Meet David Sumpter, professor of applied mathematics at Uppsala University (Sweden) and author of the book "Football Mathematics: Mathematical Adventures in the Beautiful Game. Thought that mathematics should help to beat bookies, and decided to check it in several ways.
We couldn't miss it - here's what the professor found out during his research. It will hardly be a discovery for experienced gamblers, but it turns out a great guide for beginners in soccer betting.
Beginning. Choosing the bookies
David Sumpter has never been into betting and didn't particularly share the soccer hobbies of some of his buddies. He's been to the dog races a couple of times and has seen his friends bet on soccer a few times to add to the adrenaline of watching. But he considered personal involvement, much less personal credit card involvement, a clear overreach.
Everything began to change when the professor sat down for his book "Soccer Mathematics": before publishing it was necessary to test his research on the real market, so he had to personally find out whether mathematics can beat a bookie on a long distance. The English Premier League season 2015/2016 was chosen for the experiments. Bets were placed exclusively on the main outcomes (win/draw/loss).
To begin with - a tip for beginners: Professor Sumpter advises to play with several bookmakers at once, because each bookmaker has its margin, and it varies from match to match.
Usually the margin is 5%, i.e. only 95€ out of 100€ bets are involved in the game. That is, the average player, if he is not much smarter than the bookmaker, simply will not be able to beat him. But different studios give different odds and if you always choose the highest one, the chances increase by 2-3% on average.
Below we will tell you about the game models that the Swedish mathematician used in his attempts to beat bookmakers.
Model #1. Following the expert
NBC journalist Joe Prince-Wright was the top predictor for 2014/2015 APL season. David Sumpter decided to start betting on his advice. The venture failed rather quickly. The author of the book "Footballomatics" believes that media personalities make predictions only for the sake of advertising: they are interesting to read, but playing on them is not able to bring the player to the plus.
Model #2. Bets on European club rankings
The European club ranking includes all teams from the top soccer divisions in Europe. In its formation various statistical indicators of each club are taken into account: the results of recent matches, performance, performance in home and away games and so on. According to Sumpter, the data of this rating can really be useful in betting, but it should be borne in mind that they are used by the bookmakers, and therefore they do not provide a significant advantage over the player's bookie.
Model #3. Expectation of goals
One of the most interesting models of the Swedish professor. It is based on calculating the probability of goals from different positions: kicks from outside the box, from free kicks, from set-pieces, etc. For example, if a team scores 12% of its goals from outside the box, it gets 0.12 expected goals per game. There are many factors to consider, including the opponent's defence (using the same principle).
The start of the 2015/2016 season with this strategy yielded inconsistent results: Sumpter managed to predict the sensational fall of Chelsea, but the system overestimated the capabilities of Arsenal and Liverpool. This model, according to the mathematician, allowed not to lose money but did not expect serious profits from it.
Model number 4. Using the bias of odds on favorites
According to David Sumpter, the strategy of using a bias in the odds of favourites is the most profitable. Many betting lines are not designed to play the same favourite over the long term. That is, for each match odds on the favourite is calculated without reference to the previous odds and the game. According to the mathematician, this is a mistake of bookmakers. Based on the 2014/2015 season, regular bets on wins of Chelsea, Manchester City and Arsenal in each round gave their small, but profitable. Surprisingly, the fact is that the top clubs win more often than bookmakers predict.
However, the strategy of biased odds on favourites failed in the 2015/2016 season because of the phenomenal Leicester, the unstoppable West Ham and several other interesting clubs in the English Premier League, which regularly took points away from the top teams.
Model #5. Betting on draws
Since the previous method did not give Sumpter the expected profits for the 2015/2016 season, the professor decided to improve it, so as not to depend on the capricious Arsenal strikers or Manchester City defence failures. Sumpter suggests betting on draws in top matches.
Mathematician found that English players do not like to bet on draws during matches of top teams. Meanwhile, the clubs play about the same level, and the factor of own field in modern soccer plays less and less role. Since the players bet on the victory of one of the teams, the odds on a draw with bookmakers increases, and it gives more profitability betting. Very many top APL matches of the 2015/2016 season ended in a draw, and the author claims to have been in profit by 27% at the end of the season (from the starting pot of £400 received a net profit of £108.33). Toward the end of the season, Sumpter could have made as much as 200% net profit on this model but abandoned betting.
According to Sumpter, no one bets on draws in top matches also because of the media: journalists very vividly oppose one team to another, because... well, who is interested in predicting a draw?