase based on variation from a moving current
win rate based on last 10. 20, 30 or 50 bets.
i-TSM Staking is a staking plan from the Staking Machine that takes a lot of good aspects of
other methods and makes a new and very interesting staking plan. It has percentage staking to
protect the bank. It has stakes based on historical average win rates similar to Kelly staking, so
that you can bet more on higher value bets. i-TSM staking also introduces the ability to reduce
stakes when you are above the winning average. In Up-Down staking it could never go below
the original stake levels, but i-TSM allows us to reduce the stakes below the original stake level
in certain conditions. If you know that it is unlikely that a winning run will continue forever,
then it makes sense to reduce stakes the longer the winning run goes on. One problem that i-
TSM and Kelly both have is the need to know the historical average of your betting system’s
win rate. If that data is wrong, then already we have a weak starting point. But I really do like
the idea behind i-TSM. The i-TSM staking plan works well with consistent, stable betting
systems. With value betting systems such as the systems created from big data like Betaminic’s
systems, there is no expectation that all bets will win. There will be winning and losing months,
but over the long term, there is an expectation of a certain yield and win rate. In a way, it is
almost like the stock market. The win rate will sometimes be above its average and sometimes
below, if I could bet more when it is below the average and bet less when it is above its
average, it would be like buying more when the market is low and buying less when the market
is high. Value betting with big data betting systems is almost like the index funds of the
betting world. The i-TSM staking method seems like the right kind of methodology to take
advantage of that. The only two drawbacks of the method are that it is stakes focused and
does not take into account the odds which are the implied risk of each bet, and also that it is a
little complicated to use and requires record keeping to know your winning rates both
currently and historically. My own new staking plan “Whitaker Staking”, which is explained in
the next chapter, seeks to solve those two problems.
246
Example of i-TSM Staking
In this example we increase stakes by 5% each time the win rate drops below the average in
specified rolling time periods. Those periods are the overall win rate and the win rate for the
last 10, 20, 30 and 50 bets. We also decrease stakes by 5% each time the win rate is above the
average in those time periods.. The increases and decreases are cumulative
Data Set Estimated Strike Level (ESL) based on historical data: 45.8%
“ASR/-5/-5/-5/-5/-5” ASR = Above Strike Rate Last 1-9 bets = -5% change
Last 10 bets = -5% reduction of stake Last 20 bets = -5% reduction of stake
Last 30 bets = -5% reduction of stake Last 50 bets = -5% reduction of stake
“USR/5/5/5/5/5” USR = Under Strike Rate Last 1-9 bets = 5% change
Last 10 bets = 5% increase of stake Last 20 bets = 5% increase of stake
Last 30 bets = 5% increase of stake Last 50 bets = 5% increase of stake
Win Rate Win Rate Win Rate Win Rate Win Rate
Bet Stake Odds Result Profit Balance
Overall Last 10 Last 20 Last 30 Last 50
1 1 1.99 Win 0.99 100.99 100%
2 0.95 1.99 Win 0.94 101.93 100%
Bet No. 1 won, so the overall hit rate is 100%. This is above the ESL of 45.8%. So the stake is decreased by 5% to 0.95.
3 0.95 2.39 Loss -0.95 100.98 67%
4 0.95 3.57 Loss -0.95 100.03 50%
5 0.95 3.43 Loss -0.95 99.08 40%
6 1.05 3.38 Loss -1.05 98.03 33%
Bet No. 5 lost, so the overall hit rate is 40%. This is below the ESL of 45.8%. So the stake is increased by 5% to 1.05.
7 1.05 1.95 Loss -1.05 96.98 29%
8 1.05 1.74 Loss -1.05 95.93 25%
9 1.05 2.08 Win 1.13 97.06 33%
10 1.05 1.94 Win 0.99 98.05 40% 40%
11 1.1 2.37 Win 1.51 99.56 45% 40%
After Bet No. 10 the overall hit rate is 40% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.05.
Also, we have 10 bets of data in the set, so we can calculate the win rate from the last 10 which is also 40%. This is under
the ESL of 45.8% so the stake is increased by another 5% to 1.1.
12 1.1 2.02 Loss -1.1 98.46 42% 30%
After Bet No. 11 the overall hit rate is 45% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.05.
After Bet No. 11 the Last 10 hit rate is 40% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.1.
13 1.1 2.5 Loss -1.1 97.36 38% 30%
14 1.1 3.42 Loss -1.1 96.26 36% 30%
15 1.1 3.3 Loss -1.1 95.16 33% 30%
16 1.1 1.95 Loss -1.1 94.06 31% 30%
17 1.1 3.28 Loss -1.1 92.96 29% 30%
18 1.1 1.85 Loss -1.1 91.86 28% 30%
19 1.1 1.88 Win 0.97 92.83 32% 30%
20 1.1 3.43 Loss -1.1 91.73 30% 20% 30%
21 1.15 2.02 Win 1.17 92.9 33% 20% 30%
After Bet No. 20 the overall hit rate is 33% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.05.
After Bet No. 20 the Last 10 hit rate is 20% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.1.
Also, we have 20 bets of data in the set, so we can calculate the win rate from the last 20 which is also 30%. This is under
the ESL of 45.8% so the stake is increased by another 5% to 1.15.
22 1.15 3.55 Loss -1.15 91.75 32% 20% 25%
23 1.15 1.7 Loss -1.15 90.6 30% 20% 25%
24 1.15 2.31 Win 1.51 92.11 33% 30% 30%
25 1.15 1.91 Loss -1.15 90.96 32% 30% 30%
26 1.15 3.52 Loss -1.15 89.81 31% 30% 30%
27 1.15 3.54 Win 2.92 92.73 33% 40% 35%
28 1.15 1.2 Win 0.22 92.94 36% 50% 40%
247
29 1.05 3.22 Win 2.33 95.27 38% 50% 40%
After Bet No. 28 the overall hit rate is 36% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.05.
After Bet No. 28 the Last 10 hit rate is 50% and now over the ESL of 45.8%. So the stake is decreased by 5% to 1.0.
After Bet No. 28 the Last 20 hit rate is 40% and still under the ESL of 45.8%. So the stake is increased by 5% to 1.05.
