Fundamentals of Sports Betting (Betting on Over/Under ‘X’ Goals) is unlike any other book you will find elsewhere. It contains insights into how mathematics and statistics are applied to bookmaking, information that is rarely (if at all) found in the public domain.
Have you ever wondered how bookmakers set their odds? They must be pretty good at it to remain trading in a high risk industry built on small margins!
Remember folks it’s all about finding the “edge”.Image: Cartoonresource (Shutterstock)
It is common knowledge that the gambling industry as a whole relies to a large extent on the ignorance of its customers and, only by understanding how bookmakers think and act, will you ever be able to compete with them on a level playing field.
The Most Valuable Book About Betting You Will Ever Buy!
This course is designed to give you the essential, fundamental knowledge necessary to understand odds calculation and the bookmaker market.
It deals with relatively simple descriptive statistics and teaches you how to look at data sets, calculate your own probabilities and odds, analyse the market odds on offer, and make informed decisions when predicting football results.
Amongst the topics you will work through are distributions, deviations, graphs and charts, odds calculation, financial terminologies, risk management, and of course, how to identify ‘value’ in the betting market.
You will gain a deep understanding of the many different elements required to understand the bookmaker market and odds calculation. Many false beliefs that the majority of gamblers and fans of football have about betting will be exposed and stripped away.
Readers will also be enlightened to learn about how odds are set in the market, where to find ‘errors’ in market prices, how to evaluate data and graphs, and much, much more.
What Do You Get For The Money?
The course comes in electronic format and the bundle includes the course book (PDF*) and a Bundesliga cluster table (Excel) for studying. This course employs a didactic method of teaching, which is an educational technique. It is a very structured style of learning and as such, each section of the course should be mastered before moving on to the next.
To further encourage you to learn, there is a plethora of exercises to practice what you have learned, and the solutions to the exercises are found at the end of the book, sometimes embellished with further explanations.
All of the course lessons are presented in a pragmatic, easy-to-follow, step-by-step fashion, with no more than passing respect towards the sport of football and the passion of its fans. After all, this book has been written by a lady who loves maths but dislikes football…!
In addition, you will receive the very latest Bundesliga table for the current season, so that you can put into practice what you have learned and within the course, you will find a discount code enabling you to download for free the German Bundesliga sheet in future seasons, meaning that you will always have access to the most current Bundesliga cluster table, forever: it doesn’t matter when you bought the course.
*By the way, the PDF is a professionally formatted document and if you have a double-sided printer you can print your own book for studying.
Fundamentals of Sports Betting
Course with Cluster Tables for the German Bundesliga
Over/Under ‘X’ Goals
PRODUCT SUMMARY
- Format: PDF & XLS
- Download Size: Course (PDF): 5.2 MB & Excel file: 782 KB
- Publisher: Soccerwidow Ltd; 2nd Edition (July 2016)
- Pages: 169 (almost 38,000 words)
- Tables and graphs: 70
- Example tasks: more than 80 (with solutions)
- Language: ENGLISH
Only £59.00*
PDF document professionally styled for easy reading and printing
For traditionalists, the PDF has been professionally formatted for double-sided printing. If you like, you can print your book in full-colour with a cover and add slick spiral binding.
Those who prefer to read the document on their computers will find the chosen font (Myriad Pro) easy to read on the screen. Myriad Pro has a clean sans-serif aesthetic which makes it highly accessible.
The electronic version also contains a plethora of easy-to-navigate links to help the learner find his way around the document.
Contents at a Glance
Section A: Basic Knowledge
Probability, Betting Odds, Value, Yield, Profitability
Basic Statistical Terminology
- Goal Distribution and Percentage Calculation
- Deviation from the Mean
- Standard Deviation – The Main Measure of Variability
- Precision, Trueness and Accuracy
Betting Odds Calculation
- Probability and Betting Odds
- Opening Betting Odds Range
- Calculation of ‘Zero’ (Fair) Odds
Risk Management Control
- Financial Terms: Yield, ROI and Profitability
- Risk Forecasting and Evaluation: Value of a Trade (Bet)
- Preventative Measures: Setting the Starting Bank
Section B: Developing a Betting Strategy
Market Dynamics, Cluster Groups and Betting Tables
Market Dynamics
- Betting Odds are Prices of Bets
- Falling Odds represent Price Increase
- Bookmakers adjust Betting Odds to Public Opinions
Building of Cluster Groups
- What are Cluster Groups?
