In anticipation of an opening weekend of great football, I thought it'd be fun to take a quick look at fouls committed in the Premier League last year. Who fouled the most and the least?* Take a look.
Spurs led the league with the smallest number of fouls called per match at 9.5; at the high end, Bolton were called for infractions an average of 12.63 times per match. That's a difference of about 30 percent. But what's also noticeable about the graph is that the range of fouls committed by PL teams is actually quite small, on the order of 10-12, and that there seems to be little connection to how teams fared in the league overall.
What this graph hides, however, is that some teams consistently foul less or more, while others commit lots of fouls on occasion but have other matches when they commit very few fouls.
Below is a graph of teams' variability in fouls committed (measured by the standard deviation), but ordered by foul frequency. You can see that some teams consistently foul very little - as can be seen in the smaller standard deviation for Spurs, for example - while others like Everton consistently fouled quite a bit, and others still (like Blackpool) generally fouled little but had some matches where they committed lots of fouls.
Here's another look at the same data - this time lining up teams by their mean foul performance and how variable the number of fouls per match was over the course of the season. In the lower left hand corner are the teams that consistently fouled less - these include Tottenham, WHU, and Birmingham. In the upper left corner are the teams that consistently (as expressed by the small standard deviation) fouled more - here, we see Everton, Manchester City, Liverpool, and Blackburn, for example. In the upper right are the teams that fouled just as much or more (Bolton, Fulham, Wigan, etc.), but their infractions were much more variable from match to match. Finally, in the lower right are the teams that generally fouled less, but on occasion fouled a lot - these include Blackpool and Chelsea.
* I'm assuming for convenience sake that referees are randomly, i.e., equally, biased against all teams.