248
Positives Negatives
Bets more when wins are due. More complex than some staking plans
Bets less when losses are due. Stakes do not take the odds and risk into
account.
Takes advantage of a betting system’s short Needs knowledge of your historical winning
term performance with knowledge of its average and also the current series winning
long-term performance. average over a certain number of bets.
Can bet less than the base staking level if
needed.
Uses percentage staking.
Can set different increments for various
degrees above/below the historical win rate
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
TSM-24 (s15-ESL84- 17,358,638,753,
ASR/0/0/0/0/0- 130 7,720 5 809,600,000,000 1401 8301 44.69% 0
USR/5/5/5/5/5-LC) AD1.25 ,000
TSM-20 (s12-ESL84- 5,305,842,726,5
ASR/0/0/0/0/0-
USR/5/10/15/20/25-LC)
132 7,549 5 78,500,000,000, -99 8501 56.96% 0
AD1.25 000
TSM-23 (s14-ESL84- 5,647,380,790,9
ASR/0/0/0/0/0- 132 7,558 5 57,170,000,000, 3601 7913 28.53% 0
USR/5/5/5/5/5-LC) AD1.25 000
TSM-21 (s13-ESL84- 3,777,893,186,2
ASR/0/0/0/0/0-
USR/5/10/15/20/25-LC)
133 7,500 5 95,720,000,000, 1 8601 57.33% 0
AD1.25 000
TSM-15 (s11-ESL84- 3,047,406,405,9
ASR/0/0/0/0/0-
USR/5/10/20/30/40-LC)
134 7,469 5 46,850,000,000, 701 8573 52.70% 0
AD1.25 000
TSM-14 (s11-ESL84- 2,052,784,116,2
ASR/0/0/0/0/0-
USR/0/10/20/30/40-LC)
135 7,412 5 23,610,000,000, 601 8637 54.21% 0
AD1.25 000
TSM-19 (s11-ESL84- 2,024,522,830,6
ASR/0/0/0/0/0-
USR/5/10/15/20/25-LC)
135 7,410 5 84,200,000,000, 101 8101 53.98% 0
AD1.25 000
TSM-16 (s10-ESL84- 1,317,298,148,4
ASR/0/0/0/0/0-
USR/5/10/20/30/40-LC)
136 7,348 5 33,580,000,000, 101 8301 55.80% 0
AD1.25 000
TSM-17 (s10-ESL85- 1,354,332,440,2
ASR/0/0/0/0/0- 136 7,352 5 32,990,000,000, 1 8201 55.77% 0
USR/5/10/20/30/40-LC) AD 000
249
TSM-22 (s13-ESL84- 1,195,473,302,9
ASR/0/0/0/0/0- 136 7,334 5 67,300,000,000, 7070 7707 4.34% 0
USR/5/5/5/5/5-LC) AD1.25 000
TSM-11 (s10-ESL84- 977,780,075,447
ASR/0/0/0/0/0-
USR/0/10/20/30/40-LC)
137 7,305 5 ,823,000,000,00 3201 8413 35.67% 0
AD1.25 0
TSM-13 (s12-ESL84- 991,429,376,650
ASR/0/0/0/0/0-
USR/0/10/20/30/40-LC)
137 7,307 5 ,126,000,000,00 1 8601 58.85% 0
AD1.25 0
TSM-18 (s11-ESL80- 443,677,074,078
ASR/0/0/0/0/0-
USR/5/10/20/30/40-LC)
139 7,191 5 ,177,000,000,00 401 8192 54.17% 0
AD1.25 0
TSM-12 (s15-ESL84-
ASR/0/0/0/0/0- 1,692,493,721,3
185 5,391 5 -99 8231 77.26% 0
USR/0/10/20/30/40-LC) 06,530,000
AD1.25
TSM-10 (s10-ESL84-
ASR/0/0/0/0/0- 31,887,846,522,
208 4,818 5 -99 8541 89.66% 0
USR/0/50/50/50/50-LC) 951,100
AD1.25
TSM-8 (s1-ESL84-
ASR/0/0/0/0/0-
USR/0/100/100/100/100-LC)
402 2,486 6 3,045,128,979 2155 2792 12.81% 0
AD1.25
TSM-5 (s1-ESL84-
ASR/0/0/0/0/0-
USR/0/50/50/50/50-LC)
613 1,632 6 8,181,059 1455 1798 10.51% 0
AD1.25
TSM-7 (s1-ESL84-ASR/0/-50/-
50/-50/-50-
USR/0/50/50/50/50-LC-
717 1,394 5 1,571,658 1147 1647 17.93% 0
MinBet1) AD1.25
TSM-2 (s1-ESL84-
ASR/0/0/0/0/0-
USR/0/10/20/30/40-LC)
891 1,122 6 238,537 1044 1200 6.95% 0
AD1.25
TSM-6 (s1-ESL84-ASR/0/-10/-
20/-30/-40-
USR/0/10/20/30/40-LC)
1,181 847 6 35,459 698 980 16.65% 0
AD1.25
TSM-1 (s1-ESL84-
ASR/0/0/0/0/0-
USR/0/10/20/30/40-NLC)
1,259 794 5 24,557 738 838 6.30% 0
AD1.25
TSM-3 (s1-ESL84-
ASR/0/0/0/0/0- 1,773 564 5 4,987 542 582 3.55% 0
USR/0/0/25/0/0-NLC) AD1.25
TSM-9 (s10-ESL84-
ASR/0/0/0/0/0-
USR/0/100/100/100/100-LC)
1,783 561 4 4,884 -99 3301 303.03% 0
AD1.25
TSM-4 (s1-ESL84-ASR/0/0/-
25/0/0-USR/0/0/25/0/0-NLC) 1,912 523 5 3,753 486 558 6.88% 0
AD1.25
Staking Plan Name Key:
“TSM-6 (s1-ESL84-ASR/0/-10/-20/-30/-40-USR/0/10/20/30/40-LC) AD1.25”
TSM-6 = i-TSM Staking Plan Profile No. 8 s = percentage of bank staked
ESL84 = Estimated Strike Level 84%
“ASR/0/-10/-20/-30/-40” ASR = Above Strike Rate Last 1-9 bets = 0 change
Last 10 bets = -10% reduction of stake Last 20 bets = -20% reduction of stake
Last 30 bets = -30% reduction of stake Last 50 bets = -40% reduction of stake
250
“USR/0/10/20/30/40” USR = Under Strike Rate Last 1-9 bets = 0 change
Last 10 bets = 10% increase of stake Last 20 bets = 20% increase of stake
Last 30 bets = 30% increase of stake Last 50 bets = 40% increase of stake
LC = Link stake to cumulative total (percentage staking) NLC = Not linking stake to cumulative total
AD1.2”5 = Artificial Data Set of 10,000 bets where all bets are at odds of 1.25 1^12 = +12 zeros (1,000,000,000,000)
Best Plan Profitable Plans Inefficient Plans Bankrupt Plans
There are so many variables that it needed a lot of research to try and find the best settings for
this bracket. Thus the large amount of results displayed. Finally it becomes clear that both
increasing stakes when below the average and decreasing stakes when below the average
sometimes cancels each other out when different time periods are both positive and negative.
