**After testing it on paper, the researchers began to gamble real money, and it turned out that the method worked. But as soon as the runners detected their success rate, they began to block their bets. Now the team is asking governments to regularize the system**

If you’ve ever succumbed to a hunch, you know perfectly well that bookmakers and casinos manipulate the odds to turn them against you. The clearest example is roulette, where there are 36 red and black numbers and two green numbers, one for 0 and one for 00. So there are 38 possibilities in all. When betting on red or black, the odds of choosing correctly are 18/38, so a reasonable reward for a bet of one euro would be 2,111 euros. However, the bank only pays two euros and keeps the difference. In this way, a profit is guaranteed.

A similar bias occurs in the quotas offered by bookmakers in horse racing, football and any other sporting event. Bookmakers always make sure the odds are in your favor. But establishing these odds is more difficult than roulette because the calculations are more complicated.

And that raises a tempting possibility. Would it be possible to find a better way to calculate the odds and beat the bookmakers?

Today we have an answer thanks to the work of University of Tokyo (Japan) researcher Lisandro Kaunitz and some colleagues. The team has discovered a way to earn money on a regular basis in the online betting market for football.

But your work comes with a serious warning. Researchers say that as soon as the bookmakers realized their success, they prevented them from continuing to gamble.

Players have always flirted with stratagems to beat the odds, but rarely successfully. This is because bookmakers work hard to calculate the exact odds. They usually use statistical teams to study historical data from a sport such as football and then develop sophisticated models to determine the appropriate probabilities for each match.

The team explains that, according to their data, no one has been able to beat this system by developing superior statistical models. But despite their sophisticated approach, there is a weak spot in bookmakers’ methodology regarding how they cover their bets to protect themselves against the possibility of big rewards.

For example, when two teams play a football match, the bookmakers establish the odds that each team wins, loses or draws. Sometimes a large number of people can bet on a particular outcome for reasons unrelated to probabilities: that team might be more popular than expected, for example. In that case, the bookmaker is at risk of having to face a large payment if that result occurs.

To minimize that risk, bookmakers can cover their bets by offering more favorable odds on the opposite outcome. In this way they attract bets that cover at least some of the possible losses.

Kaunitz and co-authors assert that this process also creates an opportunity for anyone who can detect it. The trick that researchers have perfected is to devise a method that detects when odds favor the gambler over the bookmaker.

Its method is simple. They begin by assuming that the bookmakers themselves are good at establishing odds and that the prizes they offer are an accurate reflection of the real odds of winning, drawing or losing, plus their own margin.

In that case, a good measure of these odds is a simple average of the odds offered by all the bookmakers, a kind of crowd wisdom. This gives the average odds, which are a remarkably accurate reflection of the real odds, according to the researchers.

Then, it’s easy to analyze all the odds offered and find out the outliers. Afterwards, Kaunitz and his colleagues calculate how favorable the atypical odds are. If they are good enough, then the bet should pay off, at least in the long run.

And that’s exactly what they’ve done. They created a web tracker that gathered the odds offered by online betting companies at football matches around the world. They calculated the average odds, found any outliers, and then calculated whether a bet would favor them or not.

Before betting real money, researchers tested the idea of a 10-year history of closing probabilities and results from 479,440 football games played between 2005 and 2015. This simulation generated payments 44% of the time and yielded a return of 3.5% over the 10-year period. «For an imaginary amount of €50 per bet, the winnings generated would be €98,865 in a total of 56,435 bets,»they say.

An important question is whether this result could have been pure coincidence. Could they have just been lucky? The team then compared their results with 2,000 simulations in which they placed random bets on the same matches. In this case, the bets generated profits 39% of the time with a return of -3.2%, which is equivalent to a loss of 93,000 euros. That allowed the team to calculate the probability that their first result would be a stroke of luck. The research details:»The probability of obtaining a return greater than or equal to 98,865 dollars euros in 56,435 bets using a random betting strategy is less than one in 1,000 million».

So Kaunitz and his colleagues began to trust that their method would work in the real world, but there was a problem. Normally, gamblers may not always be able to bet on the odds at close, which can vary significantly during the pre-match period. So the team decided to create a simulation of this situation. The article describes:»We decided to carry out a more realistic simulation in which we place the bets on available odds one to five hours before the start of each match.

The way odds vary in the pre-game period is not publicly available, so the team created a bot that compiled these odds from betting websites around the world from September 2015 until the end of February 2016. They then tested their approach on all of this information.

The results were even better. Their bets paid 47.6% of the time with a yield of 9.9%. The team celebrates:»If every bet made was 50 euros, our strategy would have generated 34,932 euros in winnings in a total of 6,994 bets.

Interestingly, a random betting strategy with the same data yielded a return of 0.2% and a gain of 825 euros. This could be the result of intense competition between online betting companies, which sometimes offer more favorable odds to attract gamblers in a friendly loss-making leader policy.

The team then tested the approach using a strategy known as «paper trading,»in which they made dummy bets using real-time data rather than historical data. This is important because it allows them to verify whether odds quoted with an online bookmaker are actually available.

And what they discovered was that these odds changed 30% of the time when they tried to check online. In those cases, they ruled out the bet. But still, the strategy remained profitable. After three months of paper trading, their bets earned a 5.5% profit, with a revenue of 1,128.50 euros for 407 bets of 50 euros. «At this moment we decided to make real money bets,»Kaunitz and company say.

So they repeated their approach for five months, using the same procedure, only that a human trader would make a real online bet of 50 euros after verifying the odds. During that period, their bets were paid 47.2% of the time and they made a profit of 957.50 euros in 265 bets. That’s an impressive 8.5% yield.

Advised readers will have noticed that the number of bets they made was significantly lower than during the period of the wagering on paper. The researchers explain:»The reason is that we didn’t have a specific operator who bet on all the available opportunities 24 hours a day and, as a result, we missed many of the bets that appeared. But the smaller number of bets didn’t matter. In his opinion, his’ betting activity on paper and real confirms the profitability of the strategy’.