:format(webp)/cdn.vox-cdn.com/uploads/chorus_image/image/25460535/20131207_krj_sv7_147.0.jpg)
Puck possession really isn't everything... but it is a large chunk of the pie (the largest in fact). A team that scores a lot but gets out-shot and out-chanced will still end up losing more games than they win, unless they have superb or lucky goaltending. This is why you can predict a team or player's success by Corsi% better than any other currently known metric (and can even improve those metrics with minor adjustments). The only problem with individual player evaluation is that it is difficult to separate who is driving the boat.
Tyler Dellow (of mc79hockey) wrote a response to AIH's own Tim Bonnar's piece The Real Problem with the Winnipeg Jets. He used a table to show how the Jets have been performing this season, while comparing top players to the bottom players. Of course, there is the problem with small sample sizes.
So, let's expand to all seasons as Jets...
Fig.1 2011-14 Winnipeg Jets forwards 5v5 Corsi% with defensemen
* click data to expand; WOWY % is how often a player is above 50% when together; avg Corsi% is average of all teammates numbers with (not weighted average); all >= 50% cells are red, all bellow are blue and white is for insufficient data
Fig 2. WOWY% and Avg Corsi%
| WOWY% | Avg Corsi% | WOWY% | Avg Corsi% | |||
| Andrew Ladd | 85.7% | 53.2% | Tobias Enstrom | 68.4% | 51.8% | |
| Kyle Wellwood | 83.3% | 53.8% | Dustin Byfuglien | 52.6% | 50.6% | |
| Alexander Burmistrov | 83.3% | 53.4% | Zach Bogosian | 52.6% | 48.7% | |
| Blake Wheeler | 71.4% | 51.6% | Grant Clitsome | 36.8% | 49.2% | |
| Eric Tangradi | 66.7% | 53.5% | Ron Hainsey | 33.3% | 47.1% | |
| Bryan Little | 57.1% | 51.4% | Mark Stuart | 15.8% | 45.9% | |
| Olli Jokinen | 57.1% | 49.5% | Paul Postma | 83.3% | 52.0% | |
| Devin Setoguchi | 50.0% | 51.5% | Keaton Ellerby | 25.0% | 46.6% | |
| Nik Antropov | 50.0% | 49.9% | ||||
| Evander Kane | 33.3% | 50.6% | ||||
| Tim Stapleton | 33.3% | 48.6% | ||||
| Chris Thorburn | 33.3% | 44.5% | ||||
| James Wright | 16.7% | 46.6% | ||||
| Michael Frolik | 16.7% | 44.8% | ||||
| Matt Halischuk | 16.7% | 44.6% | ||||
| Antti Miettinen | 0.0% | 46.8% | ||||
| Mark Scheifele | 0.0% | 46.1% | ||||
| Jim Slater | 0.0% | 42.8% | ||||
| Tanner Glass | 0.0% | 41.8% |
Thoughts
* Context of usage is missing. There are no score effects, line matches, quality of teammate or zone start adjustments. No taking in Hainsey's tough zone starts or high QoC. No adjustments for Postma's easy zone starts. No adjustments for Frolik, Scheifele and Halischuk being joined at the hip by Noel. Etcetera.
* The players are placed in order of WOWY% and then Avg Corsi% for tie breakers (exception: Postma and Ellerby).
* Postma's and Ellerby's sample spaces were extremely small, so they are left on the side more for curiosity sake.
* The pattern is as expected, with more red in top left, and more blue in bottom right.
* Remember, it's best for a player to promote possession and score; however, having one is far better than none.
What are your thoughts? What patterns do you find interesting? Let us know in the comments!