Also, reducing stakes, similar to the results seen in Phantom Bet Staking and in 4321 Staking
leads to lower profits overall since we bet less in some cases and miss out on winnings. The
best settings are when we increase stake size when the current win rate is below its average in
various periods. In fact, the average total profit (7,720) is better than the LP28 Staking total
(7,516) and i-TSM takes its place in the recommended staking plan leader board. It also does
this with a staking level of 15% of the bank. It is, however, worth noting that the TSM plan
has a wider range of results (44.69%) compared to the LP28 plan (4.12%)
251
For Type 2 systems i-TSM (2,483) does not beat LP28 (2,498) but it comes close. TSM staking
5% of the bank only has a variance of 10.07%, which is slightly higher than LP28’s 9.25%. But
LP28 is betting 6% of bank to get those results. When TSM bets 6% of bank, it bets LP28 in
total profit but some simulations come out as nearly bankrupt and a much higher variance of
63%. So LP28 Staking retains its place in the leaderboard.
Again TSM (1,891) comes close to beating LP28 (1,919) but not quite.
252
Again TSM (1,487) is close to LP28 (1,535) but does not beat it. As the odds increase, the
stake is brought down to 3% of the bank.
For Type 5 systems suddenly TSM (611) does not get even close to LP28 (1,031). LP28 makes
almost twice as much profit staking 2% of the bank. TSM’s 1% staking is not close and 2%
staking is inefficient.
As the odds get higher, it is harder to even stay efficient. Staking 0.5% of the bank is the best
setting that can be found.
253
34.1.7 Type 7 Systems: Very High Odds Range 10.00-30.00
Ave.
Ave. Low High Range Bnkrpt
Staking Bets to Ave. Let It Ride
Total Profit Profit from Chnce
Plan Dble ROI Balance
Profit Range Range Ave. % %*
Bank
TSM-51 (s0.5-ESL7-
ASR/0/0/0/0/0- 6,369 157 4 297 -1 316 100.96% 0
USR/5/5/5/5/5-LC) AD15
TSM-57 (s0.25-ESL7-
ASR/0/0/0/0/0- 7,463 134 5 253 97 232 50.37% 0
USR/5/5/5/5/5-LC) AD15
TSM-58 (s0.25-ESL7-ASR/-5/-
5/-5/-5/-5-USR/5/5/5/5/5- 7,634 131 5 248 56 236 68.70% 0
LC) AD15
TSM-50 (s1-ESL7-ASR/-5/-5/-
5/-5/-5-USR/5/5/5/5/5-LC) 10,638 94 2 192 -99 711 430.85% 0
AD15
TSM-49 (s1-ESL7-
ASR/0/0/0/0/0- 12,048 83 2 178 -99 701 481.93% 0
USR/5/5/5/5/5-LC) AD15
Type 7 also has the same problems as Type 6. The higher odds do worse with TSM staking.
This is possibly due to the lack of odds sensitive staking. If some kind of Target Profit aspect
were introduced, it may solve this problem. The best results are achieved staking 0.25% of the
bank.
Interestingly, in the mixed odds version TSM (3,876) beats LP28 (3,516) easily but does not
come close to the fantastic performance of Secure Staking (13,766). However, Secure Staking
(50.12%) has a much wider range of results than TSM (4.51%) or LP28 (2.73%). Secure
Staking’s lowest result is still a healthy 2,709 points which is above level stakes 2,531 points
profit.
254
Ave. Bets Ave. Low High Range Levels Level
Staking Ave.
to Dble Total Profit Profit from Stakes Stakes
Plan ROI
Bank Profit Range Range Ave. % Profit Increase
Sec1 ADmx 436 13,766 5 2709 14301 50.12% 2,531 443%
TSM-39 (s1-ESL47.9-
ASR/0/0/0/0/0- 1,548 3,876 5 3703 4053 4.51% 2,531 53%
USR/5/5/5/5/5-LC) ADmx
LP28-S15 (s1-LC) ADmx 1,706 3,516 5 3440 3632 2.73% 2,531 38%
Overall i-TSM staking shows excellent results for odds ranges below 3.00, but becomes
inefficient as it reaches higher levels of 7.50 and above. It does well with mixed systems, too.
255
34.3 Applying to Real Betting Systems
Ave. Rang Ave.
