Wednesday, April 9, 2014

Does possession matter?

Last month when Michael Bradley commented that possession is overrated, it rekindled an age old debate whether more possession leads to more wins. The MLSsoccer.com article took the opposing side, citing a graph showing the percentage of teams with more than 50 percent of the possession who made the playoffs. Around the same time, the MLSsoccer.com also published an article that San Jose Earthquakes, the winner of Supporter’s shield in 2012 was the best attacking team in the MLS history. When checking their stats, I noticed that they managed to win many times despite having less possession. I digged deeper - and San Jose Earthquakes are not alone, so did New York Red Bulls on their way to winning Supporter Shield in 2013 - in fact they had a very convincing win over Dynamo (3-0) with as little as 39% possession. I picked up few more games at random, and sure enough there were plenty of examples in MLS last season where side with lower possession won.

At this point, I decided to investigate further, collecting more data points systematically. I picked up three rounds (2, 18 and 32) in the beginning, middle and end of the season and looked at all the games that produced a result.What I found was that there is some positive correlation between more possession and winning, unfortunately the correlation is not very strong, as shown below(I didn't include ties):

 More Possession  Equal Possession(49-51%)  Less Possession 
13 wins7 wins12 wins

So, I would rather agree with Michael Bradley, possession is indeed over rated. It is a factor but not an overwhelming one, at least in MLS. In search of other factors,  I looked at few other statistics like total shots, shots on goal, duals won etc. So far I have not been able to find any single overwhelming factor but found that shots on target has higher correlation with win than possession. I took the same games from last year and compared the winning and losing team's shots on goal data.

 More Shots on Goal  Equal Shots on Goal  Less Shots on Goal 
17 wins8 wins7 wins

As a side note, I observed that only three teams in 32 games managed to win in spite of having less possession and also with fewer shots on goal. For one of them, possession was almost equal (47% vs 53%) and difference of shots on goal was just 1;  in another instance, the winning team’s goalkeeper was MOTM, and in the third case the winning side had two stoppage time goals. I'll be sure to revisit this topic in a future post with more factors and data and hope to pinpoint three things which a team must focus on in order to win.
*Here is a link to the data which I have collected

No comments:

Post a Comment