Stu wonders: "Face-off specifics. This is one of the NHL’s best areas of stats-keeping, but it can get even better. It would be helpful to learn where the face-offs are won and lost, because one taken in the neutral zone is not as crucial as one in the defensive zone. Adding won-lost records for defensive and offensive zone draws for teams and individuals would be a great addition — and so would face-off situations during special teams play. And just to make face-off figures even more complete, let’s compile won-lost match-up records that show how well centers do when they’re matched against other centers. If Anze Kopitar takes 12 draws in a game against Jonathan Toews, you can find that in the game faceoff report, but I’d like to know how many each won in the times they faced off during the season or their even their careers."
Once again, we already have this data:
One of Vic Ferrari's greatest contributions to NHL analysis was noticing how much coaches tend to rig their players' offensive and defensive zone starts (Alain Vigneault is going for a record with Manny Malhotra this season). What we haven't seen is any particular talent for taking defensive ones as opposed to offensive ones - face-off-winning talent is zone-agnostic. We also just don't see centers feasting on non-centers in the faceoff dot - centermen face a wide range of opposing centers and there's a very small spread of average opponent faceoff skill relative to league average.
As regular readers of this blog will gather, I am also not a fan of small sample size head-to-head matchups. They are completely overrun with luck and have zero future predictive value - it's not as though Anze Kopitar "owning" Jonathan Toews has any meaning the next time they play each other.
Verdict: the NHL should keep this stat, but for the purpose of showing whether players start their shifts in the offensive of defensive zone. Head-to-head matchups and the notion of defensive vs offensive zone faceoff skills are dead ends.