Staking Ave. "Let It Low High Brpt Ave. Level
Bets Ave. from Bets
Plan Total Ride" Prof. Prof. % Mnths Stakes
to Dbl ROI Ave. per
Profile Profit Bal. Rnge Rnge to Dble Increase
Bank % Mnth
TSM-24 (s15-
ESL82.5-
ASR/0/0/0/ 85,763,2
78 1,971 8 1931 2218 7.28% 0 19 4 1750%
0/0- 47
USR/5/5/5/
5/5-LC) Col3
TSM-59 (s5-
ESL61.5-
ASR/0/0/0/
0/0- 247 802 8 25,957 720 920 12.47% 0 24 10 456%
USR/5/5/5/
5/5-LC) Pro1
TSM-53 (s4-
ESL53.9-
ASR/0/0/0/
0/0- 321 879 7 44,264 785 1020 13.37% 0 34 9 360%
USR/5/5/5/
5/5-LC) Col5
TSM-54 (s3-
ESL53-
ASR/0/0/0/
0/0- 109 646 21 8,803 604 694 6.97% 0 9 12 354%
USR/5/5/5/
5/5-LC)
BTMO2.5
TSM-42 (s1-
ESL33.6-
ASR/0/0/0/
0/0- 828 843 9 34,489 791 902 6.58% 0 84 10 40%
USR/5/5/5/
5/5-LC)
Col17
TSM-64 (s0.5-
ESL27.6-
ASR/0/0/0/
0/0- 679 208 20 423 193 223 7.21% 0 18 38 52%
USR/5/5/5/
5/5-LC) Pro8
TSM-61
(s0.25-
ESL18.8-
ASR/0/0/0/ 5,445 85 6 180 51 149 57.65% 0 165 33 18%
0/0-
USR/5/5/5/
5/5-LC) HRB
TSM-63 (s1-
ESL45.8-
ASR/0/0/0/ 300,722,
712 2,152 10 2061 2249 4.37% 0 199 4 167%
0/0- 212
USR/5/5/5/
5/5-LC) Mix6
The TSM staking plan works very well with all the real systems except for the very high odds
horse racing system.
256
34.4 Recommended Staking Plans Leaderboard
iTSM Staking replaces LP28 Staking for Type 1 betting systems, and runs close on Type 2 and
Type 3.
257
35 Whitaker Staking
Rating
Overview Combines the successful characteristics of the best staking plans to create
the simplest and most efficient staking plan ever made.
Concept: Use Percentage Staking or Target Profit Staking depending on the odds of
the bet.
Other names: None
Variables: Odds range for each bracket
Stake settings for each odds bracket.
Having analyzed all the major staking methods out there, the leaderboard from the end of
Chapter 34 shows us the best staking plans for betting systems of various types and odds
levels. (i-TSM Staking, LP28 Staking, Target Profit Staking, Secure Staking)
The problem is that sometimes we don’t know what kind of system we are using. If our
historical data is flawed or incorrect we may end up using the wrong staking settings. Another
problem is that complicated staking plans can lead to mistakes or bettors giving up as they
become too time-consuming to monitor, track and plan bets. If we are using different betting
systems together, how can we plan our bank then? Do we have separate banks and calculate
different stakes for all of them? Splitting up betting banks into smaller betting banks is over-
complex and reduces capital efficiency. We need a staking system that takes the
successful characteristics of the best staking plans and makes them into a simple, easy
to use staking plan that does not require historical data or separate banks being
logged. My new and original staking plan does just that.
258
Whitaker Staking combines Percentage Staking, Target Profit Staking and Secure
Staking to create a new, simple and efficient staking plan. All you have to do is look at
the odds of the bet you are making at that moment, then either stake a certain percentage of
your bank or stake a certain amount to win a target profit. Whether you are making your first
bet in a series or the 100th, all you need to look at is the odds, then apply the right staking
setting to that. The stake depends on the odds bracket. This can all be done with one
betting bank. It does not matter if you are betting with various systems at the same time,
your own personal bets, tips from paid sites or big data based betting systems from Betaminic.
It does not matter if you are mixing sports or even multiple bets. You just bet based on the
odds of that selection. The four odds brackets are so simple that after a while you will be able
to remember them without even looking.
For odds under 2.00 we use Percentage Staking because Target Profit Staking at low
odds can lead to very high stakes being used. The recommended staking systems of i-
TSM and LP28 for odds under 2.00 are over complex and require knowledge of your average
win rate (i-TSM) and needs bet tracking so that you know your current win rate (i-TSM) or
what bet you are in the recovery series (LP28). i-TSM becomes much easier when you use
staking software like The Staking Machine, but it requires that extra step of logging each bet
before you know the stake for the next bet. If you see a good opportunity on live betting on
your mobile when you are away from your software, it becomes harder to place bets on the
go. With Whitaker Staking, you can easily calculate the percentage of your current bank.
Betting 15% of your bank may seem a lot, but actually, some bettors say that it is difficult to
follow staking plans that recommend only putting small amounts on sports events we are very
confident about. If you have 100 euros and you limit yourself to bet a maximum of 5 euros on
Manchester City at home to Sheffield United, then you can feel tempted to break your rules
“just this one time” and bet more, which are the first steps to abandoning a staking plan.
When your staking method actually allows you to place 15 euros, a much bigger chunk of your
bank, on such fixtures, then it becomes easier to keep the rules of the staking plan and follow
it to success. Also, if you know the success rate of a certain sports trading technique on
betting exchanges, you can also use that to guide how much or your current bank you risk on
the next trade. For example Lay The Draw (LTD) trading where you trade out for a loss after
60 minutes if the game is a still a draw has a an average success rate of 80% which equates to
odds of 1.25. This means that LTD traders can feel safe risking 15% of their bank on each
trade.
259
For odds over 2.00 we use Target Profit Staking to win a percentage of the bank. This
means we get risk adjusted stakes for higher odds. Again we avoid the problems of LP28
which requires record keeping to know your position in the recovery series.
Both the Percentage Staking and Target Profit Staking are for percentages of the
current bank. As the bank increases, so do the stake sizes, as the bank gets smaller,
the stakes reduce. This reduces the chances of bankruptcy.
The one other system that did well for a mixed range of odds was the Secure Staking method.
In the table below, it shows that Whitaker Staking (which uses Target Profit Staking instead
of Percentage Staking for odds over 2.00) is an improvement on Secure Staking in all
cases (20 out of 20 real betting systems tested). Whitaker Staking also improves on the
Recommended Staking plans in 15 out of 20 real betting systems tested. The chart compares
Whitaker Staking with Basic Level Stakes, the Recommended Staking plans from the
leaderboard (i-TSM Staking, LP28 Staking, Target Profit Staking, Secure Staking) and Secure
Staking.
260
Colossus 10 Away Dog
3.83 209 452 216% 450 215% 452 216%
Draw
Pro5 Undefeated
4.11 340 645 190% 596 175% 515 151%
Home Dog
Pro8 Underestimate
5.2 136 395 290% 394 290% 305 224%
Underdog V2
Horse Racing -85 -72 -90
13.05 72 -215% -100% -125%
Tipster Boris inefficient inefficient inefficient
Positives Negatives
Staking proportions of the bank protects against Working out Target Profit stakes
bankruptcy. requires an extra step of calculation.
