Measuring Freedom

What exactly is a ‘free role’ ? Why does Kevin De Bruyne have more freedom? Can we put a number behind it?

Swaminathan Nanda-Kishore
4 min readDec 29, 2020

‘The gaffer has entrusted him with more freedom today’ , ‘he has been given the creative license, let’s see what he does’. Freedom in football is common parlance. It’s usually associated with creative midfielders who move into channels to receive passes. Still despite being such an easily understandable concept, there isn’t exactly a way to definitively say that ‘player X’ has more freedom than ‘Player Y’.

I propose a new Metric, called FR (Freedom Rating) that puts a mathematical number behind how free a player’s role truly is.

Ok maybe the guys over at statsbomb already have something like this, and this article isn’t exactly a game-changer that I hope is. Having said that, nobody I know openly talks about such a metric, and I hope that my simple model can spark more research into this field.

The metric that I chose to start with is passes received. Obviously a more free player would move into more channels and throughout the course of the season, would receive passes all over the pitch. Kevin De Bruyne, a midfielder is a lot more free than Wan Bissaka, a right-back who cannot move around the pitch to much due to defensive commitments. The following visualisations help illustrate.

Kevin De Bruyne’s heat-map for passes received in the 19/20 season
Aaron Wan Bissaka’s heat-map for passes received in the 19/20 season

Now, I simply took the standard deviation of passes received, first with the length and then with the width. I assigned both weights, the length received a weight of 0.8 and the width received a weight of 1.2. I added the two and divided by 6, since that normalised almost all of the values to a scale of 10.

So here’s the first look at FR for the 25 players who received the most passes in the 2019–20 season.

Obviously the limitation of a season-wide FR calculation is that it tends to favour those who play multiple different positions, for example — Cesar Azpilicueta, who filled in both as a left back and a right back for Chelsea last season.

A closer look at his season heat-map confirms our hypothesis

So instead of taking the standard deviation for the season as a whole, I took the average of the standard deviations and called the new metric FR_A or Freedom Rating Adjusted. I made sure to keep a minimum touch requirement per match of 12.

And here we go , here’s the FRa ratings for the top 20 player, KDB comes out as the player with the highest FRa among all players with atleast 1400 passes.

And here are the most ‘free’ players who have received at-least 500 passes in the 19–20 season. Bruno Fernandes shows up in the top-20 and Moritz Leitner is the best in the league, (Ozil in the top 5 btw).

I’ve also realised that this metric does tend to over-value forwards thanks to the weights and the relatively low number of spread out touches they have per match. Still it’s a very useful tool for comparing player positionally. At the same time, models in the future could use a higher pass cutoff than 12 to fix this

Thought I’d finish this piece with a Fr_A comparison for Pogba and Bruno last season for you to draw your own insights.

After Bruno joined, in every match, he was given more freedom than Pogba. Still, Pogba’s FR_A numbers just barely reduced, showing that they Bruno’s freedom didn’t necessarily take away Pogba’s freedom

Thanks for reading ! You can interact with me on my Twitter accounts : nandy_sd and NdyStats

--

--