- Clustering Depending on ‘Strength’ of the Team
Betting Tables: Over/Under ‘X’ Goals
- Goal Distribution by Team
- Team Calculation Tables
- Standard Deviation Tables
- Using Calculation Tables to Determine Betting Odds
Finding Value Bets
- Everything that Glitters is not Gold
- Method I: Value Betting using Cluster Group Tables
- Method II: Value Betting using the Value Calculator
- Cluster Groups, Value Calculator and Bets Identification
…or ask any questions in the comment section below.
Hector
Andrew
Pete
Sumiko
Ian
Janice
Mark
Hi Soccerwidow,
Nearly at the end of the course! Just stuck on question (D) of exercises 7.4 (page 130) where we put the value calculator into action.
For Hertha Berlin vs Stuttgart
I’m getting different results in my value calculator compared to the answers on page 151. I notice from the comments section that another of your students had similar issues but I wasn’t sure which was right.
For true odds using the value calculator for Hertha vs Stuttgart, I had:
0 – 0: 20.00
U 1.5: 3.49
U 2.5: 2.07
U 3.5: 1.49
U 4.5: 1.16
U 5.5: 1.01
U 6.5: 1.01
Not the same in the answers on page 151, which had
0 – 0: 23.98
U 1.5: 3.20
U 2.5: 2.00
U 3.5: 1.58
U 4.5: 1.19
U 5.5: 1.01
U 6.5: 1.01
What worries me is that I also got different results for Hannover vs Dortmund when I used the Value Calculator then (page 72). Since I want to try to use the value calculator as my main tool to detect value bets, I’m a bit worried I’m doing something fundamentally wrong. I don’t seem to get too much wrong when I do other parts of the course, by contrast. But I’m yet to get the same answers as the back of the book with the value calculator, despite diligently watching your excellent YouTube tutorials.
I can email you the spreadsheets or screenshots of them. Likewise, if you have other past examples of using the value calculator for specific matches, I could have a try at them and we could compare notes. Or if you plan on using the value calculator for any matches this weekend, I could have a try at those specific matches too.
I’d be most grateful for ANY assistance. Sorry for being a nuisance… again.
Kind regards,
Audiendi
Hi Audiendi,
this exercise in the course has unfortunately a bug. When putting together the examples I have for some reason missed one season what means that the 25 games that form the basis for the example calculations are incorrect.
Therefore, you have probably done everything right. I’m very sorry!
Just carry out the calculations with the figures shown on the screenshots in the course (the numbers for the last 25 respective games). Then, when you fully understand the calculations you can repeat the exercises with the correct data and from your provided example, you can already see that the ‘correct’ results only differ very slightly. That is why I haven’t worried about updating the course and rectifying that error.
The course is about understanding odds calculation, even if the example is ‘made up’ like it unfortunately happened in this one exercise.
The good news is that there aren’t any more errors in the course, at least none of which I’m aware.
Best wishes,
Soccerwidow
Hi Soccerwidow,
Just wondering for Chapter 7.2- Value Betting using Cluster Group Tables, I’m having a bit of trouble understanding the Cluster Groups.
I’ve opened the accompanying spreadsheet and gone to the Betting Tables tab. Taking Hannover at home, as an example. So, on page 120, there’s a copy of (part of) the table from the spreadsheet.
But what I don’t get is how we have separated Hannover’s 85 home games into four quartiles? And what’s the significance of each quartile?
So for example, with the first quartile. I’m at a loss to understand why the HO/AO range is from 0.000 to 0.604. Or why that first quartile is reflected by odds from 1.01 to 2.20.
I can sort of understand how, as Hannover is perceived as one of the weaker sides, it gets put in the bottom row (and it’s HO/AO quotient, against Dortmund, as 5.065 fits into that bottow row, between 1.1031 and 100.000). But I don’t understand why it’s those number specifically.