Risk adjusted staking
Simple
Does not require complex record keeping
261
Can use multiple betting systems with one bank
Can be easily memorized for on the go betting
Betting larger amounts on low odds and smaller
amounts on high odds is mentally easier to follow.
Stakes increase as the bank increases
The most capital efficient staking method
Since Whitaker Staking depends on the odds, the uniform odds from each artificial data set
mean that Whitaker Staking generates the same data from Percentage Staking and Target
Profit Staking in previous chapters, so I will not repeat those results here. Only the Type 8,
the mixed data set, changes. This is why it easier to see the benefits of the Whitaker Staking
plan by looking at mixed artificial data sets (below) and real betting systems (above) that have
a range of odds values in them.
SB* (Separate Banks) The Whitaker Staking plan is not in any currently available staking
software, so these figures were calculated by adding together the 6 separate data sets of the 6
betting systems. This in effect means that 6 separate betting banks were used and then added
together. For the sake of direct comparison, I did the same thing with Secure Staking SB*.
This shows that Whitaker Staking is still an improvement on Secure Staking. It has higher
highs, higher lows and also a narrower range of results. This is all evidence that Whitaker
Staking is a more efficient way to use betting capital than Secure Staking. However, both of
them are excellent ways of using multiple betting systems in an easy and simple way.
Hopefully, soon this new staking plan will be added to the TSM (The Staking Machine)
software and the results can be calculated for one bank with even more certainty.
262
35.2 Recommended Staking Plans Leaderboard
Whitaker Staking replaces Secure Staking for Type 8 Mixed Odds systems.
263
36 Final Ranking of Staking Plans
We have analyzed 33 staking plans. Let’s see how they rank for each odds range and type of
betting system before finally seeing the most important ranking which is that of mixed betting
systems with the widest variation in odds. This is the main aim of this book, to show once and
for all the best and worst staking plans and to do that on a level playing field where all the
staking methods can be compared fairly and equally. If we have a 100 point bank, which
staking plans do the best and worst. Here are the final ranked tables.
Ranking Order
The staking plans have been ranked from lowest to highest average number of bets to double
bank. This, by default, also means that they are ranked from highest to lowest average total
profit.
Special Rules
If a staking plan has a higher average profit than another staking plan but was inefficient or
bankrupt in the Monte Carlo Simulations, then it automatically drops below the other plans.
So there is another order of efficient staking plans, then inefficient staking plans, then
bankrupt staking plans, which are then ranked in order of highest average profit in turn. This
is because even if a staking plan has good results, if it is inefficient or goes bankrupt
sometimes, then it is not better than another staking plan with lower profits that improves on
basic level stakes in all cases.
264
36.1 Final Ranking of Type 1 Systems: Low Odds Range 1.01-1.49
265
Real Betting System Ranking: Colossus 3 (19 bets per month average)
266
36.2 Final Ranking of Type 2 Systems: Odds-On Range 1.50-1.99
267
Real Betting System Ranking: Pro 1 Away Day Blues (24 bets per month average)
268
36.3 Final Ranking of Type 3 Systems: Evens Range 1.90-2.10
269
Real Betting System Ranking: Colossus 05 Against the Trend Overs
(34 bets per month average)
Ave.
Ave. Range Bkrpt. Ave.
R Bets P/L P/L P/L
Staking Plan Total from Chnc. Mnths
# To Dble Low High Incr.
Profit Ave. % to Dble
Bank
1 Coup Master Staking 298 945 836 1212 19.89% 0 9 394%
2 Whitaker Staking 302 932 900 1090 10.19% 0 9 388%
3 Percentage Staking 306 920 880 1209 17.88% 0 9 381%
4 LP28 Staking 307 919 806 1118 16% 0 9 381%
5 Target Profit Staking 310 910 870 1115 13.46% 0 9 540%
6 i-TSM Staking 321 879 785 1020 13.37% 0 9 360%
7 Secure Staking 337 836 810 938 7.66% 0 10 337%
8 Professional Staking 343 823 304 1154 51.64% 0 10 330%
9 Rolling Doubles Staking 345 816 459 1134 41.36% 0 10 327%
10 Parlay Staking 364 775 458 1058 38.71% 0 11 305%
11 Bookies Bank V2 Staking 369 763 388 1166 50.98% 0 11 299%
12 4321 Staking 426 661 425 866 33.36% 0 13 246%
13 Bookies Bank V2 Phantom Bets 487 579 298 798 43.18% 0 14 203%
14 Kelly Staking 514 548 382 1102 65% 0 15 186%
15 1234 Staking 723 390 111 601 62.82% 0 21 104%
16 Square Root Staking 1,072 263 -7 593 114% 0 32 37%
17 Level Stakes 1,128 250 196 296 19.84% 0 33 31%
18 Simple Level Stakes 1,476 191 191 191 0.00% 0 43 0
19 Recovery Type 3 Staking 162 1,743 -99 2501 74% 0 5 812%
20 Stop at a Winner Staking 447 631 184 1084 71.32% 0 13 230%
21 1326 Staking 614 459 148 799 70.92% 0 18 140%
22 D`alambert Staking 662 426 62 812 88.03% 0 19 123%
23 Phantom Bets Staking 921 306 85 517 70.59% 0 27 60%
24 Recovery Type 1 Staking 946 298 79 502 70% 0 28 56%
25 Reverse Labouchere Staking 1,389 203 145 567 103.94% 0 41 6%
26 Up-Down Staking 1,431 197 184 214 7.61% 0 42 3%
27 Labouchere Staking 1,476 191 127 256 33.77% 0 43 381%
28 Retirement Staking 1,985 142 104 164 21.13% 0 58 -34%
29 Bookies Bank Staking 2,104 134 129 135 2.24% 0 62 -29%
30 1234 Winner Staking 2,202 128 -1 229 89% 0 65 -33%
31 Fibonacci Staking 9,397 30 -6 48 90% 0 276 -84%
32 Martingale Staking 18,793 15 15 17 6.67% 6.5 553 92%
33 Recovery Type 2 Staking NA 7 -1178 44 8728% 2.4 NA -96%
270
36.4 Final Ranking of Type 4 Systems: Over-Evens Range 2.00-2.50
271
Real Betting System Ranking: Betaminic 5 Over 2.5 Against the Trend
(9 bets per month average)
272
36.5 Final Ranking of Type 5 Systems: Medium Odds Range 2.50-4.50
273
Real Betting System Ranking: Colossus 17 Dog Draw (84 bets per month average)
274
36.6 Final Ranking of Type 6 Systems: High Odds Range 5.00-10.00
275
Real Betting System Ranking: Pro 8 Underestimated Underdog v2
(18 bets per month average)
276
36.7 Final Ranking of Type 7 Systems: Very High Odds Range 10.00-30.00
277
Real Betting System Ranking: Horse-Racing Tipster “Boris”
(165 bets per month average)
278
36.8 Final Ranking of Type 8 Systems: Types 1-6 Mixed
279
Real Betting System Ranking: 6 Betaminic Strategies Mixed Together
(199 bets per month average)
280
37 The Best Staking Plans
What is the best staking plan? It depends on your betting system, your betting style
and if you know your betting history.
