How do Trends Help to Win in Betting?

Sometimes a mathematician can leave beautiful ivory towers and test his skills on the stock exchange, in a casino or at sports betting.

It also happens that he manages to earn, and he is ready to share the secrets of his not entirely fortuitous good luck with others. For us, of course, such stories are of interest not only from pure curiosity. And here’s one of them.

Case study

Professor Uppsala University David Sumpter decided to test his knowledge of statistics and luck in football betting, and at the same time earn extra money. With £400 on hand, the professor earned a net £108.33 for two months, which is 27%. This is not a jackpot, but I can’t imagine how you can increase this amount better, staying within the law.

According to the professor, bookmakers usually pocket 5% (office margin), which corresponds to a coefficient of 1. 90 for two equally probable outcomes (for example, on the outcome of a tennis match).

That is, if you simultaneously bet $50 for each of these two outcomes, then after the game on your hands you will have 95 instead of the original $100 with any result, and $5 will settle at the BS. So pretty soon you will be left without money. However, thanks to the competition between bookmakers and the existence of low-margin bookmakers, the odds on individual sports events can have a margin and 1.5%, and on top events like the World Cup finals, football – and even lower. In this case, you need to beat the bookmakers at least 2% in order to stay in profit at the sports betting. This was the aim of the Swedish professor in his field studies.

The mathematician used various statistical models, of which, in practice, he used four, and only one helped to make money from them. But first, about the models of the professor, who did not live up to expectations and did not bring profit.

Model winnings

Model #1. There is an index of European football clubs, which is called – Euro Club Index. The index grows when the team wins, and decreases when the team loses. Details of the calculations of the professor in this model are unknown, but there is reason to believe that this is a kind of logistic regression equation Bradley-Terry.

Forecasts based on the Euro Club Index in the professor as a whole were not bad, but it was not possible to beat the bookmakers’ margin – he suffered a permanent loss.

Another tactic that did not justify itself was based on expert opinions – in this case on the forecasts of NBC journalist Joe Prince-Wright, who had previously successfully predicted the position of the Premier League clubs in the final standings of the championship. The result of the game on this model did not satisfy the professor. The mathematician soon changed tactics, concurrently drawing a conclusion: the experts’ forecasts are not bad as entertainment, but usually they do not help to beat the bookmakers.

The third model, more successful than the first two, used the Poisson distribution, with which we recently made a successful forecast for the match Liverpool – Arsenal. Each strike on the goal was given a certain value, which was determined by the historical statistics of the blows made in the same position (for example, inside or outside the penalty area). The author of this method helped at the beginning of the season to predict the failure of Chelsea, but led to a reassessment of the chances of Arsenal and Liverpool. He did not lose money, but the forecasts came out so boldly that the professor decided that this model is unlikely to be useful at a distance.

The fourth method, which was formed by the professor by the middle of the season of the Premier League, was successful – the identification of long-term tendentious expectations. The scientist of Uppsala University noticed in different years a clearly expressed tendency to underestimate the teams-favorites and major clubs.

This was, for example, the season 2014/2015, when the bets on the victory of Arsenal, Chelsea and Manchester City over teams with a lower position in the standings would bring a small income. That is, England’s top clubs have won slightly more often from outsiders than what the coefficients said. The professor gave this explanation: players in pursuit of large profits neglected a small gain, reluctantly betting on strong teams, which influenced the coefficients. However, in the season of the Premier League of 2015/2016 there was an adjustment, and this trend was reversed – the major clubs- favorites were revalued in the coefficients, which were crowned by the championship of “Leicester” at a factor of up to 5001.0.

At the same time, another long-term trend was discovered by the scientist, which proved to be the most enduring: the likelihood of a draw in the top matches is underestimated. There are reasons for this. The media and social networks foment passions and irreconcilability before the game, the players themselves do not like to bet on a tie. Famous English football hooligans also do not add to the peace of mind in the Premier League. It is the long-term trends, including the underestimated odds of a draw, that allowed the professor to win at bets often enough to prove the existence of regularity.

It is useful to remember this when making bets on the football battles of the Premier League. About this and many other things the professor of mathematics David Sampter wrote in his book Soccermatics. It can’t be downloaded for free, but I think it’s worth your money if you know English.

And finally, the secret fifth forecast model, which the professor used, and it’s called – “ask your wife”. However, as it turned out, the professor’s wife from Sweden also teaches mathematics and knows how to make bets. In any case, within a month she received £17 profit with an initial amount of £100. Perhaps this is the secret ingredient of a successful forecast strategy: combine statistical methods and use predictions from experts.

The bookmaker coefficient and the probability: how to translate one into another – and back

Beginners of sports betting choose an event that arouses their sporting interest, and they decide to take a chance. In this case, such players usually do not understand the relationship between bookmaker odds and probability. Let’s correct this misunderstanding and learn how to translate the BS odds into chances and back.

What is the bookmaker coefficient?

In a general sense, a coefficient is a numerical expression of the probability of an event. For example, the victory of one of the teams in the match. It can be called a proposal, which the bookmaker does to the bettor. The office evaluates the odds, lays out its margin and offers you a bet.

Before we see these numbers in the line, there is a difficult process of determining the coefficients. The modern bookmaker counts them automatically using mathematical models. They take into account a huge array of statistics. However, the role of a trader who perfectly knows his sport or individual tournaments is still important. Especially in niche leagues.

The task of the coefficient of the bookmaker office

The dependence of the coefficient and probability is inversely proportional: the higher the coefficient, the less likely, and vice versa. At the same time, the lower the probability of an event and the higher the odds, the greater the potential winnings that the bookmaker promises to you at a bet.

Here we come to the problem of the bookmaker office coefficient. It’s about telling the player how much he can win.

How to find out potential winnings? To do this, multiply the bet amount by the coefficient. For example, the bet on 1000 dollars on factor 2,0: 1000 dollars x 2,0 = 2000 dollars. This amount is the total potential payout. Net profit you get if you take away the payment of the amount of the bet: 2000 dollars – 1000 dollars = 1000 dollars. Well, and if your bet wins.

It is worth the odds to exceed the psychological mark of 2. 0, most players light up their eyes and turn off caution. Therefore, it would be nice to use simple manipulation to determine the probability of the event, calculated by the bookmaker, and compare it with your own estimate. Perhaps, you will take a fresh look at the chosen bet.

You should bet when you evaluate the chances of an individual event higher than the bookmaker.

How to calculate the coefficient BS in the probability and back?

Extremely easy. To find the probability of an event – an approximate one, because it will have a margin – it is enough to divide the unit by a factor and multiply by 100. We get the percentage probability value. Example with a factor of 2.0:1/2.0*100=0.5*100=50%

Now we will understand how to convert probability into a coefficient.

This is also easy. It is necessary to divide 100 by the probability value. Example with a probability of 75%:
100/75=1.33.

That’s all the wisdom. In other materials we will consider what are the formats of bookmaker coefficients and how to translate them into the usual decimal.