Number Chains - June 9, 2012
Robert Vollman
June 09 2012 08:41AM
This new, regular feature on NHLNumbers will share interesting stats-related posts from around the web every week.
Player Usage Charts are here!
2011-12 NHL Player Usage Charts for every NHL team, accompanied by expert insights from a field of 20+ analysts, are now available for download here
Obviously we believe very strongly in the usefulness of these charts. We've already used Player Usage Charts repeatedly on NHL Numbers, like when Josh Lile looked at the Dallas Stars, or when I looked at last year's rookies, or the league's top defensive pairings for example, and very rarely does an article go by without at least some mention of one of its core components (Offensive Zone Starts, Quality of Competition or Relative Corsi). In fact, we might be giving away our secret by providing these ground breaking statistics in such an understandable fashion - we might have to soon rely solely on Wanye's wit for our readership.
Others have picked up on Player Usage Charts since it was unveiled June 1st, including:
- Daniel Wagner covered the Vancouver Canucks over at Pass it to Bulis Vancouver Sun Sports Blogs.
- He also previewed the New Jersey Devils and Los Angeles Kings series over on the score.
- The Tampa Bay Lightning were covered over at Raw Charge.
- Dirk Hoag covered the Nashville Predators over at On the Forecheck
- Graeme Nichols handled the Ottawa Senators over at the 6th Sens (I see Daigle people).
- And Derek Jedamski covered the Buffalo Sabres for the Buffalo Examiner
Several of us contributed to this project, including me, Kent Wilson, Jonathan Willis, Eric Tulsky, Corey Sznajder and Josh Lile, and we almost all use them on a regular basis, so we're quite proud to have unveiled this last Friday. So download the PDF, print it off, grab a cold one, sift through it this week-end, and let us know what you think.
Also unveiled this week was this year's look at zone entries over at Broad Street Hockey. The fine folks over there, including Eric T and Geoff Detweiler, examine every single Philadelphia Flyer zone entry and then conduct some fascinating analysis on the consequences of various strategies. Truly ground-breaking stuff here.
Part 1
Part 2
While on the topic of how teams act differently depending on the score and the impact that it can have, the imcomparable Tyler Dellow over at MC79 hockey looked at the most dangerous lead in hockey.
"The really funny thing about this is that playing conservatively seems to have little in the way of benefits for the team leading. If you calculate the points on the basis of regulation win being worth two points and a regulation tie being worth 1.5, you would expect the teams leading by one to average 1.66 points. Actual average points collected by teams leading by one heading into the third: 1.67. The real beneficiaries are the teams trailing by one heading into the third period. Their expected points per game is 0.56. They actually averaged 0.64 points per game."
Finally, Brodeur is a Fraud took a look at Darryl Sutter's effect on goalies
"Throughout his career, Sutter's goaltenders have routinely been above average. Sutter-led teams have only posted a below-average save percentage in two out of his dozen seasons as an NHL coach, and in every one of the remaining ten his team was at least .006 above the league benchmark in save rate"
We'll end this week with a cheap plug of my own work over at Hockey Abstract where I've collected lots of raw statistical data, including Player Usage Charts, for every single pending Unrestricted Free Agent. Download a copy and figure out who should go where.
Great question!
I generally favour Corsi percentage to showcase the difference between players who are high event and those who are low event, and I'll definitely try that on the Player Usage Charts.
As for the use of Relative Corsi instead of straight-up Corsi (for both the bubbles and the Quality of Competition), it tends to weed out the team effects.
Relative Corsi, after all, is just the team's on-ice Corsi minus the off-ice Corsi. Not only does that help you compare players from team to team, but also players who were with a team for different parts of the same season (when the team was perhaps playing much differently).
Why do you use CorsiRel in the charts instead of straight Corsi/60 or Corsi percentage?
I don't understand how CorsiRel can help compare players from different teams.
If Player A has CorsiOn of 10 and a CorsiOff of 10, while Player B has a CorsiOn of 5 but a CorsiOff of -15, if ZoneStart and QualComp are the same across Player A, Player B and their office teammates, do we really think that Player B is twice as good at puck possession as Player A? How did Player B's teammates while he was off the ice make him a superstar?
Great question.
First of all, straight-up Corsi would have this same problem - and worse!
Player B appears to be making the best of a bad situation, while Player A is offering his team no particular possession advantage over their other players, but straight-up Corsi suggests that Player A is the better possession player.
It's possible that Player A is indeed the better possession player, but he plays for the best team in the league, while Player B is on a team full of stiffs that would make anyone look good. That's partly why we said Relative Corsi helps you compare players across teams, not that it fully allows you to do so.
As with all statistics context is everything and you need to supplement your analysis appropriately.
@Tach
You can't necessarily look at it as Player B's teammates while he was off the ice making him look like a superstar. If a players off-ice Corsi is significantly worse than his on-ice Corsi, is it not a reasonable conjecture that over a large enough sample size, he had something to do with the increased possession numbers with him on the ice?
Off-ice Corsi is a measure of how the team plays without him, on-ice Corsi is a measure of how the team plays with him. As Rob pointed out, Player A in this case seems like a somewhat replaceable asset. He doesn't appear to really be pushing possession.
Pure Corsi numbers are subject to team influences, CorsiRel diminishes that effect to a certain extent.