Exceeding Expectations: Top Corsi Overachievers

One of the best ways to evaluate a player's performance is to count the amount of time his team and his opponets had the puck while he was on the ice.  Nobody collects that data, but counting the total number of shots for and against while the player is on the ice is a good proxy for it.  This shot differential has come to be known as the "Corsi number", after its presumed originator, Buffalo Sabres goaltending coach Jim Corsi.  (We don't actually know who invented it, and Jim's not taking credit...)

One thing that's clear is that you can't use a Corsi number without context.  If a player plays against soft competition and gets a lot of offensive zone draws, then he'll have a better Corsi than someone who starts in his own end against the other team's top line.  One thing we can do is build a simple regression model for what a player's Corsi should be given his strength of competition and what faceoffs he lines up for.  Here are the top 30 in the league in terms of exceeding their expected Corsi, all at 5v5:

It should be clear that the whole can be greater than the sum of the parts.  Nashville has clearly found a couple of forward lines that are incredibly successful together - there's no guarantee that these players would do as well individually with other pairings.  But separating the individual contributions of players who line up together almost all of the time is very difficult.

For completeness, here are the bottom 30 of the league:

No big surprises here, but I would draw your attention to the number of Colorado Avalanche players on the list.  The Avs have been fighting the percentages all season: they give up a ton of shots at 5-on-5 and get outshot, but they've had amazing goaltending and serious luck in the other team's end.  Since their 14th game, they've essentially been even on goal differential, hardly what you'd expect from a team that could easily finish the season with 100 points.