When I did the first thing on HO/AO quotient on page 105, I thought it made sense to me. But those clusters seemed to be based on bookmaker odds for specific matches between
team x and team y. Overall, I just can’t get my head around how the quartiles are arranged from 85 matches one team has played in, precisely what they signify and exactly what the quotient ranges and odds ranges refer to.
Sorry for such a long post from me which may have seemed a bit jumbled.
Kind regards,
Audiendi
Hi Audiendi,
The division of HO (Home Odds) by AO (AO Odds) reflects the “strength” of the teams perceived by the public. For example, if there is a match with “equally strong” teams, the HO/AO quotient is roughly between 0.8 and 1.2.
When building the quotient we’re not asking ourselves the question, “Are these teams really equal?” We do not calculate the 1×2 odds either, but simply use market prices and assume that the Bookmakers have reflected public opinion as well as possible.
Team ranking and public opinion is taken permanently into account in the 1×2 odds. That’s how bookmakers ensure that they stay in business and make their money!
‘Public Opinion’ for the 1,530 Bundesliga matches, seasons 2012-2017 (HO/AO quotients)
20%: 0.2918 The home team is the clear favourite with a very good chance of winning (much felt certainty for the outcome of the game)
40%: 0.5292 The home team is ‘definitely’ stronger than the away team, but there is also a good chance of a draw (opinion swinging between home or draw)
60%: 0.7830 The home team could win, but it could also be a draw or an away win (possible outcome is not really clear)
80%: 1,3172 The chances of a draw are pretty high; teams are felt to be equally strong (feeling is that both teams are evenly matched and the draw is probably the most likely outcome)
100%: 23,9400The home team is weaker than the away team; it ‘could be’ an away win(the favourite is the away team)
Take any Bundesliga weekend and cluster the matches according to the above HO/AO quotients (before kick off) and check against the distribution of results. You will see that they won’t distribute according to the implied probabilities of the AO/AO clusters.
There is a correlation between the HO/AO quotient and the O/U odds! I am currently expanding upon this whilst writing an update for the German version of the O/U course. Unfortunately, the explanation is not short and/or sweet enough to pack it into a comment. You’ll just have to believe me that the HO/AO quotient is a very important factor when judging if OU odds contain value or not.
Hi Soccerwidow,
OK. So, I calculate the HO/AO by dividing the 1 x 2 odds. Then I look at the cluster tables. I look at the the appropriate row (or quartile) of the table for, say, Hannover, to find out which odds correspond. And then I make sure it’s lower than the market odds and then do the same with Dortmund. That’s it?
But what I don’t understand is how to calculate the four quartiles of the HO/AO for each team, myself. What’s the process for separating the 85 previous matches into four quartiles? How does one build the cluster tables, oneself?
Cluster tables are crucial to a successful betting portfolio using the 5 seasons method, right? Let’s suppose I want to do this for Serie A. I’ll need to make a set of cluster tables, won’t I?
Kind regards,
Audiendi
Hi Audiendi,
you can unprotect the sheets (the passwords you’ll find in the legal tab) and then see how the formulas work and replicate them yourself for any league you like or, you can save yourself countless hours and simply buy the tables you’re interested in: http://www.soccerwidow.com/products/over-under-betting-cluster-tables/
Sorry Audiendi, this is a Betting course and not an Excel course. I’m not even going to try to explain advanced Excel formulas and how they work, and of course there are plenty of other websites dedicated to teaching Excel.
Hi Soccerwidow,
All good, all good. Sorry I’m sorry for being incredibly annoying. I’m such a stickler for trying to grasp all the finer details. I thought I had made a mistake by not knowing how to construct the cluster tables myself. So, the course doesn’t explicitly teach how to construct the cluster tables (in the way it teaches SD, RSD, %CV, expected losing streaks, etc.)?
I’d just buy the cluster tables, sooner than trying to replicate them. Much easier.
So, basically to put the five seasons method into practice, it’s necessary either to buy the relevant cluster sheets (or to replicate them)?
