While there is nothing wrong with gut betting, it is not a viable strategy for long-term profit. If you are serious about betting, you will analyze the event and create your estimate of what the odds should be in order to find out if there is value in the bookmaker’s offer. Read this article to learn more about how to properly evaluate a soccer match.
Why bother pricing a football match?
To find a valuable bet, you need to compare the odds you are betting with what you think is a more accurate reflection of the true probability of the event. If the odds available lower the probability of the event coming from your estimate, this will give you a positive expected value.
While this is a fairly straightforward concept, most players fail to correctly assess the event. Comparing odds will certainly help you find the best odds for betting, but creating your own probabilities and comparing them to the odds available is what will help you find the right odds for betting in a given market.
How to bet on sports
Many people do not realize how difficult it is to win in betting, and you will not be able to find profitable bets until you start evaluating the matches yourself. However, you need to start somewhere, and this will certainly help improve your understanding of probability. Once you develop your knowledge, access better information, and experiment with different inputs and pricing methods, you can begin to find legitimate betting opportunities.
While we will provide an example of pricing for a football match in this article, it is important to note that you do not have to do the “hard work” of pricing in the betting market yourself. Some people prefer to trust the market using information provided by an effective bookmaker, such as Bookmaker Pinnacle, and look for discrepancies in their offerings with other bookmakers.
You may be using Pinnacle’s tennis odds and rescaling them to the odds of other bookmakers, then placing your bets. Or, you can use NBA team metrics to calculate the odds of one team overtaking another (and by how much) to calculate the odds of a money line or handicap bet. However, we can consider such examples (and their potential pitfalls) in other materials, and today we will talk specifically about the assessment of football matches.
Where to start?
The prospect of doing what the bookmaker does (judging the odds of an event) with incomparable resources is likely to be a daunting task for many players. Provided that you are willing to put the time into learning, and are willing to make mistakes and accept failures (because they will happen anyway), there are useful things to learn from pricing the markets yourself.
In recent years, the general idea of a successful sports player has changed. While traditional bettors who base their calculations on their own knowledge and experience still exist, it is now more about individuals or groups who build their own models with large datasets and complex algorithms.
If your goal is to reach the level of professional players that many strive to be like, you need to understand that you have to start somewhere. You don’t just run programs — Excel, R, or Python — load the data and wait until it gives you a signal to bet. Start with the basics, a simple approach with little data, and work your way up.
An example of how to rate a football match
To explain why the pricing of a soccer match is important when you want to place a bet, we used a simple example to show you how. It should be noted that this approach has many disadvantages (which we will talk about later) and, when used on its own, will not help you find value in the soccer betting markets.
We used the Poisson model to create the 1X2 odds of the English Premier League matches (for this example, the first round of the 2019/20 season was taken). How Poisson’s distribution can be used in betting is described in more detail in a separate article, but we’ll cover the basics here as well.
Using Infogol data on expected goals from the previous Premier League season (2018/19), we were able to calculate the “attack power” and “defense strength” of each team for both home and away games.
This gave us a relative measure of a team’s ability to score and concede goals using the ratio of the team average to the League average. Using expected goals instead of real ones will give a more accurate representation of the teams’ performances and to some extent eliminate the randomness and chances of ordinary luck that we see in a season of 38 games.
The formulas for determining these indicators are as follows:
Home Attack Strength = Team’s expected goals in one home game on average / Average expected goals of all League teams in one home game
Home Defense Strength = A team’s expected conceded goals in one home game on average / Average expected conceded goals of all League teams in one home game
Away Attack Strength = Team’s expected goals per away game on average / average expected goals of all League teams per away game
Away Defense Strength = Team’s expected conceded goals per away match on average / Average expected goals conceded by all League teams per away match
All these indicators are summarized in Table 1:
We can now use the home and away attack strengths to calculate how many goals the home team should score (and vice versa – using home defense and away attack strength – to calculate how many goals the away team can score).
This is how the process will look like in the match of the first game week of the 2018/19 Premier League season between Leicester City and Wolverhampton:
Leicester Goals = Leicester Home Attack Strength X Wolverhampton Away Defensive Strength X average expected league goals per home game
0.903 x 0.808 x 1.649 = 1.203
Wolverhampton Goals = Wolverhampton Away Attack Strength x Leicester Home Defense X Average expected League goals in a single away match
0.937 x 0.814 x 1.326 = 1.011
This tally gives us the number of goals each team would have to score if they played against each other (1.203 for Leicester and 1.011 for Wolves). However, the game cannot end with a score of 1.203-1.011, so we need to find the probability distribution in the range of outcomes.
We can use the Poisson function in Excel to calculate the probability distribution for the different number of goals each team can score in a match (we used a range from 0 to 5 for clarity). Using the example above, this is what this distribution would look like ( Table 2 ):
In order to calculate the probability of only a home win, a draw and an away win (1X2), we need to calculate the probability for each of the potential outcomes.
Leicester – Wolverhampton 0-0 = Probability of Leicester to score 0 goals x Probability of Wolverhampton to score 0 goals
0-0 = 0.3002 x 0.3639 = 0.1092 or 10.92%
We then repeat this for all possible outcomes where both teams can score from 0 to 5 goals (36 total – six draws, 15 home wins and 15 away wins). This is what the likely outcomes of this match will look like ( Table 3 ):
This gives us the following probabilities for each outcome:
Leicester win 40.23%
Draw – 28.84%
The victory of the “wolves” – 30.66%
We can then convert these percentages to odds and convert the bookmaker odds to percentages to compare them and try to determine the bets that have value. Table 4 below compares Pinnacle’s opening odds for the first round of the 2019/20 Premier League season and the odds derived from the following model of expected Poisson goals:
How to find an edge in the betting market?
If this were a real-world example that we experimented with, it would take us some time to evaluate how accurate these odds are compared to those provided by the bookmaker. It all looks logical and it would be nice to find such inconsistencies, but if the bookmaker is more accurate than you, you will not win in the long run.
The temptation may be to start betting money on outcomes that you consider to be valuable bets, but even with small amounts it can be too costly (we need a lot of bets to start making any meaningful observations). Thus, back testing is the most effective approach to see how viable the method is.
Comparing the odds this model would have given for past events and comparing them to the Pinnacle’s close line will help us see how good this pricing strategy is. However, it is not enough to simply say that it works or not. The most important thing is to understand why it works or why it doesn’t work.
There are many reasons why the Poisson Expected Goal Model used above is not a good way to evaluate a football match. Using data from the past season and not using rolling data means it will quickly become obsolete. Failure to account for transfers and appoint new coaches can distort the measure of a team’s strength and its chances of winning the match. These are just a few examples of what the model does not account for.
If, hypothetically speaking, we did find any advantage in this model, then it is important to understand where it came from. Is it just something that the bookmaker or other players haven’t taken into account? Does it depend on when you place your bet? Can you improve the quality of the data to increase your bid? If we have a legitimate advantage and we know how it is created, it is imperative that you manage your bankroll properly to maximize that advantage.
What do you do when you start winning?
This may come as a surprise to some, but the hard work is not over once you find a successful betting strategy. In fact, for many people, this hard work is really just beginning. Unfortunately, some bookmakers will restrict those who manage to make more accurate predictions for matches than the odds offered on them.