What I have noticed from profiling European teams’ style of play

Plus a tool to track how they have evolved in the last 4 seasons

The value of free, easily accessible football data sources like Statsbomb’s on FBref is that you can quickly find yourself spending weekends just digging into the weeds of it.

Well. That’s what I did.

Back in June, I looked at pre-tournament performance stats to profile the 24 national teams that qualified for the Euros.

With the new season about to kick off, I wanted to replicate and expand the analysis to the top 5 European leagues.
The idea here was to have a more robust set of data to better define styles of play and effectiveness in and out of possession.

I expanded the number of ratings (from 8 to 14), took some liberties (you’ll notice them), and spent some time looking at the data to see what interesting insights could come out of it.

How did I put this together?

The starting point is the data. FBref’s catalog goes back to 17/18 for the more advanced stats. I took that as my first season and 20/21 as the last one.

I didn’t want to have overall ranks for each team as in 4 seasons too many things can and did change. Instead, I defined the overall team ratings by ranking their performance for each stat in each of the individual seasons.

A percentile rank of 100 indicates that the team has outperformed all the other teams in each individual season. Ranks are first calculated at stat level and then aggregated for each rating (some of them are based on 2 or more stats, some others only on one).

For possession and attacking stats, high ranks indicate a strong performance. In the case of defensive stats like non-penalty xG against or npxG per shot against, percentile ranks are reversed to keep the same high rank-strong performance link (teams with strong defensive solidity and low-quality chances conceded had high percentile ranks).

I cherry-picked the Statsbomb data on FBref and grouped them into 3 macro-categories, 13 ratings, and a bonus one (you’ll find it at the end of the article).

Style in possession:

• Possession dominance
• Territorial dominance
• Progression via carries
• Direct play

Defensive approach and effectiveness:

• Defensive solidity
• High press
• Intensity of pressing
• Quality of the chances conceded

Attacking solutions and threats:

• Attacking threat
• Crossing game
• Reliance on dead balls
• Goal efficiency
• Quality of chances created

Style in possession

Possession dominance

Stat used
Team’s share of possession
Question the rating is meant to answer
Did the team tend to have the control of possession in that season?
What caught my attention
Atléti is now more than just a defensive fortitude.

This doesn’t come as a surprise but I’m glad to see it popping up in the data as well. The shift came in the 20/21 season as Simeone managed to provide more options to a team that is now more comfortable in controlling the game through longer spells of possession.

Territorial dominance

Stat used
Field tilt (team’s completed passes in the final third/team + opponent’s completed passes in the final third)
Question the rating is meant to answer
How dominant was the team when it comes to possession higher up on the pitch?
What caught my attention
Whether it’s pressing, creating chances, or dominating possession, Sarri’s Napoli kept popping up while I was looking at the data.

In the last 4 seasons, it’s been the team with the highest percentage of pressures applied in the final third. A record favored by the territorial dominance that they imposed on all Serie A teams between 2017 and 2018.

With a field tilt just above 74%, Sarri’s Napoli also sits among the giants of European football, just behind 3 of Pep’s Manchester City teams.

Progression via carries

Stat used
Share of progressive distance from carries
Question the rating is meant to answer
How does a team progress the ball through the pitch compared to the other teams?
What caught my attention
Burnley is the only team who is represented 3 times in the bottom 10 for share of progressive distance from carries in the last 4 seasons: 21.22% in 17/18, 21.86% in 18/19, 21.54% in 20/21. Percentages that are far from the 29.19% median value for teams of the big-5 teams since 17/18.

Direct play

Stats used
Share of long passes (attempted long passes/total attempted passes), share of total passes completed in the attacking third.
Question the rating is meant to answer
How quickly did the team attempt to get the ball into the final third?
What caught my attention
When it comes to getting the ball into the final third, Getafe doesn’t really mess around.

In the last 4 seasons, they have been one of the teams with the most direct approach, averaging a 98.6 percentile rank for direct play. Long passes account for more than 25% of their total attempted passes while the share of passes completed in the final third gets close to or surpasses the 10% mark. Only 5 teams in the last 4 seasons had more than 10% of their passes completed in the final third. Among them, 19/20 Getafe was the one with the highest percentage (11%).

