The obsession of betters and some researchers – to beat the bookmaker in English sounds somehow belligerent: “beat the bookie”, since the first word means not to beat, but to beat. More than once I had to deal with articles compiled by obscure authors, scientists and researchers who set themselves a similar goal. One of these attempts was somehow mentioned in our previous article. It dealt with a strategy taken from the book “Soccermatics”. The author’s tendency is based on a long-term bet. It is more likely that the author managed to take this step, but the main mechanism for identifying and subsequently exploiting long-term trends has not been disclosed.
The rest of the authors decided to use several complex models: collecting social media keys, technological learning, and more. In the experiment, examples were given that allegedly could bring a stable income of 70% if profitable methods were used. Most likely, this is how it is presented in the framework of experiments. But the only problem is that this experiment and a long experience in participating in casino online quebec are completely different things. This is another important circumstance. It is unrealistic for an inexperienced user to check all the calculations on their own. Before you can collect 100,000 keywords from Twitter before the start of the match, you need to calculate them in detail.
And this is what happens, where it is practically not necessary to wait for profit. Three Japanese scientists decided to do the impossible, they conducted a study and got a positive result when participating in sports betting. Soon, all the data obtained during the work were posted for public use. The model turned out to be simple and did not require 150 knots and 25 scientific hands. They were able to get a decent win, which is why bookmakers began to block them. I thought it was time to stop at this experience.
Description of the main activities
The researchers did not work on a new version of the calculation and make predictions on the likely winnings in football matches. The essence of this strategy is that the personal calculations of bookmakers are taken as the basis so that they can surpass the performance of other users. From the collection of important information on certain rates, it is usually possible to isolate the probability to identify rare rates that may have significant deviations from a certain value. The calculations were carried out for ten years, they were reinforced by 5 months of participation in sports betting for real money in actually operating offices.
Important Login Information
The authors who conducted the study decided to collect 479,440 football matches, 818 from all this League, collected from different countries of the world. The betting information was collected from the databases of the bookmakers themselves. The duration was taken into account from 01.2005 to 06.2015.
Methodology
At first, scientists tried to reveal the main information, for which they decided to analyze 80 teams with a probability of 0.0125. A matched sports bet is placed on 32 bookmakers, has an inversely proportional probability, compiled according to the formula.
Pmean=1/mean(Ω)
For any of the 80 teams, we decided to single out only 2 indicators:
Agreed value on average tactics.
Accurate forecast for the outcome of a certain football match, which has a true ratio between teams on their own field, with draws and victories on a foreign playing field.
Now let’s take a look at some strategies that can lead to winning bets. In certain situations, bookmakers try to provide bettors with promotions for certain actions. Let’s say that in a situation where there is a favorite in one match, then everyone is usually in a hurry to bet on him. Bookmakers in this case can increase the payout per outsider in order to attract even more bets and reduce the payout as a result. Sometimes it happens that rates can drop below market value.
Pcons=Preal-α:
- Preal is the probability of the result of the match;
- α is the correction taken into account, its size is usually 0.034, 0.057, 0.037 for the following positions: 1, X and 2.
Technical specifications
Any data, and this is code and documents, are on GitHub, which is the main portal https://black-mantra.com/ of the operator and programs with open source code. Information, in particular, is duplicated in several convenient repositories, and a link to them can lead to a drive in Google. The files usually weigh less than 2 GB, which is quite acceptable to install on your own computer. Any information will be saved automatically, but in case of unforeseen situations, you can contact a lawyer.
All the necessary tools for personal work are offered as free open source programs:
- MySQL – database of important information;
- PHP is the main programming language;
- Octave is a math package.
What is the result?
Any assumption related to the fact that the bookmakers are well aware of their own business, as a result, was justified. A linear relationship was found between the betting probability of estimates and the result in the match.
- R2= 0.999 – correlation indicator for operators’ winnings;
- R2= 0.995 – correlation indicator for a draw;
- R2= 0.998 – correlation indicator for getting the winning result of the guests.
At first, the results of the run-in were carried out on a simulator program, where final rates were used for certain bookmakers. Accurate predictions resulted in about 44.4 percent, and the norm for all this is a virtual profit of 3.5 percent or $98,865 for 56 attempts with $50 for a single game. In comparison, the forecast at random gives only 38.9 percent with an accurate hit on the target.
As a result, inspired by their own success, the better decided to use not final, but intermediate rates, rightly believing that it would not always be possible to apply final investments. Then bets are used with a distance of 1-5 hours and the initial stage of the game. The result in this case will be encouraging.
Another important stage of the online game on the exchange is taking place, and as soon as it ends, a completely different game begins, already for real money. The researchers, acting on their own principles, decided to bet $50 each through a virtual bookmaker’s office, where the program automatically found the most profitable football match options, using certain criteria related to the selection of a particular option.
For five months they also placed bets to compare the results according to the table on the site. Random bets were taken into account at the time of a fictitious game on the stock exchange for the period 05.2015-07.2015.
Strategy Income Accurate Prediction Win Total Number of Games Bet
Connoisseurs 8.5% 47% 957.50 USD 265 50 USD
Randomness -0.7% 38.7% -670 USD 672 50 USD
The result turned out to be impressive, but the probability that this strategy in terms of mathematical calculations does not differ at all from random ones can affect several times. According to scientific indicators, it looks something like this: p = 1//11 ≈ 0.089. In addition, in certain academic circles, with the de facto standard, the main value is p = 0.05.
Although pi is a little high in this case, the bookmakers were still seriously alarmed, they decided to block all the accounts of scientific users, which significantly complicated their lives, since the selection of tournaments has shrunk, as you can see, even a small selection can lead to a slight error.
Conclusion
It is most likely that this study will result in a significant event in sports betting. A simple tactic, this is a conscientious work with the disclosure of accurate data and miscalculations, where there are no black boxes, attempts to cast a shadow on the player. In addition, any negative reaction from the bookmakers means only one thing – the researchers hit the target. Most likely, after a certain time there will be an answer to any question.