Referees have a damn hard job in football, and the last few weeks, with Mike Dean being "demoted" (how does this even work? how is working in a similar environment, with perhaps more incidents per game in the Championship work? why is this a demotion? The answer, as ever, is money, power and exposure. But I digress) the work referees do and how they perform is increasingly important.
There's a couple of things which have been on my mind lately, and it stemmed from someone in the Analytics community (I spent a bit looking for the actual tweet in question, but couldn't find it...) saying how Ander Herrera is notoriously effective at fouling multiple times without picking up a yellow.
This is interesting to me - in big games, physicality can be a great weapon (of course, and dangerous, anti-football etc etc.) and knock the big players off the stride. Cristiano often talks about having lumps kicked out of him as defenders would either try to get under his skin, or knock him off his game. Indeed, I recall Atleti doing something similar to nullify Messi in a Champions League game (if my memory serves me correctly)
The ability to get away with a foul uncarded, though, depends on the league. Some leagues are more lenient, not wanting to disrupt the game or unnecessarily break the match, and others are a lot more card happy. This is what this post is going to focus upon, and potentially build out and look at individual players in a follow-up post.
Let's talk about leagues.
The way I've gone about this is quite simple - Instead of looking at individual players, and which players seem to have such a skill at a very, very high level, we can start by looking at teams, and therefore, leagues.
The metrics I've used are very basic - I've taken data from this season for 7 leagues; Premier League, Championship, Bundesliga, Eredivisie, Serie A, Ligue 1 and La Liga for this season - and the only fields I've looked at are;
- Games played
- Cards (Yellow + Red)
- Fouls/Fouled
Within this, I've look at fouls/fouled per game, yellows and reds per game as well as the number of fouls committed for every card received. Let's get stuck in.
I'll post the dashboard here, and if you scroll down, you'll find discussion points from me, with a few screenshots.
The first story point (there are two, one in each tab along the top), focuses on a scatter plot. This contains every team from every league mentioned above for this season (before the round of games for the weekend of 21st Jan 2017)
On one axis is the cards per game, and the other has fouls per game. Tab through and find your own team, and compare it across the leagues.
The second set of charts is the dot-plots, and I've added a median line, with an confidence interval of 95%. These I've gone through in a bit more detail below.
Fouls per Game
The leagues are each sorted by the median of the league metric - so the league with the highest number of fouls per game is the Bundesliga, whilst the Premier League is the one where there are less fouls awarded per game. My favourite team here is Stoke - Who knew that the old school, long ball, physical team would be the cleanest in the league a few years on! PSV, as we'll see throughout, have a knack of not fouling often, whilst Real Madrid, Napoli, Bayern and Derby are the "bigger" teams in the league who come out with the lowest fouls per game in their respective leagues. Game effects? A better team spends less time defending, so wouldn't have to worry about making fouls? This also gives an inkling of the management of the games in each league - which referees play advantages vs. blowing the whistle for any sort of foul play.
Cards per Game
When we look at fouls, we then have to look at how trigger happy referees are when pulling out their cards. La Liga has the highest number of cards awarded across the league, whilst referees in the Eredivisie and Ligue 1 1.5-2 cards per game.
Fouls per Card
The danger here, of course, is inferring that teams which foul and are carded less are lucky, and those who foul less and are carded more and unlucky - what if the latter have some tough tacklers who often make bad challenges? What if the former are pushes in the back, and body checks by various players across the teams. This is where the data should be investigated to explain the phenomenon. Drilling down to player level will support this.
What's the point? Where's the value?
With any sort of data analysis, particularly in sport, there are two key questions. The first is, what's the point in doing it? The second is, if this is useful information, how can we use it to our advantage as a team?
The answer to the first question is simple; there are certain teams across every league with a style of play where tactical fouls are common. Is this because of their aerial style of football? That they play with 2 robust central midfielders or defenders? Or are they simply bullies, who have the ability to shut down flair players through sheer physical player.
This doesn't even start to answer those follow up conversations or questions, but it does do something quite key - it starts those conversations. It gets people thinking from a set of basic metrics which are easy to understand. And this is where true buy in can begin within clubs, and indeed, within fan bases.
As a team, the advantage of this information would be for a pre-match scouting report. How can this help us identify the style of play that the team plays? One step further... Who are the key players who help them play like this? Again, basic analysis breeds the desire for more context to inform better, smarter decisions - and in a shorter space of time. And further? Once you start getting a good idea, a smart analyst would begin to look at manager styles (hat tip to Statsbomb Services who advertise that they do this sort of thing), and then even look at referee styles & data, to see if that can be gamed. Edges and gamification of league rules.. That's the way sports teams get ahead, right?
Have a play with the dashboard above, and do be in touch any thoughts, ideas or feedback.
0 comments:
Post a Comment