Kukla's Korner

The Puck Stops Here

Worst 20 Raw Corsi Ratings

As I continue my summer posts on sabermetrics and hockey, today I will look at the worst 20 players in 2010/11 by their raw (unadjusted) Corsi rating.  Here at the top 20 players; today is the flip side.

This group of players are generally not well known hockey players.  Many play tough defensive roles on weaker teams and thus have the situation in which they play as an explanation for their poor rating.  Those players who do not play tough roles are some of the weaker players who played regularly in 2010/11.

Here are the worst 20 players by unadjusted Corsi rating in 2010/11:

2010/11 Worst 20 Raw Corsi Ratings
Rank  
Player   
Team   
Corsi 
1Greg ZanonMin

-429

2Cam FowlerAna

-381

3Zenon KonopkaNYI

-318

4Eric NystromMin

-300

5Todd MarchantAna

-288

6Brandon McMillanAna

-276

7Luca SbisaAna

-270

7Saku KoivuAna

-270

9Paul MaraMon

-257

10John MaddenMin

-256

11Joffrey LupulTor

-245

11Matt MartinNYI

-245

13Brent SopelMon

-239

14Andreas LiljaAna

-235

15Ryan JonesEdm

-234

16Keith AulieTor

-233

17Theo PeckhamEdm

-225

18Matt CullenMin

-224

19Kevin KleinNas

-219

20Tim BrentTor

-216



As with the top 20 players, there are several teams represented multiple times.  There are six members of the Anaheim Ducks (Fowler, Marchant, McMillan, Sbisa, Koivu and Lilja), four members of the Minnesota Wild (Zanon, Nystrom, Madden and Cullen), three members of the Toronto Maple Leafs (Lupul, Aulie and Brent), two members of the New York Islanders (Konopka and Martin), two members of the Edmonton Oilers (Jones and Peckham), two members of the Montreal Canadiens (Mara and Sopel) and one Nashville Predator (Klein).  Clearly team effects account for a significant portion in many of the players who appear here.

It is worth noting that Greg Zanon’s league worst rank of -429 and Cam Fowler’s -381 are well worse than any positive Corsi in the league - Ryan Kesler led the league with +337.

In the future I will apply some adjustments to this list in an effort to better isolate individual skill and take a look at some of the more interesting cases.

Filed in: | The Puck Stops Here | Permalink
 

Comments

Da lil Guy's avatar

Interesting that Konopka rates so high -

He was brought in to Ottawa in part due to his faceoff skills, and people have been talking about how he will be useful to the Sens as they try to play a better puck possession game.

This abysmal Corsi number would seem to suggest puck possession is not a strong point.

Between 10-20% of his ice time was short handed, though.

Posted by Da lil Guy from Guelph, Ontario on 07/14/11 at 01:35 PM ET

Avatar

Zenon wins faceoffs, not puck possession. These are unadjusted Corsi so they don’t factor in defensive role.

Posted by hockey1919 from mid-atlantic on 07/14/11 at 03:51 PM ET

PuckStopsHere's avatar

Winning a faceoff is a way to gain puck possession. 

It is true that Konopka starts a lot of shifts in his defensive zone, when that is adjusted for, he doesn’t appear this far down a worst list.

Posted by PuckStopsHere on 07/14/11 at 03:59 PM ET

Avatar

Where are these numbers coming from? I can only find up-to-date Corsi lists that take a player’s average per 60 minutes, not ones that show the raw totals.

Posted by Sven22 from Grand Rapids on 07/14/11 at 04:02 PM ET

PuckStopsHere's avatar

These numbers come from the same place behind the net gets their numbers (which are reported per 60 minutes).  They come from some work gathering data from the game statistics on nhl.com.

You could probably recreate most of these numbers from the behind the net numbers if you multiply the per 60 minute numbers by games played and ice time.

Posted by PuckStopsHere on 07/14/11 at 04:06 PM ET

Avatar

I’m aware that I could calculate a raw number for a given player based on his ice time / games played. I’m was just wondering if there’s any place that conveniently lists these raw numbers as such—alas, it appears there is not, if I understand you correctly.

