Scientists at the Alan Turing Institute in the UK have developed an AI algorithm to predict which team is most likely to win the Qatar Soccer World Cup. It mainly takes into account the results of previous championships. However, factors as important as the performance of the individual footballers. Therefore, it would be interesting if it could be complemented with another algorithm, developed in 2020 by data scientist Carter Bouley. In it, he analyzes the different types of passes that players can make, to calculate what would be the best strategies.
It should be noted that this algorithm was not carried out for the Qatar Soccer World Cup, nor for any specific championship. It’s just a way to use the statistics and the computer Sciences to develop the best game strategies for soccer players.
This is mostly based on passing, so it could be very useful for teams like the Spanish selection, whose passing game is often among the keys to his successes. The ideal is to optimize them, so that they involve the minimum expenditure of energy and, in addition, they can be linked accurately until they end up in the goal. There are no magic formulas; but, at least, with this algorithm one can help to Design the best possible strategies.
The algorithm for the perfect passes, even beyond Qatar
Logically, not all soccer players are the same. There are more or less skilled and more or less trained. However, optimizing their passing strategies can help all of them.
That is the goal of this algorithm, which was trained using data from 358,753 passes, made in 380 games, in which 20 teams participated. Several factors were taken into account. To begin with, if the players were in their own field or in that of the opposing team. On the other hand, the results minute by minute and with the complete game. In addition, the passes were drawn graphically, where the ends of the field would be the X and Y axes. Finally, the type of pass was taken into account: normal, header, cross, corner, launch, goal kick or free kick.
With all these data, the artificial intelligence to look for patterns that associated a specific type of pass from the players with the best results. discovered data as a large proportion of passes are missed at a very short distance, below 5m. Furthermore, between 15 and 30m, “many more passes are completed than missed, and after 30m the proportion of completed passes drops dramatically, while missed passes start to level off.”
Another key factor turned out to be the place on the field where the pass is made. For example, the closer they get to the opponent’s goal, the more missed passes are made. Logically, it is a very important area, so it is important to know which strategies work best right there.
Special attention to footballers
This algorithm takes into account the individual role of footballers. As Bouley himself explained at the time, “if the model predicts that a pass will occur with a probability of 0.8 and it is completed, 0.2 is added to the footballers’ pass score”. On the other hand, “if the pass was not completed, minus 0.8 for the players to pass the score”. This is then averaged over how many passes the player makes, to define a average risk score exceeded. “This score makes it possible to compare footballers through the risk assumed and overcome in the pass.”
Because, logically, it’s not just about knowing which are the best passes. Good footballers are also needed to carry them out. This also means that they are able to take risks, but without being too daring. In the medium term is virtue. Also to win a football match. It doesn’t matter if you’re in the Qatar Soccer World Cup or in a neighborhood championship.