DETROIT - MARCH 22:  Henrik Zetterberg #40 of the Detroit Red Wings turns up ice with the puck during an NHL game against the Pittsburgh Penguins at Joe Louis Arena on March 22, 2010 in Detroit, Michigan.  The Red Wings defeated the Penguins 3-1.  (Photo by Dave Reginek/NHLI via Getty Images)

I often talk about the value of Corsi (and Fenwick and other shot-based metrics) in evaluating teams and players, because I believe that they do a good job of showing possession ion the offensive zone, and that possession of the puck in the offensive zone leads to goals, and prevents goals against.


Using Corsi the other day over in this thread, I was chided by a commenter who tossed out this criticism:


You know what Willis always says…

"Corsi wins championships"

… Oh right, no one says that, because that’s retarded


Because I’m me, I decided it might be interesting to look at the shot data for the Stanley Cup champions, going back as far as the NHL makes the records available (1997-98).  One quick caveat – this isn’t Corsi, but rather straight shot differential, which is less precise.  Here are the regular season and playoff results for out-shooting, per game, for Cup-winning teams going back to 1997-98:


Season Team Regular Season Shot Differential Playoff Shot Differential
1997-98 Detroit +5.1 +5.1
1998-99 Dallas +3.6 +4.8
1999-00 New Jersey +7.5 +8.3
2000-01 Colorado +4.1 -0.7
2001-02 Detroit +4.7 +5.3
2002-03 New Jersey +8.1 +2.4
2003-04 Tampa Bay +4.6 -0.6
2005-06 Carolina +0.6 +1.8
2006-07 Anaheim +4.1 +2.9
2007-08 Detroit +10.9 +12.9
2008-09 Pittsburgh -1.3 +3.5
Averages   +4.7 +4.2

It’s an interesting chart, I think, and one that shows at least a strong indication that outshooting is incredibly important.  10 of 11 teams out-shot in the regular season, and nine of 11 outshot in the post season., while the average team fired nearly five more shots per game than their regular season opponents, and just over four more shots per game than their playoff opponents.


Next up: comparing shot differential with other things (offence, defence, special teams, goaltending, etc.) that reputedly win championships.

Comments (12)

  1. Betcha if you had Corsi going back that far, the handful of outshot results go away.

  2. Doogie2K: It wouldn’t surprise me, although a strong run of goaltending can compensate for getting outshot. The 1999 Sabres, for example, were a pretty lousy team but they still went to the Finals thanks to Hasek.

  3. I wouldn’t be so sure. The 1984, ’85, ’87, ’88 and ’90 Edmonton Oilers were outshot every single regular season. They had better results in the playoffs, suggesting they turned it up a notch (“it” meaning defensive play), although in ’90 they were outshot in the postseason as well.

    The Oilers were the exception that were much more about shot quality than quantity. Having watched them do it, and succeed, for all those years is one of the reasons why I’ve been something of a foot-dragger on Corsi. I’m starting to come around, although I do not feel it’s the be-all and end-all. Got some research in the works that I’ll post on Copper & Blue some time fairly soon.

  4. Who is claiming Corsi to be the be-all end-all anyway?

    I found that, at player-level the Corsi / Zone Starts combo are fairly instructive. One gives context to the other. But at team level, Zone Starts looks pretty screwy.

  5. BruceM:

    Interesting; I had no idea that was the case. Maybe that explains the organization’s obsession with the mythical “one-shot scorer.”

    Looking now, I see that Jari Kurri had seven consecutive 20.0%+ shooting percentage seasons, which is unheard of now. Messier has three consecutive seasons, Gretzky had six in seven years. Even for the ’80′s, that’s nutty.

  6. Olivier: I don’t think anybody is, although I’d have been very surprised to see a different result than what came up above. Outshooting has a strong connection to outscoring, which in turn has a strong connection to winning games; it’s not the only factor but it’s certainly a significant one.

  7. more than anything, I think scoring and goaltending wins championships.

  8. noskillgill:

    Of course, there’s so much variance in goaltending that it’s difficult to predict which goaltender will be good in whcih playoff year.

  9. Sigh, 2001 :(

    They should have won it, and the Devils went ahead 3-2 only to blow it.

  10. Shots on net is an OK stat, but not perfect.

    Scoring chances is a much better stat, and I wish we had those numbers.

  11. What about the “Error Stat” David? Getting any traction with that?

  12. J-Dub: For 7 straight seasons between 1981-88, Oilers team Sh% was never lower than 15.5%. They were very consistent, between 15.5 and 16.1 every year but one – they shot a lights out 17.1% the year they scored the 446 goals, but it was only about 1% above their own norm. The league shooting percentage (including the Oilers) over the same time ranged from 12.2 to 12.9%. The Oilers led the league in Sh% every year, and annually scored 20% to 32% more goals per shot than the league average. They were truly an exceptional scoring machine, and their work was based in shot quality over quantity. If we had today’s metrics I would bet they have a high ratio of scoring opportunities per shot on goal, but wold also have a high percentage of goals per scoring chance. They really made their shots count like no other NHL team I have ever seen.

    From 1983 – when Sv% started to be tracked – until 1988 they had a PDO# between 1.033 and 1.052 every year, averaging out at 1.043.

    Now it may be they were the one and only, the true “exception that proves the rule”, but at the very least they proved it’s possible to very consistently outscore without outshooting.

Leave a Reply

Your email address will not be published. Required fields are marked *