@SteveBurtch As for the hockeygraphs article, it's logical that qualteam is easier for coaches to control than qualcomp, which is why qualcomp is harder to measure for stats guys, which is why they wrongly assume it has no impact just because they can't measure it as effectively. But if you look over the data, you will find qualcomp is also controllable, so conclusion of that article is parsing the data incorrectly as well.

Looking at the Sharks, Vlasic has the hardest corsi rel qual comp, and also has the highest qualcomp of any Sharks dmen. Braun is 2nd among Sharks dmen. So the Sharks "top shutdown pair" has the hardest qualcomp of any of their defense pairs. Clearly the coach has some control.

Martin and Burns have the second hardest of Sharks defense pairs. Polak and Dillon have the easiest. Just as you would expect if the coach was drawing it up.

Looking at Chicago, Hjalmarsson is at the top, then their top two-way winger Hossa, then Keith, then Toews at #4. You can fiddle with the order, but those are the four names you'd expect if a computer drew it up. At the bottom you have Gustafsson, Ehrhoff, Roszival, players the coach was clearly trying to shelter.

Now I'm checking LA. Doughty has the hardest, then Kopitar. At the bottom are Dowd, Mcbain, Greene, Scuderi, Forbort, etc.

See a pattern? My guess it he hockey-graph writer never looked on a micro level like this. He probably did as too many stats writers do and grouped 700 players together, didn't see any obvious patterns from 30,000 feet above the ground, and then assumed there weren't any. But when you look from ground level, you will see the patterns that you can't see from a macro view.

Reply · Report Post