3 Responses to “Finally…Tanguay”

1. Olivier says:

   August 31, 2009 at 6:12 pm

   I came here trough your latest piece at puckprospectus.

   Tanguay is “soft”, so he is underrated; dud is in the top 10 in ESP since the lock-out, always played north of 14.5 min at ES since 2001, is 29 years-old, but Carbo insisted on playing him 10-11 minutes at ES a night (so he wouldn’t eat away Higgins and Latendresse and whomever minutes I suppose). Oh, and he’s as good on the LW and the RW.

   The word best apt to describe the way the habs handled Tanguay last year is “clusterfuck”. I find it somewhat depressing that he’ll easily outscore Gionta at half the price…

   Anyway, ESTR is an interesting stat. I always have the same problem with rate stats based on GF/GA when it comes to position players: the variations in S%; the habs overall S% at ES last year was .922, but D’Agostini had a .895 on-ice s%, tagging him with 10 extra GA. So there you have a guy with a pretty good Corsi, who obviously was sheltered when it comes to defensive zone draws, but gets murdered on ratings based on GF/GA.

   Now, I’m perfectly willing to accept, even without demonstration, that a guy can have such an effect on his team ES Sv%, but I look at Laraque’s .933 on-ice sv% (as horrendous defensively and sheltered a player as can be) and I can’t help but have a doubt about the validity of on-ice/off-ice goal-differential as a basis for rating individual players. There seems to be a randomness about it that just doesn’t compute for me.

   I used to think Corsi would thus be a better basis for such a rating, but it’s just a gut feeling and I don’t trust it.

   Of course, I didn’t invent anything. I’m just half assedly referring to the work put down by guys like JLikens at objective NHL and Ferrari and others.

   I’d be very curious to know what you think about that kind of stuff…
2. Timo Seppa says:

   August 31, 2009 at 9:30 pm

   Excellent thoughts, O. I do appreciate the feedback. Hopefully, by adjusting equitably for other players on the ice, ESTR will succeed where those other stats failed you. As you know the Habs well, tell me how these ratings sit with you (ESOR/ESDR/ESTR):

   LW Alex Tanguay +1.26/+0.19/+1.45 (1st on MTL)
   RW Matt D’Agostini -0.29/-0.84/-1.13
   RW Georges Laraque -1.61/+0.09/-1.51

   You can see that ESTR does agree on the poor defensive assessment of D’Agostini, but it considers Laraque essentially average. Keep in mind that Tanguay and D’Agostini logged less than 600 minutes of ESTOI each, enough to have some confidence in the ratings, but I’d feel better with higher ESTOI. Laraque’s 242 minutes are a pretty small sample. I wouldn’t be surprised to see that ESDR dip, per your perception, over a larger sample.

   By the way, your new right winger Brian Gionta came in at +0.45/-0.01/+0.44 last season with the Devils.
3. Olivier says:

   September 1, 2009 at 12:36 pm

   The more I think about it (paralysis by analysis! paralysis by analysis!), the more I simply think it’s a matter of cautions; not many guys get bitten by rotten on ice s% each year (on the Devils, ony Colin White really looked bad on that aspect, being nearly 20 points under the team overall score), but some of them do as some of them get the dice rolling their way (Holik on the Devils, Lapierre on the Habs…).

   Is that a skill? I’m not convinced, and if it’s not a skill, then on each teams you will have a few players who are over/under rated by the metric due to the defensive side of the puck, even tough they meet an acceptable threshold ov ESTOI. Say you get 1.5 plyers on each side of the puck per team, well that’s 90 guys whose ratings are off…

   Altough it really does seem to even out once the sample size gets big enough (say, 800 or 900 minutes on ice…), a solid string of good/bad luck will carry a long way.

   I guess my point is, overall, the ratings looks pretty darn sound to me and having a couple of guys fall in the cracks is inevitable; I remain skeptical because said cracks, absent a demonstration of a given player’s ability to influence on-ice Save%, seems to be pretty darn wide sometimes. Here are the 10 guys who a) played more than 40 games ranked by difference between their on-ice and team overall ESSv%:

   Name Team OnIceSv% TeSv% Dif
   BOOGAARD,DEREK MIN 0,976 0,920 0,057
   BERGLUND,PATRIK STL 0,958 0,904 0,054
   ROY,ANDRE CGY 0,956 0,903 0,053
   JONES,RYAN NSH 0,962 0,913 0,048
   FROGREN,JONAS TOR 0,937 0,892 0,045
   EAVES,PATRICK CAR 0,958 0,915 0,044
   BUTLER,CHRIS BUF 0,956 0,913 0,043
   ELLIS,MATT BUF 0,955 0,913 0,042
   PARRISH,MARK DAL 0,935 0,894 0,041
   IVANANS,RAITIS L.A 0,944 0,905 0,038
   PERRON,DAVID STL 0,942 0,904 0,038
   WILLSIE,BRIAN COL 0,937 0,899 0,037
   LEBDA,BRETT DET 0,939 0,903 0,035
   MCAMMOND,DEAN OTT 0,940 0,905 0,035
   OSHIE,T.J. STL 0,939 0,904 0,035

   Now, that’s a whole bunch of bottom feeders, but also there is this guy at #2 whose name seems familiar… And you keep going down, and a few others blues pop up quickly; wonder what went on over there? Did they have such a crappy bottom 6?

   And you look at the bottom of the list:

   HEDICAN,BRET ANA 0,887 0,916 -0,029
   D’AGOSTINI,MATT MTL 0,889 0,915 -0,026
   ANTROPOV,NIK TOR 0,870 0,892 -0,022
   MCDONALD,ANDY STL 0,882 0,904 -0,022
   BOUILLON,FRANCIS MTL 0,895 0,915 -0,020
   BRUNNSTROM,FABIAN DAL 0,875 0,894 -0,019
   HINOTE,DAN STL 0,885 0,904 -0,019
   STORTINI,ZACHERY EDM 0,900 0,919 -0,019
   HANSEN,JANNIK VAN 0,899 0,918 -0,019
   LISIN,ENVER PHX 0,895 0,912 -0,018
   METHOT,MARC CBJ 0,893 0,911 -0,017
   KOZLOV,VIKTOR WSH 0,896 0,911 -0,016
   STASTNY,PAUL COL 0,884 0,899 -0,015
   NISKANEN,MATT DAL 0,879 0,894 -0,015
   MOSS,DAVID CGY 0,888 0,903 -0,015

   Sorry for the crappy format; this time, Lang (35) and AKostitsyn (37) are somewhere nearby… Maybe there is a team effect of some sort, I don’t know. But Berglund’s presence at the top of your ranking tells me your method isn’t completely immunes to these variations.

   Amazingly interesting conversation.