Kukla's Korner

The Puck Stops Here

Greg Zanon Has The Worst Corsi Rating

In yesterday’s sabermetrics and hockey post, I listed the worst 20 players in 2010/11 by their unadjusted Corsi rating.  These are the players who were on for the highest difference between attempted shots (shots on goal, missed shots and blocked shots) for their opponents and their own team in 5 on 5 situations.

The player with the worst Corsi rating was Greg Zanon of the Minnesota Wild.  He had a -429 Corsi rating.  This is slightly over five extra shots taken per game by Zanon’s opponents. 

Zanon is quite typical for a low Corsi player.  He is not a player who drives puck possession.  Last season he scored 7 points (all assists) in 82 games.

He was second in ice time on the Wild behind Brent Burns and played a shutdown defenceman role.  Burns frequently started shifts in his defensive zone making his opponents more likely to take shots than his team.  His Minnesota team is one of the weakest teams in terms of puck possession in the NHL.  This all sets up to make Zanon the worst player by Corsi rating in the 2010/11 season. 

Zanon is clearly not a good puck possession player but this is not a sign that he is the worst player in the league.  Adjusting his rating for team and zone starts clearly shows this.  For Zanon to be a successful player he must be able to reduce opponent’s scoring chances significantly because he does not produce them.  The fact that he does not play against tougher than average opposition does not help his case. 

Greg Zanon is a poor puck possession player.  His team (Minnesota) and his high number of defensive zone starts combine to make his Corsi (which would likely not be good in any situation) the worst in the league.  This makes Zanon a defensive specialist who does not have tremendous value to his team.

Filed in: | The Puck Stops Here | Permalink
 

Comments

Avatar

Wow, you should just stop trying to talk about sports and players you have no clue about. Have you ever even watch Zanon play? The guy is a monster, and has saved a number of games single handedly with his play. This article was a waste of everyone’s time with your nonsense and complete lack intelligence on the subject at hand.

Posted by HansonJJ from MN on 07/15/11 at 03:40 PM ET

PuckStopsHere's avatar

Yes i have watched Zanon play many times. 

Your comment is typical of a fan who feels the need to defend his player because somebody correctly points out a significant flaw in one of their players.  Your comment was a waste of everyone’s time with your nonsense and complete lack intelligence on the subject at hand.

Seriously, do you want to address what was written or just rant because you can’t accept that Zanon is bad at puck possession and plays on a team and in a role that makes his puck possession numbers look even worse?

Posted by PuckStopsHere on 07/15/11 at 03:47 PM ET

Avatar

I agree with just about everything EXCEPT the last nine words of the piece.

Again, I think this is part of the danger of putting too much stock in Corsi. I do not think you have provided enough context to make the claim that Zanon “does not have tremendous value to his team.”

Despite being a poor Corsi player, Zanon was fourth among full-time Wild players (I define this as having played at least 40 NHL games over the course of the season) in terms of goals allowed per 60 minutes of ice time. In over 1,500 minutes at even strength (the equivalent of over 25 full games), the Wild only surrendered 52 goals while Zanon was on the ice.  Zanon was also third among full-time Wild players in terms of his goaltender’s save percentage while he was on the ice.

Zanon also led the Wild by a wide margin in blocked shots. Blocked shots are defensively valuable, but count against a player’s Corsi because Corsi attempts to measure balance of puck possession. He also had relatively few giveaways considering his TOI.

While all of these additional stats are, like Corsi, subject to fair amount of error and interpretation, I believe they paint a general picture of Zanon as a defenseman who does not control the puck much, but is very effective at limiting the quality of his opponents’ shots on net and preventing goals. This picture is consistent with what I have seen and heard about Zanon’s overall game.

Again, I think two of Corsi’s major failings are that it assumes that all shots are equal (they aren’t), and that—when given too much weight—they suggest that a given player’s value is tied very closely to his puck possession ability (which isn’t true in all cases).

Greg Zanon’s job isn’t to possess the puck and create chances for his team. His job is to prevent goals against in critical defensive situations by limiting the quality of his opponent’s attempts on goal (and often, once possession swings the other way, to clear the puck and get off the ice, so that a better puck possessor may take his place).

Other advanced metrics support the suggestion that he does this job very well. Which, in my mind, makes him a very valuable player, even though he is not very skilled with the puck.

Posted by Sven22 from Grand Rapids on 07/15/11 at 04:38 PM ET

J.J. from Kansas's avatar

Zanon also led the Wild by a wide margin in blocked shots. Blocked shots are defensively valuable, but count against a player’s Corsi because Corsi attempts to measure balance of puck possession.

I would be interested to see where Zanon’s Fenwick stacks up in this case, as it’s a very good argument for a defensive defenseman’s case that a blocked shot is a benefit he provides rather than an event which should count as a negative.

Without having to dig up game-by-game logs, does anybody have a good source for this information?

Posted by J.J. from Kansas on 07/15/11 at 04:49 PM ET

PuckStopsHere's avatar

I think people are far too quick to give a pass to players who have no offensive value and their defensive value is not enough to make up for it.  If it were the reverse where a player had no defensive value and their offensive value was not enough to make up for it, people would be much quicker to accept criticism of the player in question.  The two situations are essentially the same - just mirror images of one another.  I think the problem is that coaches would rather pick the defensive player because it appears safe - he won’t make an unforced error that winds up on the highlight reel.

Zanon provides no offence.  He does provide defence, but its not enough to make up for his no offence.  He is exactly the kind of guy who will get a lot of playing time on a weak team that plays a defensive system and he is a reason they are a weak team.  Good teams are giving that playing time to better defencemen than him.