It would be unfortunate (and some what ironic) if the only way to get “unadjusted” Corsi numbers, aside from manually counting every attempted shout oneself, is to “d-adjust” the average numbers posted elsewhere.

Did you have to derive these numbers yourself?

Posted by Sven22 from Grand Rapids on 07/14/11 at 04:21 PM ET

PuckStopsHere's avatar

I got these numbers from the NHL game statistics and collected them with a script that I have written.  I checked a few cases with the behind the net numbers to see that I wasn’t making an error.

I don’t know of any convenient online sources that do this for you.  I don’t think I am likely to put in the time to manufacture one myself (unless there was some clear payoff for doing so).

So the short answer is that there isn’t an easy online source.

Posted by PuckStopsHere on 07/14/11 at 04:27 PM ET

Avatar

I realize that you are a true believer, but this stuff is just plain nonsense.  It is an attempt to quantify an abstract.  If this Corsi number were even remotely valid, there would be no need to hire and pay pro scouts.  These numbers could be used to evaluate potential draft choices and get rid of the amateur scouts too.

Think about it for a while.  Compare the results that have been listed with the reality of what your eyes see.  The numbers do not correspond, do they. 

When I took basic statistics a long time ago in college, we used a book entitled “How to Lie with Statistics” as our main text.  Ever since, I have had a healthy dose of skepticism regarding manipulating numbers to prove an abstract.  It cannot be done.

I would much rather trust the judgement of a Kirk Maltby or a Pat Verbeek than a page of numbers.

Doc

Posted by DocF from Reidsville, NC on 07/14/11 at 04:37 PM ET

PuckStopsHere's avatar

Doc

You couoldn’t be more wrong.

I will merely quote a responce to the simkilar tomment you gave yesterday that was given by Ralph:

Doc: the top Corsi team going back lost in the finals, won, lost in the finals, and won. Puck possession wins games. Detroit, Chicago, Vancouver, San Jose, Pittsburgh, Philadelphia, and Washington are the best possession teams there are. Corsi correlated very strongly to zone time numbers which the NHL published for a few years and correlates about r^2=.55 to points. Shot domination nowadays is the single most important sustainable skill to winning.

This correlates strongly with what your eyes see.  You have offered nothing but a factless rant to say otherwise.

Posted by PuckStopsHere on 07/14/11 at 05:39 PM ET

Avatar

To be fair, PSH, I think you and Doc are arguing about slightly different things.

You say the top Corsi team is likely to be successful, which I think is a non-controversial statement. If you control more of the puck and generate a greater volume of shots than your opponent, consistently, you’re likely to win a lot of games.

Doc seems to be arguing that Corsi is less valuable as a measure of individual talent or contribution. Again, I think this is not a terribly controversial statement, but I concede that I put much less faith in individual on-ice Corsi numbers than you do, and WAY less faith in team-adjusted Corsi numbers.

I have not seen any convincing evidence to suggest that, even within the confines of a single team, Corsi is a reliable measure of who the best players are. If that were the case, one would have to conclude that Jonathan Ericsson was much better than Nicklas Lidstrom in 2010-11, a ridiculous claim. The TEAM may have had a better shot differential when Ericsson was on the ice compared to Lidstrom, but any observer would be able to see it had little to do with their respective levels of ability.

I also generally distrust team adjustments as a means of comparing players across teams, not only for the reasons above, but also because I think it gives an unfair advantage to talented offensive players on top-heavy teams that have many Corsi-poor players. For example, if Ryan Getzlaf (who has a negative raw Corsi, but a very high relative Corsi due to the ineptitude of the bottom half of Anaheim’s roster) played with the Vancouver Canucks, his raw Corsi would likely be MUCH higher, but his relative Corsi would likely be lower, as Vancouver would be a good puck possession team with or without Getzlaf on the roster or on the ice.