Kind regards,
Audiendi
Hi Audiendi,
this course is designed to introduce readers to the essential, fundamental knowledge necessary to understand odds calculation and the bookmaker market. The intention of the course is to introduce the subject of odds calculation to punters. It is therefore not a “five season method”; it teaches distributions, graphs, financial terminology, etc… all with the goal to show the reader how professionals (such as bookmakers) approach setting their odds.
Please look at the image above… to put the course into perspective… The course is an ‘introduction’ and if there would be ‘betting degrees’ at universities then students would learn the contents in the Bachelor Year 1.
Therefore, please work fully through the course first, until you understand absolutely everything. Then try your knowledge on the Bundesliga. If you feel that you would like to learn more, then consider purchasing the EPL as it is a much more liquid league on the exchanges.
To summarise, you must thoroughly understand the fundamentals before mastering odds calculation, and I’m afraid that like everything worth having in life it will take time to accomplish.
Hi Soccerwidow,
Me again, sorry. I wondering if there’s a mistake in the book or if I’ve miscalculated somewhere. It’s only a small difference, but I wanted to be sure I’m on the right track.
On page 89, for (G) (i) we have to calculate the probability of winning a bet for ‘Over 2.5 Goals’ using the Five Seasons Method compared to the Value Calculator Method. You’ve provided Zero Odds for both. Zero Odds for the Five Seasons Methods are 1.60. While Zero Odds for the Value Calculator method are 1.43.
So, to calculate the probabilities, I went…
Five seasons. Probability = 1 divided by betting odds
= 1 / 1.60
= 0.625
= 62.50%
Value Calculator. Probability = 1 divided by betting odds
= 1 / 1.43
= 0.6993
=69.93%
But in the answers (on page 146), it says the respective probabilities are…
Five seasons method: 62.36%
Value Calculator method: 69.89%.
Have I made a mistake somewhere?
Kind regards,
Audiendi
Hi Audiend,
Please remember how to calculate ‘Zero’ odds…
(1) Calculate the probability
(2) Convert the probability in odds
Not the other way round!
Five seasons method expected ‘Zero’ probability: 62.36%
1 divided by 0.6236 = 1.603592046 (rounded: 1.60)
Value Calculator method expected ‘Zero’ probability: 69.89%.
1 divided by 0.6989 = 1.43081986 (rounded: 1.43)
Hi Soccerwidow,
Thanks for that. I understand where I went wrong. All that was written, in the question, I’d forgotten that we’d already carried those calculations earlier. And it makes sense how a statistical probability is a more precise representation than the betting odds. As such, I understand what you mean by the idea of calculating the probability first and, then, converting it into odds.
Kind regards,
Audiendi
Hi Soccerwidow,
Another question, sorry 🙁
On page 72, exercise (D) asks me to ‘Calculate the game between Hannover and Dortmund using the Value Calculator Method and compare its suggested odds with the highest market odds.’
But, again, where do I source the relevant stats for the value calculator then? Thanks for your help last time. I figured out that the accompanying spreadsheet enables me to switch between Bundesliga teams to generate those stats (and then simply go back and substitute the relevant data). But, as far as I can tell, that spreadsheet is only suitable for the last 5 seasons approach. I can’t find the snapshot of the value calculator for these two teams.
Kind regards,
Audiendi
Hi Audiendi,
Sorry for that confusion. You are making me aware that I put too much information into the course and packed it full with knowledge that is very obvious to me but not necessarily for the reader. I will take that into consideration once I start reviewing the course. I’ll have to extend on the explanations!
Regarding your question… Data is always football-data.co.uk; there is no other source of free and reliable data available on the Internet. To help people with the Value Calculator here’s a video: How to prepare data for easy copying and pasting into the Value Calculator.
Hope that helps. 🙂
Hi Soccerwidow,
That’s okay! Actually, (I feel a bit guilty saying this), I’ve watched that video before and have learnt how to use the Value Calculator (although I’m not betting until I finish the course). I’d just forgotten how to use it and I wasn’t sure if the 25 match results were included in the accompanying spreadsheets.