Simple is Best
If you don’t have a betting system and just bet based on your experience, sports knowledge
and gut feeling, or if you don’t have time to keep records of your bets or just don’t want to,
then Whitaker Staking or Secure Staking are the best staking plans for you. They adjust the
stakes according to the odds and are thus risk sensitive. They bet more on safer bets and less
on risky bets. The stake sizes changes as the betting bank does and they are very easy staking
plans to use. Look at the Type 8 Betting System results to see how they fared and see their
chapters to see the right settings.
281
More Effort for More Reward
If you do have a betting system and are able to keep accurate records of your betting history,
then more advanced staking plans tailored to your betting system could squeeze even more
profit out of a winning system. By knowing your average betting odds and range of odds, it
may be more effective to use more complicated staking plans. Look at the Type 1 to 7 Betting
System results to see how different kinds of staking plans work well for different odds ranges.
LP28 Staking, Target Profit Staking, Recovery Type 3 Staking, i-TSM Staking,
Bookies Bank V2 Staking, Parlay Staking, Rolling Doubles Staking and Coup Master
Staking are all worth considering. If you do have staking software, then these are the staking
plans to back test your data with in addition to Whitaker Staking and Secure Staking.
282
4321 Staking A simple but effective staking plan that reduce stakes after a winner to take
14
advantage of consistent betting systems.
Square Root Staking
15 A controlled way to increase stake size when in profit.
Simple Level Stakes All betting systems must make profit at simple level stakes. It is hard to go
16
wrong with this staking plan.
Phantom Bets Staking An interesting idea that manages to increase the ROI, but unfortunately
17
reduces overall profits due to a smaller bet volume.
Bookies Bank Staking
18 A recovery plan that is less aggressive than others.
1234 Staking
19 Chases losses. Not good for systems with most odds over evens.
Professional Staking This is an aggressive recovery staking plan that increases stakes not just to
20
win back losses but to also achieve a target profit
D`alambert Staking A simple and poor recovery method. The stake increases are too aggressive
21
making it inefficient for almost all odds levels.
Bookies Bank V2 Phantom Bets To bet more after a winner by aggressively recovering losses we didn’t
22
actually have.
Recovery Type 1 Staking The stakes increase too much in this plan, meaning the initial stake levels are
23
too low and overall profit is reduced compared to other staking plans.
1326 Staking
24 A terrible, illogical staking plan that can only work if lucky.
1234 Winner Staking A terrible system if used in practice, but created and analysed here for
25
comparison to 1326 Staking.
Retirement Staking
26 A capital inefficient loss recovery plan that does not beat simple level stakes
Reverse Labouchere Staking
27 A stakes focused plan that seeks to take advantage of winning runs.
Labouchere Staking A recovery method that focusses on stakes used instead of odds and profit,
28
which leads to rapidly rising stakes and bankruptcy.
Up-Down Staking Adjusts stakes by increments depending on wins and losses. Essentially a
29 recovery system that doesn’t reset to the original stakes after a win, but
merely reduces the stake level.
Stop at a Winner Staking
30 Martingale with stop losses combined with Target Profit Staking
Fibonacci Staking
31 A dangerous, aggressive progressive recovery staking plan.
Recovery Type 2 Staking An unwise plan that bets large amounts after a win, effectively needing two
32
wins in a row to recover previous losses.
Martingale Staking
33 A terrible, dangerous and ruinous staking method that should never be used.
Color Key
Best Plans Great Plans Good Plans
Use anytime. Worth Considering Useful in some cases.
OK Plans Bad Plans Worst Plans
Not bad. Not worth using.. Never use.
283
38 Practical Staking and Betting Issues
38.1 Simultaneous Betting (Placing multiple bets at the same time.)
A lot of staking plans, such as percentage staking, are calculated as if each bet is placed and
resolved before the next bet is made. Thus the result from the previous bet is then used to
calculate the stake for the next bet. But the practical realities of sports betting are that several
bets are often placed at the same time and then the next bets placed while those bets are still
unsettled. The problem then arises, how do we use staking plans with simultaneous betting?
If you can bet each selection one by one, then use Whitaker Staking LC settings.
If you bet multiple selections simultaneously, use Whitaker Staking DP or NLC settings.
Whitaker Staking LC Whitaker Staking NLC
(Betting 1-by-1) (Simultaneous Betting)
Odds Bracket Settings Odds Bracket Settings
1.01-1.49 Bet 15% of bank 1.01-1.49 Bet 4% of bank
1.50-1.99 Bet 5% of bank 1.50-1.99 Bet 1.25% of bank
2.00-9.99 Bet to win 5% of bank 2.00-9.99 Bet to win 1.25% of bank
10+ Bet to win 6% of bank 10+ Bet to win 1.5% of bank
285
38.2 Record Keeping
For professional betting, and using more complicated staking plans, keeping a record of bets is
essential. Even for the casual bettor, if it is possible to keep a basic record of what kind of
bets, odds and stakes you have used, then it can be useful to find out what kind of sports,
odds ranges and events you have a bigger edge in.
