Push Notification

Passes per 100 passes: statistical potential

In the analytical world of football, everyone aims to create reliable statistics. We have come a long way from the number of corners appearing in the match broadcasts to the player tracking data and I believe we are still at the beginning of the data-based football analysis. The great thing about this new wave is that anyone can come up with a good idea as there are loose boundaries in the data analysis in football. 

One important step in the fast-paced evolution of analysis was the ‘per 90’ statistics. This way, player stats were standardized and easily comparable whether he played 10 or 30 games in a season. The disadvantage of this method is that these metrics are biased based on the playing style of the player’s team. If a team has an average possession of 40% during the season, chances are the whole numbers of passing opportunities would decrease. This results in losing sight of some great performances and this would be the topic of my article. 

This is just a simple example but the main idea would be that event statistics should be all based on possessions. Already there are some defensive metrics that provide less biased data such as interceptions or tackles but we should create even more out of this opportunity. The Premier League contains the best playmakers in the world. I highlighted ten of them to see how their progressive passing rate changes when I use a possession-based calculation instead of the usual per 90 stat. 

The change in the ranking is visible and all in all, the second chart accounts more for the player’s individual ability rather than what he was asked from the club. Gylfi Sigurdsson improved the most as we can see that even though he hasn’t been that creative outlet in Everton, he still has the ability to progress the ball.