Rob Vollman of Hockey Prospectus ("and friends", as the title puts it) recently released player usage charts (warning: big PDF) for the 2011-2012 NHL season. This thing is chock-full of information that's fun to look at due to the friendly graphical presentation.
Basically, it's set up as follows, everything here being for even strength 5 on 5:
- The y-axis is Quality of Competition (QComp). The higher the Qcomp value, the tougher the competition that player saw during his time on ice.
- The x-axis is the percentage of times that player began his shift in the offensive zone. The higher the number, the less the coach exposed that player to defensively rigorous situations.
- The size and color of the bubble indicates the player's Corsi value, which indicates whether the player was outshot or not. Blue is positive Corsi, while white is negative Corsi, and the size of the bubble indicates how big the number is. For example, a big clear bubble means the player was outshot badly while on the ice, and a big blue bubble means the player dominated his competition.
- Forwards' names are black while defensemen's names are blue.
Some analysis after the break.
- Jordan Staal was nearly unique in the NHL in his usage pattern: a forward with less than 50% zone start and a Qcomp of about 1.5 who came out ahead. The only other player I can find with that kind of profile is Joe Pavelski of San Jose. Not only does Disco Dan trust Staal with a ridiculous job; Gronk's getting it done and scoring while he's at it. If there really is a trade in the works, Shero could ask for the moon and get it.
- Pascal Dupuis maybe shouldn't have been one of the finalists for the PensBurgh version of the Selke Trophy. Oops. It's not that he did badly, of course, but he wasn't exactly the even-strength beast he appeared to be, either. He saw tough competition and played them basically to a stalemate, but there was no dominance.
- Simon Despres was all kinds of sheltered. Not only did he start nearly 65% of his shifts in the offensive zone, he saw weak competition in the process. If you're looking for a silver lining, it's that he performed well against that competition. Sure, he might be ready for a regular shift in the NHL at this point in his career, but it is a bit optimistic to assume that he could step into a top-4 role.
- Looking at this chart makes it seem like Brooks Orpik played the whole season alongside Zbynek Michalek and got their butts handed to them while Kris Letang and Paul Martin fared well against slightly weaker competition. Given that's kind of what my "eye test" told me, I'm inclined to believe it, but I get the feeling that this will be quite a controversial data point.
- Eric Tangradi was bad. Really just not any good at all.
- Matt Niskanen, Ben Lovejoy, and Deryk Engelland all played basically the same role. Engelland came out way behind the other two. Perhaps it's time for that to change.
- As expected, James Neal and Evgeni Malkin saw pretty sheltered minutes. Dan Bylsma clearly used Jordan Staal to allow those two to get mismatches. That's pretty sweet. What's not sweet is that it's harder to force that sort of thing on the road
- All the defensemen played sheltered minutes compared to Staal and Dupuis regardless of results.
- Craig Adams and Joe Vitale aren't the defensive stalwarts people thought they might be. They both faced weak competition and their results still suffered.
Let's keep something in mind here: while all of this data is useful and interesting, we cannot forget that most of this data reflects usage, not talent or ability. For example, if Bylsma decided to use Vitale to play big minutes against the other team's top line in exclusively defensive zone starts, he could do that. Vitale's usage stats, specifically Qcomp and zone start would change accordingly. His bubble, then, would be in a different spot and through no fault of his own. However, one would expect his bubble to stay white and expand to indicate that he struggled horribly against that kind of usage.
So, bubble placement comes from the coaching staff, but bubble size and color come from the player's ability.