clock menu more-arrow no yes

Filed under:

Thursday's Numbers: Having fun with the shootout

New, comments
Getty Images

The Canucks have played in a lot of shootouts recently, scoring, if memory serves correct, twice on nine attempts while allowing goals on, again, I'm chalking this up to memory, three times on ten attempts.

The goaltending has been pretty solid. I think we have enough shootout data over 6 years of doing this shameless post-tied-game ritual to determine that the total NHL save percentage is .674, meaning the total NHL shooting percentage is .326.

Before I go on with the rest of this post, I'll mention I piggy-backed a bit off of a couple of links that Jordan posted in the Buzz section this morning, in particular this piece from Gord MacIntyre highlighting Mike Gillis' interview on Team 1040. There are a lot of things to discuss, but I figure that the shootout bit is more relevant.

In pin-pointing the problems that the Canucks have historically had in the shootout (28-37) it's a little disconcerting that, when asked whether it was shooting or goaltending, Gillis simply said "both". And he's right, sorta. The Canucks have shot and saved shots below the median:

Vancouver NHL Difference
Shooting 31.7% 32.6% -0.9%
Goaltending 63.0% 67.4% -4.4%

(UPDATED - Plugged in this season's numbers for all-time results for some reason)

So, goaltending has been "worse", I guess, but, over the large sample of games, the number is so closely resembling the median. Flip 10 coins, and you have a 12% chance of 7 of them being heads. Flip 1000 coins, and you have a probability of getting 700 heads that closely resembles Scott Gomez's goal total. So any problems will be far more noticeable over any one season, but over time, they begin to even out.

From a goaltender's perspective, using this fun little app, I figured that a goalie who has the average save percentage of .674, on 10 shots, there's a 20% chance that five or more will go in. On 100 shots, the percentage begins to closely resemble Scott Gomez's goal total. (I may call this the Scott Gomez asymptote)

The NHL, as it were, this season, slightly favours the shooter:

Van 2012 NHL 2012 Difference
Shooting 28.6% 33.9% -5.3%
Goaltending 52.9% 66.1% -13.2%

It would be easy to just say "goaltending" and call this off, but that hasn't been the problem lately. Certainly, the 0-for-3 start in the game against Pittsburgh didn't help, but the goaltending has been above-average lately. Shooting is still down. So, who should shoot? And why?

From a Botch blog:

"I'd like to have a few different groups that compete against each other" he said.

"We have enough guys who can score. We should be better.

"It's up to the coach. He thinks he has better guys. But if we [him and Daniel] had another chance, we'd love to take it."

In his career, Henrik is 0-for-3 and Daniel is 4-for-21.

In a small stretch of time, we can see some weird results. Alexander Burrows and Mason Raymond, for instance, are both 50%, which is the maximum level of observed talent among the players who have taken the most shots in their careers (the lowest being 25.5%).

Using regression calculations (if I'm doing this right) we can calculate Daniel's "expected" shooting talent to be at 29.4% and Henrik's, actually, at 32.0%, so maybe, indeed Hank ought to get a crack at an attempt. Behind Burrows and Raymond, the Canucks have players with limited success this season, so maybe there ought to be a shake-up.

We screw around looking at skaters because there's more options here than with goaltending. Gills effectively nixed the idea of bringing a goalie in cold for the shootout, and I have to wonder if that novel idea would even make a difference. Both Schneider and Luongo have career .667 save percentage records, with Luongo having far, far more attempts. We have no surefire way of telling who's better at this point.

Option A - Henrik Sedin: As alluded to above, may have some hidden scoring talent in there. His three attempts were spread out over five seasons, and he's a much better scorer today than he was five years ago.

Option B - Maxim Lapierre: Lapierre has 3 goals for 6 attempts in his career. It's a small sample, but he does have some skill, and he has yet to be tested by Vigneault as a Canuck.

Option C - Um... those are the only ideas I have.

The team is pretty set on using Ryan Kesler and Alex Burrows, and are possibly leaning on Cody Hodgson maybe a little too much because he scored a pretty one on his first attempt. However, since we're looking at a weighted coin flip, perhaps it doesn't matter who the heck is used, since everybody on the Canucks has an ability to score goals at some inherent level.

Is it worth practicing? No. It's so random. Being "historically bad" in shootouts, the Canucks are, on average, 4.3 and 5.6 per season on a wins and loss record. They are, essentially, .6 points below .500 per season on this. That is worth less than 1/3rd of a win in their history per season, even despite all the problems the Canucks have had.

Now, I'm all about adding up the small aspects of the game, but consider that a goal is about 1/6th of a win. The Canucks are so bad at something that is so marginal that it costs them less than two goals per season. Saying "the Canucks are really bad at shootouts" is exactly the same as saying, essentially, that "the Canucks are really good at bounces off the stanchion".

It's so marginal, and so random, and in the scheme of things, I'd rather the Canucks work on things where the true team talent can lie in something greater than being worth two goals a season.