FBM data analysis is completely different from all other data analysis. We do our own data acquisition, so we are not reusing data from sources like OPTA. FBM data analysis works through a Bayesian statistical model of football players. We only look at actual results on the pitch and then figure out which players attributed to that result.
Results on the pitch can be anything but we look especially at:
Getting the ball significantly closer to the opponent’s goal
Recovering the ball
Goals against the team
Allowing the ball to significantly get closer to the team’s goal
Losing control of the ball
Gaining space without the ball
Red and yellow cards
Missing a chance to score a goal
As you can see some results are negative and other results are positive. Negative results are translated in a lowering of the efficiency of all player involved in that result. Positive results are translated in an increase of the efficiency of all players involved in that result.