Tuesday, 11 September 2012

Goalscorer Footedness

For my first attempt at looking at the 2011/2012 Premier League Opta data released as part of the MCFC Analytics project, I decided to try to quantify the 'footedness' of players. The original aim was to measure this for all players, but the data was only really useful for shots attempted, and so I pared it down to goalscorers only.

Somewhat arbitrarily, I defined the footedness of a player as (R-L)/(R+L), where L and R are the number of left- and right-footed shots respectively. I like it because it means the range is from -1 to 1, where -1 is a wholly left-footed player, 0 is a player who uses both feet equally, and 1 is a wholly right-footed player. It's not completely intuitive, but as a guide players who shoot with one foot twice as often as another have a footedness value of 1/3, three times as often is 1/2, and so on. The results for all players who scored 10 or more goals are below:


Of the top 20 goalscorers, 20% are left-footed, although it might be worthwhile expanding the selection to more goalscorers to see if this tends towards the usual 10% figure quoted for the population at large, or if left-footers are more prevalent in football. Edin Dzeko is the ambipedalist of the group.

I thought it might be interesting to see how much more accurate a player was with their favoured foot compared with their unfavoured foot, which I calculated by subtracting the unfavoured foot accuracy from that of the favoured foot. It turned out that wasn't even a particularly well-phrased question, as 12 of the 20 players are more accurate with their 'wrong' foot! In some cases the difference is very slight, and in the cases of Van Persie, Sturridge, Balotelli and Graham it barely seems to matter if they shoot with their favoured or unfavoured foot.

I think the case of Steven Fletcher provides a potential explanation for what seems to be a strange quirk. He seems to be 49.1% more accurate with his wrong foot! So surely he should be shooting with his right more often? However, he is so one-footed that he hardly ever shoots with his right foot, and in fact over the entire season he only took four shots with his right foot, three of which were on target. I surmise that the only time he shot with his bad foot is when it was difficult to miss the target, meaning shots close to goal. I could probably follow up on that point at another time, to confirm it, but it's probably a decent guess.

I suspect that players are less likely to take speculative shots with their weak foot, which might inflate the relative accuracy of their bad foot. In turn, this means that the approach of forcing a very one-footed player on to his weaker foot is probably still a good idea, even if the difference in the accuracy of the two feet doesn't seem to bear that idea out.

A final note on the accuracy of the different players: Papiss Demba Cissé is the most accurate with 78.3% of his shots on target, while Steven Fletcher is bottom of this particular bunch with a meagre 32.3% of his shots on target. As a Liverpool fan, Suarez's meagre figure of 44.9% explains a lot.

1 comment:

Karl Jackson said...

We are looking at similar numbers in the Australian Football League, but with all kicks not just shots at goal. Roughly a third of players are "better" on their non-dominant foot, which I also put down to players being more likely to use their non-dominant foot for easy kicks. Glad to see analysis in other sports coming up with a similar conclusion!