To shed some light on these issues, I've used play-by-play data to separate ice time into three categories:
Offensive - the most recent faceoff was in the attacking zone and no player from that faceoff has left the ice.
Defensive - the most recent faceoff was in the Oilers' defensive zone, all players from it remain on the ice.
Neutral - the most recent faceoff was in the neutral zone and/or a change has taken place.
The Oilers have changed their lines around a lot more than the Canucks, for example, so it's not as simple to break down. I've decided to look at Shawn Horcoff, Eric Belanger and Ryan Nugent-Hopkins individually and group all the time together where none of those were on the ice. Much of that was with Gagner or Lander in the middle. Hockey is messy with more fluid positions than other sports, but this should give us a good idea who is doing what in each zone. On top of the usual teammate and opponent effects, we are dealing with small sample sizes. Try to take the exact numbers with a grain of salt and focus on the patterns. Here is a table with the Corsi rate in each type of ice time:
|None of Above||-48.4||-4.9||30.2|
jumps off the page. There's no nice way to put it; the can't-legally-drink-on-most-road-trips line has been getting thrashed after faceoffs in their own zone. By Corsi rate they have been quite bad in offensive-zone starts as well. In the neutral situations, that is after neutral-zone faceoffs or an on-the-fly change, RNH has been pretty close to average. This pattern points to faceoffs being a big liability and that is backed up by faceoff win% - Nugent-Hopkins is the worst Oiler center, winning only 37.7% of his faceoffs. Flip that for Belanger, one of the best in the league on the dot. Going by Corsi rate, he has been the best Oiler center for starts at both ends of the ice, but not as good in neutral situations.
To get an idea who should get o-zone starts, I like to look at the difference between Corsi rates following o-zone starts and those in the defensive zone. That will tell you which players are most affected by zone starts. There are matchup issues and so forth, but those for whom o-zone starts are the most important, whether that's because they are very good in the offensive zone or very bad in the defensive zone, usually should get the easier ice time. Here are those numbers:
|Center||Ozone - Dzone|
|None of Above||78.6|
These data fit the usual narratives - the young guns and fourth liners should get the protected minutes. Renney has been doing this. Personally, I wouldn't recommend giving Nugent-Hopkins the Sedin treatment because it seems bad for development to give young players so little defensive responsibility, but if all you cared about was winning right now then there is pretty strong evidence even with these small sample sizes that he rarely should start in his own end.