So I have been looking at some data on blocked shots, specifically team blocked shots percentage (blocked shots for / Corsi against).
We all know that using raw totals for blocked shots is useless, since it actually correlates very slightly with losing (since a blocked shot means the other team has the puck.) Other people around the hockey world have also been taking a look at blocked shots.
So I decided to take a look at some data from 2007-2008 season to the 2011-2012 season at the team level. I compared blocked shot percentage for (the rate at which a team blocks Corsi events against) to several advanced metrics, such as Corsi and Fenwick.
|CF60||FF60||CA60||FA60||CT60||FT60||Fenwick Shooting percentage against|
So I compared blocked shot percentage for (a defensive stat) to Corsi and Fenwick for per 60 minutes (offensive stats), Corsi and Fenwick against per 60 min, Corsi and Fenwick total per 60 min, and Fenwick shooting percentage against.
Corsi for and Fenwick for are offensive stats. I didn't expect a very high R^2 for those two, since there is a lot that happens in between a blocked shot on defense and getting a shot attempt on offense, but I compared them just for kicks. As expected, blocked shots at a higher rate has no affect on your offensive stats.
The next few columns is where I become interested. Blocked shot percentage is explaining some of the variance in the Corsi columns. Now I know that a .2 R^2 is not great, but I'd say it's definitely a decent number for this comparison and at least gives us a hint that there is some connection between Corsi and blocked shot percentage.
What we are seeing is that teams that block shots at a higher rate tend to also give up more Corsi events against. When I first saw this my immediate reaction was that the data was telling us that teams get better at blocking shots when they have more opportunities (i.e. a bad team that gives up more Corsi per game will be better at blocking shots). But if that were the case we would see the same thing in our Fenwick stats. However, we don't.
Basically, a team's ability to block shots at a higher rate is not affecting how many shots actually get past the defenders and either hit the goalie or go wide of the net. So while shot blocking at a higher rate isn't a bad thing, it also doesn't seem to be a good thing. Given that teams that block shots at a higher percentage don't give up less shots and less missed shots, shot blocking at a higher rate isn't actually accomplishing anything.
But a second question I had (spurred on by Cam Charron's article and Twitter conversations) was whether blocking shots was a net negative since it created another screen for the goalie to look around. However, the data isn't showing any connection between Blocked shot percentage for and Fenwick shooting percentage against (goals / Fenwick).
I take from this that the damage done by screening your goalie by trying to block a shot is most likely washed out by the benefit of forcing shots to go wide by attempting to block them.
Everything I have looked at so far is pointing towards shot blocking be a semi-irrelevant skill. Now some have suggested that shot blocking is only valuable when blocking non-point shots, since those shots are more likely to go in. I don't have the data to figure that out, but it would be an interesting study.