Of course, Excel is a good starting point, but there is software and websites available now
that will log and record your bets for you, making it easier to analyze your performance and to
calculate the next stakes in your staking plan.
The Staking Machine (www.thestakingmachine.com) is a great piece of software that you
can record your bets in and it will automatically give you the stake for your next bets. Once
your data is in it, then it becomes easy to analyse your betting history to see what staking plans
are best for your style. You can also import excel or CSV files of bet history into the software
for analysis.
Bettingmetrics (www.bettingmetrics.com) is an award winning website that combines record
keeping and betting history analysis with an oddschecker function that can be linked with your
Betfair or Smarkets account to place bets directly through the site or import your betting
history from those sites and analyse performance.
286
38.6 Get The Best Odds
I cannot stress the importance of getting the best odds more. There are many odds checking
and comparison sites out there now, so there is no excuse for not being at least aware of the
average market prices at the time of placing your bet if you are doing it manually. Getting just
a few ticks higher on a price over the long term can make a huge difference to your profits.
For example, if you followed the Betaminic strategy Colossus 17 Dog Draw, and bet on
Pinnacle’s closing odds, it would have made 598 points profit over 6 years, but if you had
managed to get just one tick higher on your odds for each bet on another website (e.g. for the
3-4 odds range, one tick is 0.05, that means betting on odds of 3.68 instead of 3.63), you
would have got an increase of nearly 20% in profits. If you managed to get 2 ticks higher, a
38% profit increase. Over the long term, getting even one tick higher on your odds means a
big difference in profit. So if you bet manually, it is important to check around for odds. If
you bet on exchanges, always request a higher price when you bet. Another alternative that I
have heard about but never used myself are bookmaker brokers. This kind of service gives you
one account to deposit money into and place your bets at. The broker has access to several
bookmakers. The broker service automatically places the bets with the bookmaker who has
the best odds when you request a bet.
If you have always used one bookmaker, then you are probably still eligible for introductory
offers at other bookmakers. It can be annoying to jump through the hoops of introductory
offers, but if you have the time and are able to, it will reduce your risk by getting a few “free”
bets and if you are building up a betting bank, it can really help.
At some stage (hopefully) you will be making so much profit that highstreet bookmakers may
start to limit the size of bets that they will accept from you. When that happens there are other
options. Pinnacle is a bookmaker that never limits customers. They operate on a high
volume, low margin business model and so do not care if people are winning too much
because they make money on volume, not on having losing customers. Another option is the
betting exchanges Betfair, Smarkets, Betdaq and Matchbook. These sites do not ban
winners because you are betting against other people, not the site. They want your business,
but you pay commission on your winnings. Finally, there are Asian bookmakers who handle
large volumes and are an option if you are blocked from European bookmakers. There is no
real need to worry about getting limited by bookmakers since if you get to that stage you can
use Pinnacle and Betfair to get your betting done.
287
39 Betaminic and Big Data Betting
I used the Betaminic betting systems (www.betaminnic.com) in this book because I needed
real, actual working betting systems with large amounts of back data to test the staking plans
on. But what is Betaminic?
I spent 20 years looking for a working betting system. Finally, I found Betaminic. Their big
data analysis tool enabled me to sort through over 7 years of bookmaker odds and football
data to find what betting systems worked and what did not. After I found working systems,
then I started to look for the best staking plan to go with those systems. It has taken me 1 year
to research the best staking plans and thus this book.
We know that some people consistently make money from sports betting. We have heard the
stories of bookmakers banning users who win too much or too often. We can read on Betfair
of the Premium Charge and we know that some people are paying it. All of this made me
certain that there were working betting systems out there. I had spent years researching
different systems and in all cases I needed to backtest my theories. I was often limited by the
data I had and my ability to analyse and process it with Excel. But when I found the Betaminic
website, my data-entry, number-crunching days were over. What had previously taken me days
of fiddling about with downloading data packs, editing formulas and linking database files to
check just one theory could now be done in minutes with Betaminic and its Betamin Builder
Tool. It was a big data geek’s dream. I used to think backtesting with 1,000 games was
enough, but now I can access much more data. How much data exactly?
I can sort matches and evaluate strategies with statistics that it would be impossible for me to
calculate myself:
Home win/draw/away win percentages for each league based on the last 50 matches
Goal averages per match for each league based on the last 50 matches
Win/draw/lose rate for each team based on the last 20 matches.
Goal averages for each team based on the last 20 matches.
Consecutive win/draw/lose streaks for home/away/overall games.
Consecutive scoring streaks of over or under 1, 2 or 3 goals for home/away/overall
games.
Consecutive conceding streaks of over or under 1, 2 or 3 goals for home/away/overall
games.
Consecutive over/under 2.5 goal match result streaks
Win/draw/lose percentages for the last 3,5,10 matches or for all the season’s matches.
288
Not win/draw/lose percentages for the last 3,5,10 matches or for all the season’s
matches.
Over/under 2.5 goal match percentages for the last 3,5,10 matches or for all the
season’s matches.
Head-to-head win/draw/lose percentages.
Head-to-head not win/draw/lose percentages.
Filter matches by odds for home / draw / away / underdog / favourite / over 2.5 goals
/ under 2.5 goals
Filter matches by date, season, game month, custom date range
Betaminic has given the ordinary punter an amazing tool for finding, testing and using betting
strategies based on big data that has previously only been available to those with the advanced
IT skills, the data and the time to do it. You can use the tool for free to find and test
strategies. And once you do find your perfect system, Betaminic will email you picks from
your system each day for a fee. So you will not have to waste hours trawling fixtures and
statistics websites for your next bets. It is amazing. The tool helps you sort through matches
and find the kind of matches where the bookmakers have consistently miscalculated odds with
their pricing models. It helps you find value bets. And then you can beat the bookies.
You can sign up to their website for free and get access to the Betamin Builder.
Once inside you can check the “public strategies” area to see what the best strategies are.
This red box shows a system called “Over 2.5 Total Share” which is a shared system by a use
called “Dell73”. It is a system that bets only on the over 2,5 goals market. The data set has
2,645 bets made so far over the past 6 years which is about 29 bets per month. It makes an
average of 2.99 points profit per month with average odds of 1.94. That would make it a Type
3 betting system according to my research. With Whitaker Staking, that 2.99 points could be
increased by 300% to about 12 points profit per month. The “Max DD” is the biggest losing
289
run. That means its biggest losing run was -20.73 points. In total this system has made 249.75
points profit which is yield (ROI) of 8.25%.
