There aren't a whole lot of interesting outcomes from the shot target analysis I started earlier, so I thought I'd do a data dump for posterity. First, the distribution of shots by shot type in our database over the last four season:
Here we have the percentage of shot targets by shot type. Recall that location 1 is high/right (from the goalie's perspective), 2 is high/left, 3 is low/right, 4 is low/left and 5 is the five-hole.
You can see that there are more shots directed at the goalie's right hand; not surprising given the preponderance of left-handed shots in the NHL. Exceptions to this rule are wrap-arounds (where a left-handed shot is going around to the left side of the net) and backhands (which we'd expect to be opposite.) Players rarely get wrap-arounds off the ice (no surprise) but deflections are primarily low too, which I wouldn't necessarily have expected.
The lesson here is that if you're close to the net and you get the puck up, you're likely to score - the goalie is probably down. This is less true of slap and wrist shots, where the goalie is more likely to be covering the upper half of the net.
Wrist shots look like they're much more effective than slap shots, but they're also taken closer to the net. If we normalize the set of all wrist shots as though they were taken from the same locations as slap shots (ie - use wrist shot shooting percentage and slap shot totals), we actually find that wrist shots are no more effective than slappers:
|# Slap||Slap G||%||Wrist G||%|
We can also see that wrist shots and slap shots are roughly as likely to hit the upper half of the net if we normalize locations:
|# Slap||Slap Up||Up%||Wrist Up||Up%|
Anyways, I'm all ears if you have any ideas for what we can do with this data. I was thinking of comparing outcomes by handedness and wing, but I fear that we won't get much in the way of exciting results.