This is Part 4 in an as-of-yet-unknown-number-of-parts series on advanced statistics in the NHL - here are Part 1, Part 2 and Part 3. So far, we've considered the impact of the quality of a player's competition on his performance, but we haven't considered the significance of territory in this analysis. We no longer have zone possession data by team, but at some point, Vic Ferrari popularized the notion of using faceoffs as a proxy for territorial control.
In other words, there are NHL players who go over the boards only for offensive zone draws or when it's clear that the other team is penned in in their end, and conversely, there are players who take all the tough defensive zone faceoffs late in the game. Coaches know who they want out on the ice in a given situation, and that determines the territorial advantage or disadvantage that a player starts out with before he racks up any statistics.
There's no question that this is important: a lost defensive zone faceoff costs a team approximately 0.25 shots on goal. For a guy like Kurt Sauer, who was out for 239 more defensive zone draws than offensive ones last year, that's incredibly costly - perhaps 3 to 5 goals versus an even split. Sauer, who was on the ice for 39 goals for at 5v5 and 43 goals against last year, could be a +1 instead of a -4. With six goals corresponding to approximately one win, that has a huge impact on a defenseman's apparent value. Indeed, the correlation between which zone a player starts in and the number of shots for and against his team while he's on the ice are very highly-correlated - Tyler Dellow again brings us the analysis of Corsi Number vs Zone Starts for defensemen.
Here are the Offensive Zone Start leaders for 2007-08 and 2008-09. You can see that Vincent Lecavalier gets to go out for a lot of offensive draws - 62% in 07-08 and 60% in 08-09 - while Jerred Smithson has to do a lot of the dirty work in Nashville (36.8% and 33.4% the last two seasons.) Again, keep in mind that these are only the roles that their respective coaches have chosen for them. It's possible that they could be used differently by other coaches and be more effective as a result.