Friday, August 5, 2011

Shots and Goals In the Big 4: Offensive Performance Across Leagues in 2010/11

During the past year, I have compared teams across the Big 4 leagues (Bundesliga, Premier League, La Liga, Serie A) with regard to various performance indicators. With the Bundesliga about to get underway tonight, I thought it'd be interesting to see where they all ended up by the end of the season, and to find out what we may expect for next year. So for today, here are season totals (averages per team and match) on three dimensions of offensive production - shots, shots on target, and goals - across the four leagues and 78 teams in total.

On average, teams across the four leagues took about 13 shots per match (12.96). Real Madrid led the top European leagues at almost 19 (18.9), followed by Inter, Dortmund, Chelsea, and Palermo. At the bottom end, we saw 3 English clubs, with Blackburn, Wolves, and Birmingham who were bottom in Europe at a measly 8.6 shots per match. That's a whopping 54% fewer shots than Madrid. But this did not necessarily imply an overall inferiority of English clubs, it turns out, as teams from all four leagues are distributed across the graph. Perhaps somewhat surprisingly, Barcelona came in only at number 8 overall in average shots per match, with roughly the same season production as Milan and Arsenal. What is perhaps most surprising about the graph is the relatively narrow range that the vast majority of teams find themselves in when all is said and done. With the exception of a few truly unusual teams at both the high and low end of the distribution, a large percentage of them are located somewhere between (roughly) 12 and 14 shots per game. And interestingly, some amazing teams are neighbors with some very mediocre ones. So one thing to take away from this first graph is that there is relatively little that separated the great from the not so great when only the average number of shots per season is considered.

So the next obvious question is: what makes the difference? Is accuracy is the name of the game, rather than average shot production? That is, do the teams that shoot the most also shoot more accurately, or do the best teams shoot just as frequently but more accurately? To answer that question, the next graph shows the average number of accurate shots per match in 2010/11. Keeping in mind that the overall average of accurate shots across the top leagues was 5.2, the accuracy chart displays a curious pattern. In particular, while we do see some of Europe's very best teams at the top, there is also a preponderance of English teams closer to the top than the bottom. As I have noted in previous analyses, PL shooters are on average more accurate, a fact reflected in this graph as well. And of course, to some extent you can only have accurate shots from overall shots, so it should not be surprising that the correlation between the two offensive metrics for 2010/11 is .64 (but this correlation is particularly high in England at .83 and low in Italy at .60). So I find it amazing to see Barcelona, Liverpool, Dortmund, and Wigan all clustered so closely together. 
Shooting is nice, but who actually produced goals and at what rate? Across the four leagues, the average number of goals per team and match was 1.37, yielding a match total of 2.74. The Bundesliga led the pack at 1.46, followed by the Premiership (1.4), La Liga (1.37), and Serie A (1.26). So goal-wise, Bundesliga supporters had the most fun in 2010/11. On a team-by-team basis, Real and Barca led both La Liga and the top European leagues at astonishing rates of 2.68 and 2.5, respectively. Other teams that gave their supporters reason to cheer frequently included Bayern Munich, Manchester United, and Dortmund. The only team that seems to be in the "wrong" place at the top end is Stuttgart - the Schwaben managed to score more than Milan or Manchester City. At the low end, pitiful Bari managed not even close to a goal per match at .71 full time goals on average. But Bari were not alone - ten teams had an average goal total per match of less than 1, including other relegated teams like La Coruna, Hercules, Almeria, Frankfurt, Birmingham, Brescia, and Sampdoria.
I've often said that you can't score unless you shoot, but which teams had to shoot more to score? That is, what was the statistical relationship between shots and goals in 2010/11? To get a glimpse at that, the next graph plots teams' average shot and goal numbers (per match) for the season as a whole. For orientation, I have superimposed a regression trend line that tells us if more of one thing (shots - on the x-axis) goes with more or less of the other (goals - on the y-axis). Teams on or near the trend line were most like other teams in the relationship between shots and goals; teams above the trend line were able to score more with fewer shots; and teams below the trend line had to take more shots to score an equivalent number of goals as the average team. Take a look.
Most teams cluster around the trend line, and this trend is upward sloping. This tells us that teams that took more shots on average also scored more goals. No surprise there, and the data are quite well "behaved" - meaning that most teams aren't all that far away from that overall trend line. In case you care, the correlation is a healthy .71 and the R-squared a very robust .51. And while most teams follow the general trend, a few don't. On the high end, Madrid, Barcelona, Munich, and to some extent, Manchester United scored more effectively, netting more goals with fewer shots. On the low end of things, Bari, La Coruna, Sampdoria, and Brescia just couldn't manage a score - they took fewer shots than most, and it took them more shots to eventually beat the keeper.

Good luck finding your favorite team in the maze of data points - and feel free to suggest interpretations I haven't thought of.