Betaminic helps to find “value” betting systems where the bookmaker odds are wrong and
bets can be made above the actual probability level. Over the long term profit can be made,
but it is a long term strategy. With the right staking plan, the systems that perform fairly at
simple level stakes can be supercharged to produce even more. The more advanced staking
plans work well when put together with consistent betting systems that have similar
bets at similar odds over time. Betaminic can provide exactly those kinds of systems,
which is why I recommend people take a look at it. The more people who sign up and
research their football gut feelings and ideas, the more winning systems will emerge on the
public strategies page. Or you can just follow the existing systems now by clicking on the
“follow” button. It is that easy.
One of the best ways to use big data based betting systems is to combine several together in a
portfolio of systems. In this way the risk is spread among the different systems. Betaminic has
an “aggregator” tool on the website so that you can see the effect of combining different
systems. In this book I took 6 systems from the Betaminic site, they were chosen since they
represented different odds bands, they were not necessarily the best performing systems on
the site. In actual fact, 3 highly recommended systems are “Colossus 06 Scoring Away Overs”,
“Away Win DNB strategy by benhur” and “Pro-Home”. Combining these strategies can
diversify the risk and maximize the profits.
290
39.1 Three Recommended Betting Systems from Betaminic.com
1. One of the most reliable strategies for the over 2.5 goals market is "Colossus 06 Scoring
Away Overs".
After more than 300 selections sent to users, this strategy maintains an impressive yield of
11%, which proves the strength of this strategy, even improving on the results of the historical
total, which already has more than 2,600 picks. A really important point here is that the
researched pattern (All results) has continued since the system has become active (Since
shared).
2. A strategy with solid results, where the stake is refunded in the case of a draw: "Away Win
DNB strategy by benhur"
291
Again, a strategy that since it was published, exceeds the average profitability, after 300
selections sent to users. This is a very efficient strategy for the Draw No Bet market, with
relatively high odds, but with a low Drawdown.
3. A strategy designed for more conservative bettors, but achieving a good return: "Pro-
Home"
This is a good strategy for the most conservative bettors, with average odds of 1.36, and with
a very low Max Drawdown, which translates into a lower risk for the user. Despite being a
low-risk system, the yield of 6% makes it a very interesting strategy to follow even by itself. If
used with the right staking plan, those results could be dramatically increased. Whitaker
Staking increased profits for this betting system by about 1,300% which, if the pattern
continues, could turn an average of 1.6 points profit per month into an average of 20 points
per month. However, Betaminic’s recommendation is to incorporate it into a portfolio to
reduce negative streaks.
292
By combining the three systems together, we get a very nice looking, but simple portfolio. It
has an average of 88 bets per month making an average of 6.77 points profit at average odds
of 2.00. With Whitaker Staking these results could be increased by 400-500% to an average of
30 points per month. Betting on big data based systems is a long term strategy, so the average
points per month is just that, an average, there will be losing months, but over the long term,
positive figures.
293
40 Staking Plan Analysis Service
It can seem a bit confusing with all these staking plans. If you have a betting system and you
want me to test it to find the best custom staking plan that matches your unique system, then
I offer a staking plan analysis service. All I need is an Excel or CSV file with past data of
“odds” and “result”. That’s all I need. Email me at [email protected] for more
information. You can read more information about the service and fees on my website
www.thestakingbook.webs.com.
294
41 Free Betting Systems Books Offer
I have written two other books about betting systems. They introduce the Betamin Builder
and various betting systems created with the Betaminic site’s data analysis tool.
It is possible to get free pdf versions of these books by signing up for Betaminic and
becoming my referral when you buy a pick pack. It is free to become my referral. It costs you
nothing extra on top of the pick pack, but you can get many benefits.
New Users:
1. Go to www.betaminic.com/?ref=TomW or click through my website link at
www.betaminic.webs.com to register for a free account to get access to the tool.
When you click on the link, it will take you to www.betaminic.com the main site. Don’t
worry that my referral link is not in the address bar.
2. Buy any package of picks. (Even the smallest one is okay at 29.75 EUR for 25 picks plus all
the books and bonus material now and in the future.)
3. Email me the order number from your purchase to [email protected]. I will
pass this order number to Betaminic so they can confirm the referral process is complete.
4. I will send you all the bonus material by email.
295
Existing Users:
1. Go to www.betaminic.com/?ref=TomW or click through my website link at
www.betaminic.webs.com to login to your account. When you click on the link, it will
take you to www.betaminic.com the main site. Don’t worry that my referral link is not in
the address bar.
2. Buy any package of picks. (Even the smallest one is okay.)
3. Email me the order number from your purchase to [email protected]. I will
pass this order number to Betaminic so they can confirm the referral process is complete.
The order number must be from a package of picks bought after you have clicked on my
referral link. Old order numbers will not be acceptable to Betaminic.
4. I will send you all the bonus material by email.
Regards,
Eric Cantona
If possible, please tell me which strategies you are following or plan to follow. Then I know
which strategies to pay special attention to.
Hi Tom,
Regards,
Alan Shearer
If you want to tell me more about your betting style, staking level, research methods, future
plans and how long you have used Betaminic, please do. I would love to get a good picture of
what you are doing. Also, feel free to ask any questions you have.
296
42 Conclusion: The Final Word
I started researching staking plans because I was confused by the myriad of plans out there. I
wanted to know what the best staking plan was. To do that I made them comparable by
limiting them all to 100 point banks, applying them to artifical and real betting systems results,
and then discarding bankrupt and inefficient ones, to see which would give the best results. In
the end my answer was not as simple as I had hoped, but on the other hand I found out a lot
of important features of staking. The two most important things I learned are as follows.
I hope this book was useful for you and gives you new insight into how to apply staking plans
to betting systems. Hopefully now you can be confident that you will be using the best staking
plan for your betting system and getting the most out of your edge. And to finish, just in case
there is any confusion over what my recommended staking plans are, here they are again.
298
Other Non-Betting Books
299