Posted by Sven22 from Grand Rapids on 07/14/11 at 06:57 PM ET

PuckStopsHere's avatar

Sven

I think you are giving Doc too much credit.  I am not convinced he knows what Corsi is.  He hasn’t convinced me yet.  He seems to be arguing that numbers cannot capture anything in hockey - at least that is my interpretation.  He hasn’t provided any facts to go with his rant.  He has us guessing at what his rant means - and we have different interpretations - he may not know the difference himself.

Nobody is claiming that Corsi is the be all and end all of statistics (at least not that I am aware of).  Nobody claims that about goals or points either (at least nobody serious).  What I claim is that Corsi, points, goals etc are useful and valuable to evaluate players.  The more of these we understand - both the absolute number and the reasons why the number is what it is - the better we can evaluate hockey players.

If puck possession is important on a team level, then those players who drive it are important.  On the reverse side of this, those who consistently have the puck possessed by their opponents are not doing very well.  There are reasons on both sides that we can get a player at the top or bottom of the rankings, where that ranking is not due to the player so much as the way he is used and we can do our best to correct for those and understand them.

Your Lidstrom/Ericsson example shows a very clear thing that must be taken into account when making comparisons - quality of opposition.  We know Lidstrom played against the best of Detroit’s opponents and Ericsson certainly didn’t.  You example is no better than saying goals isn’t much of a statistic because Blake Comeau scored more than Henrik Sedin last year and we all know Henrik is the better player.  It proves nothing except that we need more than one number to evaluate a player.

Posted by PuckStopsHere on 07/14/11 at 08:06 PM ET

Avatar

Yep, anyone who knows what Corsi is isn’t about to use raw Corsi to argue Player X > Player Y. Context is extremely important.

And before it comes up, PSH is simply presenting raw Corsi because it’s interesting, not because it proves anything about any single player.

Posted by Ralph on 07/14/11 at 10:22 PM ET

Avatar

To clarify my face-off versus puck possession. True he wins faceoffs and that is puck possession. What he doesn’t do is actually retrieve the puck from the other team when the puck is in play.

Posted by hockey1919 from mid-atlantic on 07/15/11 at 11:15 AM ET

Avatar

All right, PSH, I think I can get behind most of what you’ve said. I agree that Corsi can be a useful tool—although always evaluated in context, and ideally accompanied with data from the “eyeball test.”

I do agree that if team possession wins games, and I do agree that Corsi is likely the best metric we have to measure which players are contributing to that puck control—although I don’t think it does this nearly as well as we would like it to, and I think it’s probably less reliable for defensemen than it is for forwards.

I think Corsi would be more valuable if it were reported not just as a raw or averaged differential, but in three segments—shots attempted for, shots attempted against, and shots attempted differential.

That, I believe, would give us more context with which to judge certain players who might be key contributors in running their team’s offense, but might also be relied on to play in tough defensive minutes in situations where their goal is not to prevent shots against per se, but to limit the quality of those chances in order to prevent goals against.

Posted by Sven22 from Grand Rapids on 07/15/11 at 12:47 PM ET

Avatar

Sven: worth noting that no player in the 4 years so far has really separated himself in terms of on-ice save percentage or on-ice shooting percentage (odd fact: the leader in the former category is George Parros, he of the famous mustache). The best defensive players still don’t really allow the other team much, even if they don’t generate anything offensively themselves.

Posted by Ralph on 07/16/11 at 02:30 AM ET

Add a Comment

Please limit embedded image or media size to 575 pixels wide.

Add your own avatar by joining Kukla's Korner, or logging in and uploading one in your member control panel.

Captchas bug you? Join KK or log in and you won't have to bother.

Smileys

Notify me of follow-up comments?

Feed

Most Recent Blog Posts

About The Puck Stops Here

imageThe Puck Stops Here was founded during the 2004/05 lockout as a place to rant about hockey. The original site contains over 1000 posts, some of which were also published on FoxSports.com.

Who am I? A diehard hockey fan.

Why am I blogging? I want to.

Why are you reading it? ???

Email: y2kfhl@hotmail.com