Corsi Percentage: Up One, Down One and Tied
by Hawerchuk on Mar 15, 2010 9:00 AM EDT
10 comments
I just wanted to re-visit league-wide Corsi percentages at even-strength for teams in tied or one-goal games. Chicago continues their dominance on the shot board, while Toronto is improbably second:
| Up 1 | Tied | Down 1 | |
| chi | 2.3 | 8.5 | 16.1 |
| tor | -0.7 | 3.6 | 5.5 |
| nj | -2.4 | 3.1 | 2.9 |
| det | -0.6 | 3.0 | 4.0 |
| bos | -0.2 | 2.7 | 5.1 |
| phi | -0.6 | 2.7 | 7.1 |
| cgy | -1.9 | 2.0 | 8.0 |
| pho | -1.1 | 1.7 | 3.2 |
| pit | -1.6 | 1.5 | 4.4 |
| sj | -4.2 | 1.0 | 3.5 |
| ott | 0.0 | 1.0 | 7.5 |
| buf | -4.6 | 1.0 | 2.9 |
| la | -3.8 | 0.9 | 6.3 |
| nsh | -3.1 | 0.9 | 5.6 |
| stl | -7.0 | 0.5 | 4.4 |
| was | -3.7 | -0.2 | 7.7 |
| car | -3.4 | -0.2 | -1.3 |
| nyr | -3.9 | -0.2 | 4.2 |
| dal | -6.8 | -0.3 | 3.3 |
| atl | -12.0 | -0.7 | 3.9 |
| cls | -5.5 | -1.0 | 0.6 |
| van | -2.9 | -1.3 | 4.2 |
| nyi | -5.3 | -1.7 | 1.0 |
| min | -6.0 | -2.5 | 6.0 |
| tb | -3.9 | -3.2 | 1.3 |
| ana | -8.1 | -3.5 | 5.3 |
| fla | -4.6 | -4.2 | -2.0 |
| mon | -8.0 | -4.5 | -0.4 |
| edm | -4.7 | -5.0 | 0.4 |
| col | -6.4 | -6.0 | 2.2 |
| AVG | -3.85 | 0.0 | 3.85 |
So how can Toronto be such a bad team despite dominating shot counts, while Colorado is such a good team despite being at the opposite end of the spectrum? It's instructive to look at a team's shooting percentage minus its opponents' shooting percentage at even-strength in tied games:
| SH%-SV% | |
| was | 35 |
| col | 27 |
| van | 24 |
| sj | 18 |
| la | 18 |
| chi | 18 |
| pit | 18 |
| phi | 13 |
| mon | 8 |
| stl | 8 |
| ott | 8 |
| nj | 7 |
| min | 7 |
| nyi | 5 |
| ana | 1 |
| tb | 1 |
| cgy | 1 |
| nsh | 0 |
| pho | 0 |
| fla | -2 |
| car | -8 |
| dal | -8 |
| det | -16 |
| atl | -17 |
| buf | -18 |
| cls | -20 |
| nyr | -23 |
| bos | -23 |
| edm | -30 |
| tor | -32 |
There's your answer: Colorado and Toronto are completely polar opposites in the goaltending and shooting departments. Some of that is deterministic - Craig Anderson is a very good goaltender, while Vesa Toskala was embarrassingly bad. But part of it is also that Colorado has a much higher shooting percentage than Toronto, a difference that tends to disappear over time. If we played 1000 games, Toronto wouldn't be this bad and Colorado wouldn't be this good. But when you only play 82 games (or not even 82 games yet), you sometimes get something other than the most likely result.
0 recs |
10 comments
|
Email
Print
Comments
What about the Canucks?
What do you make of the Canucks results? This is the second straight season they’ve been top-3 in terms of EVSH%. Alex Burrows and Henrik Sedin are both well above the league average SH% for the second straight season and above their career averages. Do you think it’s just a fluke or is there maybe something else going on (generating higher quality chances, shot recording differences, etc)?
They were at 8.1% in 2007-08, then they’ve been at 9.6% for the last two seasons. If you put the Sedins up against crappy opposition, you can certainly get better chances.
Pittsburgh (efficiently) and Atlanta (inefficiently) are the only teams that might actually be generating better chances than average for three seasons.
So, are the Wings getting some bad luck in addition to all their injuries?
Can we draw that conclusion?
No
Bettman hates them and there is a major conspiracy making them lose.
Jibbles is an older man so you can trust what he says.
by Bob in Boulder on Mar 17, 2010 11:38 AM EDT up reply actions
are those numbers in the first chart corsi differentials (i.e. +/-) or percentages?
curious that league average is -3.85 when up one and +3.85 when down one. i wonder if third period numbers alone are skewing that. (i.e. I wonder what league average corsi diff is if we use, for example, vic’s “close” definition – which i think is something like “within one goal in the first two periods or tied in the third”).
Having a one-two center punch of Paul Stastny and Matt Duchene is going to lead to some pretty high quality scoring chances. So this whole “over 1,000 games things will even out” thing is kind of crap, no?
Jibbles is an older man so you can trust what he says.
by Bob in Boulder on Mar 17, 2010 11:37 AM EDT reply actions
Teams don’t usually maintain high shooting percentage from season-to-season. In fact, the correlation between even and odd games in a single season is really low. It doesn’t matter who’s on your team. The best forwards generate a lot of scoring chances, but the percentage that are high-quality is roughly the same regardless of skill level. Teams are much more adept at getting more chances, and not so much at making the most of the ones they get.
But Stastny and Duchene go to 11.
Jibbles is an older man so you can trust what he says.
by Bob in Boulder on Mar 17, 2010 3:24 PM EDT up reply actions
Comments For This Post Are Closed
Secondary Sidebar
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2010 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