Defensive approach and effectiveness

Defensive solidity

Stats used
Non-penalty expected goals against, percentage of clean sheets.
Question the rating is meant to answer
How good was the team at limiting the opponent’s attacking output?
What caught my attention
The 19/20 version of Pep’s City — the flawed one, the most fragile, the one constantly exposed to past demons and opponents’ counters — was more solid defensively than 92% of the big-5 teams but the drop was significant enough to make them lose the title to an uncatchable Liverpool.

The consistency in City’s solidity in the years under Pep is although just remarkable. With an average percentile rank of 97, they are the European team with the most consistent defensive solidity in the last 4 seasons.

Intensity of pressing

Stat used
Pressures applied per 1000 opponent’s touches in mid and attacking thirds
Question the rating is meant to answer
How intensely did the team apply pressure to the opponents?
What caught my attention
Packed calendars, long injury lists, little to no pre-season camps. In the 20/21 season, we saw the intensity of pressing ease in all the main leagues.

In Italy, Antonio Conte’s adjustments after a bumpy start of the season led them to the first title in the last 10 years. After acknowledging that his team was not able to protect the space between the lines as effectively as they did in 19/20, Inter starting adopting a more conservative approach when out of possession.

A shift in strategy that proved to be successful.

High press

Stat used
Share of pressures applied in the attacking third
Question the rating is meant to answer
Did the team aim at regaining possession higher up on the pitch?
What caught my attention
The 20/21 season had all the perfect ingredients to make a team opt for a more conservative approach, a more prudent attitude off the ball. Some teams did it by dropping their defensive line, some others by easing the intensity of their pressing

And then there’s Nagelsmann’s Leipzig.

In his last season in charge at Leipzig, the wunderkind pushed his team’s pressing efforts even more into the attacking third.

20/21 RB Leipzig had a higher percentage of pressures applied in the attacking third than 96% of the teams in the last 4 seasons.

Quality of the chances conceded

Stat used
Non-penalty expected goals per shot against (teams that conceded low-quality chances get a higher rank)
Question the rating is meant to answer
How good was the team at limiting the quality of the opponent's chances?
What caught my attention
They just came out of the most vulnerable season of their recent history but, in the last 4 seasons, Juventus have been the best team at limiting the quality of the opponent’s chances.

The probability of their opponent’s shots to end up in the net has only once exceeded 8% (the median value among big-5 teams in the last 4 seasons is 9.8%) and was as down as 7.3% in the last year under Allegri (19/20).

Attacking solutions and threat

Attacking threat

Stat used
Non-penalty expected goals
Question the rating is meant to answer
How threatening was the team’s attack compared to others in the last 4 seasons?
What caught my attention
It took more than 6 months for Coudet to turn Celta Vigo into a team not anymore dependant on Iago Aspas’ contributions but the 20/21 season saw them as one of the most improved teams in the big-5 leagues.

When it comes to non-penalty xG per game, Celta was actually the one with the strongest season-over-season improvement, going from the 4th percentile rank to the 47th.

Reliance on dead balls

Stat used
Share of shot-creating actions from dead balls
Question the rating is meant to answer
How much did the team rely on dead balls to create chances?
What caught my attention
A high percentage of shots coming from dead balls is generally not a good sign for a team’s attacking phase(there’s a negative correlation between reliance on dead balls and non-penalty xG per 90) but there are some teams that managed to get into the top 8 even while making dead balls a key component of their attacking efforts.

Some examples? Getafe in 17/18 (14% of SCA from dead balls) and 18/19 (14.5% of SCA from dead balls), Burnley in the 17/18 Premier League (15% of SCA from dead balls).

On the other side of the scatter — the one where teams are quite productive and tend to not rely on dead balls — Barcelona.

No other team in the last 4 seasons has had a lower percentage of shot-creating actions coming from dead balls than the Blaugranas: 4.93% in 17/18, 4.45% in 20/21, 4.44% in 19/20, 4.01% in 18/19.

