Saturday, July 11, 2009

Leveraged Icetime Leaders

I'll introduce the details behind leverage later, but for now I'll just give a short explanation. I'm going to define the leverage of a game state - that is, goal differential and time remaining in the game - as the value in Wins of scoring an additional goal minus the cost in Wins of allowing an additional goal. A simple example:

A game is tied at 19:59 of the third period. If your team scores, it will get two points in the standings = 1 Win. If your team is scored on, it will get no points. So the leverage in this situation is 1 Win, which is the highest possible leverage. Using historical data, we can calculate the leverage for every game state.

The average leverage during 2008-09 was 0.329. The top 25 players in leverage (defined as sum(leverage*shift length)/total toi) are shown below, minimum 600 minutes TOI:

Player Pos Lev. %TOI
ANTTIPIHLSTROM LW 0.3616 12.3
ANDREWFERENCE D 0.3607 20.3
BRANDONCROMBEEN RW 0.3604 20.9
SAMUELPAHLSSON C 0.3604 23.8
JEFFWOYWITKA D 0.3596 24.1
MIKEWEAVER D 0.3595 20.1
BRADWINCHESTER RW 0.3587 15.8
BARRETJACKMAN D 0.3576 38.3
ROBNIEDERMAYER C 0.3567 24.6
JAYMCCLEMENT C 0.3563 27.3
MILANMICHALEK RW 0.3560 28.5
MANNYMALHOTRA C 0.3551 27.8
TRAVISMOEN LW 0.3551 25.0
ZDENOCHARA D 0.3551 41.7
MIKEGREEN D 0.3537 34.8
JANHEJDA D 0.3536 36.6
JAYMCKEE D 0.3529 23.9
TIMGLEASON D 0.3519 29.0
SCOTTGOMEZ C 0.3518 31.9
P. J.AXELSSON C 0.3517 24.6
BLAIRBETTS C 0.3516 17.0
JERREDSMITHSON C 0.3512 23.0
DAVIDBACKES RW 0.3510 29.2
JOELWARD RW 0.3508 25.4
BRETTFESTERLING D 0.3508 13.5


We've already identified many of these players as ones who get tough icetime, either by Quality of Competition metrics or through their percentage of "late-and-close" icetime. But some, like Mike Green, didn't make either of those lists despite being on the ice in a lot of tough situations. Not every method is perfect!

Here's the bottom 25:

Player Pos Lev. %TOI
THOMASPOCK D 0.2818 15.2
BORISVALABIK D 0.2899 15.5
CHRISBUTLER D 0.2920 15.8
PHILIPPEBOUCHER D 0.2951 16.6
MICHAELRUPP LW 0.2974 12.8
ZACHBOGOSIAN D 0.2989 16.9
JEFFFINGER D 0.2991 27.1
RICHPEVERLEY C 0.2995 21.2
ANDYGREENE D 0.3001 16.1
JOHNMITCHELL C 0.3002 21.1
NATHANOYSTRICK D 0.3003 16.8
KYLETURRIS C 0.3008 16.5
DOUGWEIGHT C 0.3009 19.4
BOBBYHOLIK C 0.3009 12.8
KEITHYANDLE D 0.3010 23.0
BRIANROLSTON C 0.3011 19.5
NICLASHAVELID D 0.3011 32.4
ARRONASHAM RW 0.3016 13.6
JORDANLEOPOLD D 0.3018 31.4
TIMCONNOLLY C 0.3018 18.3
ILYAKOVALCHUK RW 0.3019 34.5
DAVIDHALE D 0.3023 14.7
OLLIJOKINEN C 0.3023 28.9
NIKOLAIKULEMIN LW 0.3024 20.3
NIKLASHAGMAN LW 0.3025 22.2

Clearly some of these players, including the large number of Toronto and Atlanta players, make this list because their teams were so bad that there weren't any high-leverage situations to play in at all. This method is probably not very useful at the bottom end, but hopefully it gives some insight into who got the toughest minutes on the best teams.

Labels:


Comments:
Interesting stuff.

Did the Blues play in a lot of tight games, or does Murray just run his bench in a very particular manner? Or both?
 
They had 37 one-goal or tied games, plus they gave up nine empty-net goals. The Bruins were 38 and 5 - so it doesn't seem like St. Louis was out-of-the-ordinary.
 
Post a Comment



Links to this post:

Create a Link



<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]