I probably shouldn’t have troubled you over such a trivial thing, given I could have figured out the answer with extra effort. Anyway, slowly but surely we get through the course. I’m nearly at the halfway point.
Kind regards,
Audiendi
Hi Soccerwidow,
I’ve slowly been doing this course. I want to say a big thank you, I think I have learnt a lot.
Just one thing is annoying me no end. On p: 62, you write
‘… we need to calculate and to make it as easy as possible, with the knowledge that calculating everything exactly is virtually impossible and unnecessary, we will therefore use the %CV of the entire Bundesliga to calculate matches of individual teams.
Then, with respect to the correction factor for the Value Calculator, on p:68, you write:
‘… looking at the contribution of the 13 ever-present teams over the five seasons, the illustration [of the distributions of the respective goal differences] became more chaotic.
Therefore, the “regression towards the means” is definitely not the distribution of the whole league.
The “correction factor” is the H2H games!’
I realise that you’re talking about different ways of calculating odds. Nevertheless, it does seem to be a bit of a contradiction to me.
How do you reconcile the notion of studying the standard deviation around the mean (the expected value) for the entire Bundesliga as the best way of determining the range of fair odds with the notion that the entire league’s distribution is too vague and chaotic to be a precise correction factor?
Sorry if I’ve misunderstood something along the way or if I’m simply confusing different concepts. I certainly don’t wish to seem rude. I’m extremely grateful as I’ve learnt HEAPS from this. This particular question just has me SO baffled.
Hope you’re well, Soccerwidow.
Kind regards,
Audiendi
Typo…
[of the distributions of X Goals scored for the respective teams]*
Sorry about that. I did not intentionally misrepresent what you are teaching us.
Hi Audiendi,
The chapter you are currently in, teaches the calculation of ‘Zero’ Odds/ Fair Odds. There are many roads that lead to Rome and the same applies to calculating odds: There are many methods to estimate the possible future result. Therefore, please continue with the course and you will see that many questions that arise at the beginning are explained later when the various puzzle pieces start falling together.
The course teaches two concepts… one using 5 seasons’ data under the assumption that the performances of teams that have been playing for consecutive 5 seasons in the same league don’t change overnight (Cluster Tables) and the Value Calculator, a tool that calculates matches with teams involved that haven’t played for 5 consecutive seasons in the same league, but often enough to have a few H2H results available as correction factor.
Of course, these two tools lead to slightly different results, but as mentioned in the course many times: From a statistical perspective, football matches do not occur frequently enough and the margin of error is therefore always going to be relatively large.
Once you reach chapter 7. (p.116+) you will compare both, Cluster Tables and Value Calculator and see, that both reach similar but different conclusions.
As mentioned in chapter 1.4. (p. 28+) – High (good) trueness of the mean is achievable, but not high precision. Nobody can calculate totally exact, not even the bookmakers…
The above graph shows the distribution of Under 2.5 results in comparison with the odds (implied probabilities). Despite the bookmakers pricing their odds close to the true odds the really observed results are still zig-zagging around the implied probabilities of the odds.
I hope this helps to answer your question.
Best wishes,
Soccerwidow
Hi Soccerwidow,
Thanks for the considered and informative response. I guess it will start to make sense further down the track.
Another question, sorry.
On page 71, question B asks me to ‘calculate the odds for the Hannover vs Dortmund game (as in Fig. 27) for all ‘Under/Over X Goals’ bets using the statistics of the last five seasons.
But where am I supposed to source the statistics? Have we gone over those teams already in the course and I’ve simply forgotten?
Sorry to be a pest.
Kind regards,
Audiendi
Hi Audiendi,
To answer that question in the exercise you will have to return to chapter 2.3.3. (p.58+) and redo all the example calculations, but with Hannover & Dortmund (instead of M’gladbach & Hamburg).
You’re not a pest! Don’t worry. Learning is hard, and I certainly prefer learners that ask questions to those who give up too easily. 😉
Hi,
I backtested this strategy for the last seasons of EPL and Bundesliga 1, but both returned -EV, EPL – 90 EURO with 10 euro stakes and Bundesliga -10 euro with 10 stakes.
