Hockey Stats Turning the Corner?
From time to time I’ve discussed the need for greater detail in hockey stats analysis. With the realities of the new CBA forcing teams to give a critical eye to dollar value of production, some teams are actually starting to work on this. Even better, they are using some of the concepts I originally developed several years ago. Here’s an article this week on the topic… During his time as GM of the Calgary Flames, one stat (Craig) Button especially liked was player production per-minute.Interestingly enough, Button is using a stat I helped develop five years ago. Alan Ryder is going to be starting a hockey stats column soon in the Globe & Mail (congrats Alan!) A few months ago Ryder authored a piece about the Pythagorean Equation, though he didn't break it down as far as I did - to the indivdual player. In any case, it looks like hockey stats analysis (hockeymetrics) is about to turn the corner. Maybe Chris Apple and I should shake the dust off our old columns and metrics and see if we can’t get back into this game. |
Comments on "Hockey Stats Turning the Corner?"
I'm not a numbers guy, but I do have a latent skepticism that you can get stats to support whatever it is you happen to believe.
The Sedins, for example, might do well with the points-per-minute stat in the Vancouver system with physical winger on their line. However, the stat geeks for a dump-and-chase team like, say, San José still can't compare their roster's stats to the Sedins with any validity. The twins never have operated in a dump-and-chase offense and would likely have their talents wasted on a team like the Sharks.
Baseball leads the world in these sort of 'red-headed stepchild' statistics. Hockey pushes the envelope enough with its 'plus-minus' figures. I just think there are too many variables for similar stats to be a reliable indicator of performance.