December 04, 2006

NHL Shootouts, Season 2: More Shots, Fewer Goals: So far in the 2006-07 season, there are more shootouts per NHL game, more attempts per shootout, and more saves per attempt. What's changed?

Last week the Boston Bruins beat the Tampa Bay Lightning 4-3. The game went to a shootout, where goaltenders Tim Thomas and Marc Denis saved thirteen of fourteen shots. The lone goal scorer was the Bruins' seventh candidate, Phil Kessel.

Somewhere in my years of watching hockey I had picked up an old saw that a penalty shot was about a 50/50 proposition. My own observations of the shootout over the past two seasons didn't match with that theory, though. I knew that the Boston-Tampa Bay outcome was a bit freakish, but my impression was that goals are scored a lot less than 50% of the time. So I got curious about the statistics of shootouts.

Using data from this website I found out that in the 2005-06 season shooters scored on just about one-third of their shootout attempts: 330 goals on 981 attempts, which gives a success rate of 33.6% for the season.

So that old saw about the 50/50 chance might be just a myth, out of date in today's NHL. Or it might just be the right answer to the wrong question. If we look at goaltender save percentages in the shootout, they come out to just better than 50%: in 2005-06 goalies stopped 514 of 981 attempts (52.4%). So it would be accurate to say that when a shootout attempt results in a shot on goal, it results in a goal about half the time. But if you have to place a wager on the outcome of any single attempt, odds are heavily in your favour if you bet against a goal.

And it looks like things are tipping even more in the goalie's favour in 2006-07.

First of all it's interesting to note that there have been more games ending in a shootout in 2006-07 than there were in 2005-06. This plot shows the number of shootouts required as a function of time, represented by the NHL game number. The blue dashed curve shows the cumulative number of shootouts for the 2005-06 season, and the red dashed curve shows the same plot for 2006-07 (through Saturday's games). In 2005-06 there was a shootout every 8.3 games, on average. In 2006-07, that's increased to one shootout every 7.6 games.

There has also been a pretty big increase in the number of attempts taken, even after the increased number of shootouts is taken into account. For one thing, there have been more long shootouts. There were seven games last season that had at least fourteen attempts. There have already been seven shootouts this season that had at least fourteen attempts. (One of those, the one-for-twenty-six display of futility put on by the Rangers and Flyers on October 7, was the longest single-goal shootout so far.)

It's not just a few long ones, though; shootouts are longer, on average, by about one attempt. The solid curves in the same figure show the cumulative number of attempts taken in shootouts for the two seasons. In 2005-06, there were 6.8 attempts per shootout, on average. So far in 2006-07, that's increased to about 7.8 attempts per shootout.

The combined effect is that fans are getting to see a lot more shootout attempts. Using the numbers I quoted above we can see that there were about 6.8 attempts every 8.3 games, which works out to just over 0.8 attempts per game. This year, that number is up about 25%, to slightly more than one attempt per game.

I could come up with some theories about why there have been more shootouts this year, but I think I'll leave that to the readers for now. As to why they are lasting longer, at least part of the reason is that it's getting harder to score. This plot shows the scoring percentages (solid lines) and save percentages (dashed lines) for 2005-06 (in blue) and 2006-07 (in red). So far in 2006-07, scoring percentages are down (to 30.4%) and save percentages are up (to 56.5%) compared to last season.

The changes appear to be significant, but I'm at a bit of a loss to explain them. I guess it's possible that some shooters are using sticks with more curve, and that this is somehow a disadvantage in a shootout situation. Or it might be that the shift is due to the presence (or absence) of a few gifted players or goaltenders.

Or I could just admit that I don't know, leaving room for Spofites to insert their own theories and opinions. Anyone?

posted by Amateur to commentary at 01:37 PM - 0 comments