rulururu
Two dudes blogging and podcasting about the San Jose Sharks, straight from sunny California.

post Stats, Glorious Stats

March 31st, 2011, 10:13 am

Filed under: blog — Written by Mike

It’s been a while since I wrote a stats-related post, and I figured I’d write one to piggyback on an interesting post on the Behind the Net blog (not to be confused with Behind the Net, the numbers site).  A while back I did a purge on the RSS feeds I read, and for some stupid reason, this blog was amongst the casualties.  I really must have been in a slash-and-burn mood that day, because it’s really one of the best hockey blogs out there if you are statistically inclined. What I want to do today is highlight some of the stats talked about in the linked post, and who on the Sharks are the best at those categories.

First of all, some real quick and dirty explanations of some of the stats referenced there.  I would recommend reading more about them, but not everyone can spend hours reading about advanced hockey metrics.  And as Doug would say, why would you want to?

  • GVT – Goals Versus Threshold.  A complicated stat that tries to create one number for the value of a player, measured in goals in a season versus the value of a replacement-level player.  Similar to VORP in baseball.
  • Rating – a BTN stat that is the difference between your team’s +/- per 60 minutes when you are on the ice versus when you are off the ice.  Unlike the regular +/- stat, it helps level the playing field for those on bad teams.
  • QualComp – quality of competition.   The weighted average of the Rating of the players you face on the ice.
  • Corsi – a +/- stat that counts shots instead of goals.  Actually, it counts all pucks directed towards the net, including missed and blocked shots.  Unfortunately, it’s similar to the +/- stat in that players on good teams generally have better ones.  Of the 28 players that have played a game for the Sharks this season, only 10 have negative Corsi, and most (Moore, Mashinter, Desjardins, McLaren, Wingells) aren’t regulars.
  • Corsi Rel – The difference in your Corsi when you’re on the ice versus off.
  • Corsi Rel QoC – Quality of Competition calculated not by +/- per 60, but Corsi Rel.
  • Zone Starts – the percentage of shift-starting faceoffs being in the offensive zone.

If you’re still reading, pat yourself on the back, because that’s a load of math.  Let’s highlight the different Sharks players leading the categories in the stats that the LOES highlighted, in the order that I think is most important. The following is all 5v5 stats, and I’m not including anyone that’s played fewer than 10 games.

Corsi Rel – Kyle Wellwood – 14.6

It’s surprising, and doubtless related to a red-hot Joe Pavelski and clicking third line since he arrived.  Still, Wellwood leads the team in a stat I believe is miles better than +/-.  One downside to Corsi Rel is that time-on-ice isn’t factored in, and it should be noted Wellwood has averaged only 13.07 minutes of even-strength ice time per game, good enough for 15th on the Sharks.  For this reason, it’s worth mentioning the second place player, Ryane Clowe (14.1), who’s averaging more than two minutes more 5v5 ice time, and who I might argue is the team’s MVP.  Top Corsi Rel among defensemen? Jason Demers (8.6).

QualComp – Patrick Marleau – 0.101

Marleau is way out in front on this stat, with the second place Joe Thornton at 0.087.  Despite the fact that Marleau tends to play the wing more now, traditionally not as defensively important as center, he’s  the go-to guy when the other team’s top line is on the ice.  Top defenseman – Dan Boyle (0.062).

Corsi Rel QoC – Patrick Marleau – 0.885

I’m not sure why the LOES like Corsi so much yet mention QualComp instead CorsiRel QoC.  If Corsi is better than +/-, then Corsi Rel QoC is better than QualComp.  Maybe that’s what they meant.  Anyway, unsurprisingly, Mareau leads again, but there’s a bit of shifting under him.  Jumbo drops to 5th on the team, and Joe Pavelski (0.747) moves up to 2nd.  Boyle moves up to 3rd.

Zone Starts – Scott Nichol – 39.4

This means when Nichol took a faceoff to start a shift, 60% of the time it was in the defensive zone.  That’s a lot of trust from the coaching staff, and certainly related to the fact that Nichol is the best faceoff guy on the team. Like the last stat, it’s a way of measuring how sheltered a guy is.  It’s been calculated that you give up about 0.25 shots every time you take a faceoff in the defensive zone, so this is why Nichol’s Corsi isn’t so good.  With that in mind, it’s unsurprising that Marc-Edouard Vlasic (46.8) has the lowest zone start percentage among defensemen.

Time on Ice – Dan Boyle – 19.13

No doubt Boyle is the workhorse, and even strength is no exception.  He also plays the most PP and ES time.  Contrast this to the Ducks (for instance), with Vish leading the category, but if you look for #2, you see that Toni Lydman and Cam Fowler play about the same amount.  However, Fowler plays almost no PK, and three and a half minutes per game on the PP.  Lydman is the opposite, almost no PP time, but is way out in front of PK time. Certainly important when trying to evaluate a player.

I didn’t include GVT here because there isn’t a day-by-day calculation of GVT that I know of, and to be honest, GVT makes a lot of assumptions about the weights of various measures that I don’t necessarily agree with.  I won’t go so far as to say the attempt to create one stat that measures everything is a fool’s errand, but I feel like I get a better picture of a player when I look at several stats, and not just one.

Just a note for tonight- Jamie Benn and Alex Gologoski lead the Stars in Corsi Rel, so watch out for those guys.

Sorry, the comment form is closed at this time.

Sorry, the comment form is closed at this time.

ruldrurd