A total 139 bets were made, where I found value based on the calculated fair odds and the best odds available at the market for that match.
Do you have the results for the last seasons of EPL and Bundesliga 1 ? Am I doing something wrong?
Hi Sarkec,
no, you are not doing anything wrong. The only thing I would suggest is to leave the EPL alone. They don’t play as nicely ‘statistically correct’ as the Germans.
The main problem with betting is that even if you consistently bet with a mathematical advantage on your site, the result distribution is random. That means that everybody will be experiencing different P/L curves if they don’t chose exactly incidentally the same bets.
I’ve had a similar question to yours on another article and as reply I produced a P/L simulation. Just have a look at this: http://www.soccerwidow.com/football-gambling/betting-knowledge/value-betting-academy/simple-steps-win-over-under-betting/#comment-19567
Hello. Thank you for the reply. I imagine that the work that goes into the books and articles is considerable so if it happens, great, if not…that’s life.
I read the article and using auto translate and got the basic facts and seems interesting I will be testing this out over the next couple of weeks on the exchange. If it works (of course not every time but consistently) it could be a great tool for pre-match trading either to get a few free ticks or create free bets, either option seems very interesting.
I hope you find the time to translate the update for the Over/Under course to English. I think that the course is very good and would really like to see more of it.
Good luck with everything.
Hello. I bought the book a couple of months ago, went over it, did all the exercises. First of all I should say that I had some small prior experience with statistics but I knew very little about excel. The book is very interesting and as far as I was concerned it is not a hard read, and English is not my first language. After finishing the book I created a small version of a value calculator just for overs and unders as well as working with cluster tables and started testing.
I soon found out that back-testing would be hard for me because I have very limited access to bookmakers and have to rely on the exchange. That is why I am just writing now, because I needed some time to have at least a reasonable number of matches to form an opinion. My results so far are very interesting. I work with two probability clusters for the value calculator picks as well as using the ratio mentioned in the book for selection purposes.
The first probability cluster is >75, of course if there is value in the bet. As you can imagine the strike rate is high, losing streaks low and bank has shown consistent growth with 150 bets so far. The second cluster is >50<70 probability. Here the results have been less consistent, but still profitable after 117 bets. I should say I am using very very small stakes. It does take time, I am building my own database for results and head-to-head data for every championship that I get involved in. I also keep track of every bet I make including probabilities, final results, true odds, and exchange odds so that I can later try to identify ways to improve the selection process.
Using the clusters I do not have enough to form an opinion yet but the seem really promising. All in all the book is worth the money in my opinion. I look forward to the new book concerning 1 X 2 betting, if you actually have time to finish it.
I had one question. So far I just went for set and forget type of bets so no trading in play. What we calculate are true odds before the game starts. If i wait in-play for the odds on the 1.5 market, for example, to reach a point where they were value at the start of the game, is that bet viable?
Thank you very much for all your materials, I found them all very informative. Good luck with your future content and all the best.
Hi Vlad,
the odds in-play will go up to the expected odds; the problem is that there may be a goal the very first minute and you may not get your match. However, what you may consider, as you are interested in trading, is, to carry out the calculations one week (or at least a few days) before kick-off and then watch the odds moving in the direction of the calculated odds ranges within the last few hours before the match starts.
Sometimes they reach the calculated range before kick-off but quite often they will only move by a few ticks. However, they will still move in the expected direction. Have a look at this.
Coincidentally, I have just written an article on this topic. Unfortunately for Soccerwidow readers, it’s in German: Tipp für Trading: So werden Trades Smarter Simply run an auto translation over it. It should be understandable; I tried to write in as short sentences as possible.
With the 1×2 book English readers will have to wait a little longer. As my mother tongue and education is German, I feel much more comfortable writing in German. Sorry! The Over/Under course is currently undergoing a massive update in German (for example, a large chapter on trading has been added – above article is a part from the update). We’ll see if we get anywhere with crowdfunding to get the book in the book shops and, if the book sells well, then there will be an English translation following.
I truly hope that I will find the time to write the long promised book on 1×2 betting next year!