The word “value” means something like value, benefit or added value. The bets are intended to generate a long-term advantage at 22Bet. An example: The following odds are given for a tennis match:
Kerber to win 3.00
Victory Sharapova 1.45
The bookmakers set the following probabilities (formula: 1/odds*100):
Victory Kerber: 33%
Victory Sharapova: 69%
The over 2% is the bookmaker’s profit margin.
However, if you know that Sharapova is weaker against left-handers, you may come to a different conclusion:
Kerber win: 40%
Victory Sharapova: 60%
In this case, you would have found a value bet, as a bet on Kerber is subjectively more likely to win than the odds imply. The bet is lost at 60% – for example with a stake of €10 – but a profit remains in the long term, as the expected payout amount is (chance of winning * stake * odds) 0.40 * €10 * €3.00 = €12. Mathematically speaking, this stake increases the bankroll by €2 in the long term (€12 expected profit minus stake of €10 = €2).
What are correlations?
In mathematical terms, correlations indicate relationships between two variables. This relationship can vary in strength and is expressed with values between -1 and 1. There are therefore both positive and negative correlations. Positive correlations indicate “the more, the more” relationships, which can be seen in the example of shots on goal (below). The more shots on goal a team takes in a match, the more goals it will end up scoring (in the long term). Negative correlations are again “the more, the less” relationships, which can be seen in the second example of the number of injuries or the number of missed passes.
The more misplaced passes a team plays, the fewer scoring chances it will create and the fewer goals it will score. The total number of variables must also be considered. If 100 passes are missed out of 400 passes, the chance of scoring a goal is of course greater than if 40 passes are missed out of 100 passes. The size of the sample also plays a decisive role in the assessment of correlations. The following applies to both positive and negative correlations: the closer the value is to 1, the more the two variables correlate. Accordingly, a correlation of -0.8 is very strong, while a value of -0.05 indicates almost no correlation.
How can correlations be calculated?
This is, of course, the task of the computer. The first step is to list the data for the variables in a table. The CORREL() function is then used in a separate cell. The data for one variable must first be entered there. This is followed by a “ ; ” to separate the two variables. Now insert the cells with the dots.
Excel or other programs will immediately spit out the appropriate correlation value. Of course, it is important to ensure that the value pairs are next to each other so that an exact assignment is possible. It is easy to determine these values. The only task now is to search for and collect suitable data in order to take into account every possible factor for a game event.
If you want to visualize the correlation graphically, all you have to do is create an XY diagram with the value pairs.