Monitor the development and current competitiveness of all players of a team by extrapolating individual performance from team performance with state-of-the-art statistical methods.
Minimize risk in player recruitment with access to our database of cross-comparable ratings for 100'000+ active players in 350+ competitions worldwide. Find undervalued talents that fly under the radar of the transfer market.
Predict the future development of player performances for whole professional careers. Forecast individual games and end of season results with predictions that regularly outpferform betting markets.
The original idea for Footballgenius took root after endless discussions about whose favorite football player truly is the best. The respsective cases were usually built by comparing more or less elaborate statistics of the respective players, but none of that seemed to lead to a satistfying conclusion, as there always remained unanswered questions ("The striker scores 30 goals per season but who knows if a different striker in the same squad wouldn't score even more?" or "The midfielder has a passing accuracy of 95%, but how do we know that the passes actually contribute to their team's success?").
Being both football fans and mathematicians, we set ourselves the goal of developing a rating model which tries to capture the essence of what makes a player good or bad: Does the player being on the pitch increase the team's probability of winning or not?
State-of-the-art statistical methods allow us to decouple a singular player's performance from the performance of the whole squad and we can thus better judge the impact of a player on their team's success. Letting the algorithm run over our database consisting of more than one million football games and over 100'000 active players provides us with a meaningful performance index for each of those players: the Footballgenius Player Rating (FPR).
The FPR makes it possible to compare performances of players across different clubs, countries, divisions and age groups. Furthermore, it also provides us with predictions for the future career development of young players.
FPR recognizes current and future key players early and reliably and can thus provide football clubs with crucial assistance in their personnel decisions.
In contrast to other rating models for football players, the FPR does not measure any single action on the pitch. Instead it evaluates a player's performance by the results of all matches in which they took part, while also considering the strength of their respective teammates and opposition.
The advantage of this model is that no assumptions have to be made about which action on the pitch will influence the result in which way, as these assumptions are necessarily prone to bias. In fact, this method is precise enough that the outcome of future fixtures can be predicted with an accuracy rivaling that of the betting market, by just considering the FPRs of the players on both teams line-ups. Our calculated predictions for the end-of-season results of more than 200 football leagues worldwide may be found here.
In addition, the FPR lends itself to being extrapolated to the future; after the first ~50 games in which a player participated (which usually corresponds to about 2-3 seasons on youth level) the calculated FPR is significant enough to make reliable predictions about the further performance development in that player's career.