Crossing game

Stat used
Share of penalty area entries from crosses
Question the rating is meant to answer
Did the team heavily rely on crosses to get the ball into the box?
What caught my attention
With most of the season spent without a real no.9, City’s use of crosses changed compared to the previous seasons. Among the 38 teams competing in the top 5 leagues, in 20/21 Manchester City was the one with the largest decrease in the share of penalty area entries via crosses: from 9.66% to 5.70%, the lowest percentage for a European team in the last 4 seasons.

Quality of the chances created

Stat used
Non-penalty expected goals per shot
Question the rating is meant to answer
How good was the team at creating high-quality chances?
What caught my attention
PSG’s scoring probability in each of the last 4 seasons vs big-5 leagues average:

• 17/18: 13.13% vs 9.63%
• 20/21: 13.36% vs 9.30%
• 19/20: 14.93% vs 9.45%
• 18/19: 14.96% vs 10.16%

Top teams already benefit from players that can score goals that others can’t (and thus outperform their xGs thanks to their shooting abilities).

But, in the last 4 seasons, PSG has also been consistently good at getting shots with above-average probabilities of being scored.

Goal efficiency

Stat used
Non-penalty goals/non-penalty expected goals
Question the rating is meant to answer
How did the team perform when it comes to maximising the attacking output?
What caught my attention
This is not going to be another of those sections/graphs dedicated to Timo Werner and Chelsea’s finishing struggles in the 20/21 Premier League. But, yeah, goal efficiency can make or break a team’s season.

Take Liverpool for example. Was their “fall” entirely due to poor finishing? Of course not. Not after everything they have been through.

Liverpool’s drop in terms of goal efficiency has although somehow not being talked about as much as it should have.

Take a look at the evolution of their chart (below) over the recent years. Their goal efficiency was strong. Consistently strong. With a percentile rank above 90 in 17/18, 18/19, and 19/20.

And then the huge drop in 20/21.

Not only Klopp had to cope with a team that had lost its core structure at the back (and in midfield as a consequence). For the first time (since they started competing for titles) he also saw his team struggle to convert the (still high) volume of chances created.

Only 7% of the teams in the big-5 leagues have shown a poorer goal efficiency than 20/21 Liverpool in one of the last 4 seasons.

Bonus track

Fun-O-Meter

Stats used
Non-penalty expected goals, successful dribbles, nutmegs, and intensity of pressing.
Question the rating is meant to answer
Which teams have been the most fun to watch in the last 4 seasons?
Inspired by
Tifo IRL’s Care-O-Meter

What caught my attention
Alright. Context first. The metric is not meant to be perfect and skews towards powerful, attacking teams.

Are these the right stats to use to statistically define “fun”?

Maybe not!

I tried to think of what makes a team fun to watch and simply used the FBref stats that I thought could help me measure that.

Here below some “truths” that emerged from the Fun-O-Meter:

• 19/20 Atalanta was more fun than 20/21 Man City
  Tough to swallow? Remember: this is not about the best team or the most dominant. It’s all about fun. And 19/20 Atalanta was arguably more fun than the fluid, strikerless, and yet high-control machine that 20/21 Manchester City was.

• Man City was more fun in 18/19
  Not as dribbly as 19/20 Man City but the most intense in its pressing efforts (2.65 pressures per 1000 opponent touches) and with the 9th best attacking output among top-5 European leagues in the last 4 seasons.

• Chris Hughton’s Brighton (in the 17/18 Premier League), 17/18 Chievo, and 17/18 Amiens have been the least entertaining teams of the last 4 seasons.
  Although I haven’t spent much time watching Amiens play, I can’t say I am surprised by seeing Chievo and that Brighton at the bottom of the chart.

The Thing You (probably) Came Here For

The final product. The bite I used to get you to spend 12 of your valuable minutes to go through all these charts and data and graphs you sometimes liked and sometimes really didn’t.

There was too much meat/preferred source of protein to just throw some random charts into your Twitter feeds. I wanted to add a bit more context to explain why I picked these metrics and take some time to analyse some of the most interesting findings.

There’s more to it, though. You can now play around with the interactive dashboard and get into some of the other data points yourself.

Just click on the GIF above to access the dashboard. Hit me up on Twitter if you liked it or even more if you didn’t.