Thursday, 31 March 2011

Da’s logisch (That’s logical )

Last Tuesday the Dutch national soccer team played against Hungary. As with any match of the national team, it was on television, encapsulated in a programme in which various self pronounced soccer experts discussed and analysed the match and accomplishments of the team. It is fun to listen to the soccer gibberish used by these experts in their comments on the actions of the coach and team performance. They are always able to tell what went wrong and what would have been a better option (as could any other inhabitant of the Netherlands). A famous expert in soccer gibberish is soccer legend Johan Cruijff. He is famous for quotes like “Every disadvantage has its advantage”, “If they have the ball, you can’t win” or “That’s logical". So, being a coach is tough, especially when it concerns the nation soccer team, because the complete nation is watching you. Since much of the decisions in soccer are made through intuition and common sense it might be a good idea to put some numbers to beliefs/convictions and introduce some rigour in the decision making.

The decisions a coach needs to make are complex. Think of scouting for a new team member to fill in an open position and improve the overall quality of the team. Or deciding on the overall composition of the team selection; which players, how many for each position, etc. A decision a coach needs to make often is the line-up for the next match. A simple calculation shows that there are 308.403.583.488.000 possible combinations for the line-up for a soccer game, given 26 players in the team selection. How will a coach be ever able to identify the best possible line up?

Given that a player performance will differ depending on his assignment to a certain position in the field (a goalie is not much good in a forward position), a certain attractiveness score good be given to such an assignment. Identifying the best possible line-up can than be seen as finding the assignment of players to field positions that maximises the total score. Because every position in the field must be filled in and a player can have at most one position in the field the so called maximum-weight bipartite matching problem comes forward. Introducing a source (supply side 11) and sink (demand side 11) and giving the arcs from the source and towards the sink a maximum capacity of 1, this bipartite matching problem can be solved as a transportation problem. It could be easily implemented as an App on the coach’s iPad. The tricky part in this approach is the scoring of the assignment of players to positions. This is where sports analytics come in.

By keeping track of the player performance during matches, the effectiveness of the player can be measured. This score could than be used in the optimisation of the next match line-up. By measuring the number of intercepts, the number of safes, the number of passes, etc all kinds of statistics can be calculated. Using these statistics an overall efficiency score can be calculated, see for example the Match index of SoccerLab. Not only after the match, but in real time as well. The score gives an indication of the players’ ability and effectiveness of the assignment to a position in the field. Companies like ORTEC TSS provide the kind of statistics to calculate this kind of efficiency scores. So instead of listening to the soccer gibberish of the experts, the facts can be used and the actual performance evaluated. The coach could use the statistics together with the line-up model, during the match, to decide on which player the exchange (a much discussed subject in the Netherlands; Bosvelt for Robben EC 2004) and decide on the line up for the next match. With an analytical approach and structural analysis of player performance, deciding on the line-up becomes more fact based leading to more sustainable team results. It doesn’t however offer a guarantee on winning championships, but that’s logical.
Post a Comment