How repeatable is sh%/sv% from outside the box?

A special thanks to Dan Kennett for the data to make this post possible

Pretty simple one this. I have four seasons worth of data, (n=68 non-relegated teams) providing me with 51 pairs of back-to-back Premiership seasons. I’ve taken all shots from outside the box, as I don’t have the number of shots from outside the box that were on target, looked at the number of goals that resulted from those shots, and calculated the ‘outside the box’ sh% and sv% for each team season. So what does a plot of ‘outside the box sh% in year 1’ vs ‘outside the box sh% in year 2’ (the following season) look like?

obsh%rttm

Put simply, at the team level and over the course of a season, scoring goals from shots taken outside the box isn’t repeatable. It is in fact 92% luck driven, and 8% skill driven. Spurs benefited from this luck the last time around, we shouldn’t expect them to score as many goals from outside the box this season, Bale or no Bale.

And what if we do the same, but look at a teams sv% for shots taken from outside the box?

OBSv%rttm

This one isn’t quite so cut and dry, the correlation coefficient is an order of magnitude higher, which means that saving shots from outside the box is 77% luck driven, and 23% talent driven. Even with such a large luck component, the difference between the best team at stopping shots and the average is about 2 goals. Regress that 77% of the way towards the mean and having the best keeper/defensive system at preventing shots from outside the box going into the net may be worth half a goal a season to your team.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s