As for the blocked shots defence.  Blocked shots correlate well with losing hockey.  The reason is clear.  In order to block a lot of shots you must not have the puck.  A blocked shot can often turn into a dangerous opportunity as there is a player down blocking the shot and it often rebounds off in some random direction.  Many good shot blockers never wind up near the bottom of any Corsi lists.  Zanon does and he has been doing so since his Nashville days (though he has never finished as low on the list as this year).

Posted by PuckStopsHere on 07/15/11 at 04:52 PM ET

PuckStopsHere's avatar

JJ

I have the numbers that I could run to get Zanon’s Fenwick rating.  I am a bit too lazy to do this properly (at least right now), but it is clear looking at them that he is well into the -300’s when blocked shots (for and against) are removed, which doesn’t do much to improve his standing.

Posted by PuckStopsHere on 07/15/11 at 05:04 PM ET

J.J. from Kansas's avatar

That’s a series of good points, especially “Good teams are giving that playing time to better defencemen than him.”

It’s true; Greg Zanon would not be getting the 2nd-most ice time on a better team.  That is quite easy to accept.

The semantic difference is that you said he “does not have tremendous value to his team.”

All things considered, it does seem that Zanon has a lot of value for a bad team.

Many good shot blockers never wind up near the bottom of any Corsi lists

If blocking shots correlates well with losing hockey and bad Corsi correlates well with losing hockey, then you are crediting what is mostly a team effect to a player category as it pertains to Zanon. I believe you would call this “pretzel logic.”

Also, I’d like to see what statistics have to say about exactly how “often” a blocked shot can turn into a dangerous opportunity.  Is there any resource out there that has this information or is it simply anecdotal?  I mean, if the puck often rebounds in a random direction, then how often can it randomly become a dangerous opportunity?  Logic would seem to indicate that fewer than half of those “random” bounces would end up in the attacking team’s possession, since the defending team would more often have a numerical manpower advantage thanks to the presence of the goaltender.

By that progression, since the definition of “often” is fairly flexible, I would say that a blocked shot can often turn into a dangerous opportunity for the team who blocked the shot as well.

Posted by J.J. from Kansas on 07/15/11 at 05:17 PM ET

J.J. from Kansas's avatar

JJ

I have the numbers that I could run to get Zanon’s Fenwick rating.  I am a bit too lazy to do this properly (at least right now), but it is clear looking at them that he is well into the -300’s when blocked shots (for and against) are removed, which doesn’t do much to improve his standing.

Posted by PuckStopsHere on 07/15/11 at 03:04 PM ET

Fair enough.  If we don’t have proper league-wide context, it likely wouldn’t be too valuable anyway.

Bottom line is that I don’t have a problem saying Greg Zanon isn’t the kind of guy I would not want on my top defensive pairing (or even my top two pairings) and that’s pretty much what I gathered was the overall gist of the post.

Posted by J.J. from Kansas on 07/15/11 at 05:19 PM ET

Avatar

Where are you getting that he faced weak competition? Corsi Rel QoC has his competition 2nd among Wild D (behind Schultz, who’s a complete beast), and coupled with that zone start he doesn’t look too different from guys like Brewer, Ohlund, Weaver, and Polak, other similar guys who are made to play mostly defense with little offense at 5v5. It’s playing to his comparative advantage, which makes perfect sense for (more-or-less) one-dimensional players.

A guy who can move the puck up the ice better would be more valuable, sure, but as long as the Wild aren’t asking Zanon to play offense, everything should be okay.

Sven: in the 4 years of data behind the net has, no player has shown a significantly above or below-average ability to drive either shooting or save percentage. That is, at the season level, the quality of the shots seems to come very close to evening out.

JJ: BtN says with Zanon on ice the Wild were blocking ~16 shots per 60, with opponents blocking ~11.5 shots per 60 (5on5).

Posted by Ralph on 07/15/11 at 05:24 PM ET

Avatar

Blocked shots correlate well with losing hockey.

I don’t know about this. Last year 6 of the top 10 shblk teams were playoff teams, including the cup winner. The year before is much the same, also including the cup winner.

Just because the Islanders block a lot of shots and lose doesn’t mean shot blocking and losses have a strong positive correlation.

I understand the reasoning behind saying a shot-blocking player or strategy is inferior to one that prevents the shot in the first place through recovery and transition to attack (no matter how well the shot-blocking might be executed), but the existence and historical efficacy of strategies like a passive box or a shot-blocking shell (which is sort of a 5 on 5 passive box) suggest that’s not always the case. These configurations produce less possession, lower corsi and fewer chances, but they also produce more odd-man breaks, which are far easier to convert on than a random shot from the wall.

Posted by steviesteve on 07/15/11 at 05:43 PM ET

Avatar

All of the stats - corsi, fenwick, shots, goals - can be found at stats.hockeyanalysis.com.

The fact that he does not play against tougher than average opposition does not help his case.

But actually, he does.  Over the past 3 seasons Zanon’s opponents on average scored at a rate of 0.782 goals per 20 minutes of ice time which ranks 13th of 129 defensemen with at least 2500 minutes of 5v5 ice time.  Source:  http://stats.hockeyanalysis.com/ratings.php?disp=1&db=200811&sit=5v5&pos=defense&minutes=2500&type=goal&sort=OPPF20&sortdir=DESC

Despite when he was on the ice opponents were held to just 0.721 goals per 20 minutes, good for 32nd of 129.  http://stats.hockeyanalysis.com/ratings.php?disp=1&db=200811&sit=5v5&pos=defense&minutes=2500&type=goal&sort=A20&sortdir=ASC

Yes, he is offensively challenges, but that isn’t his role.  He is a defensive specialist and a good one at that.  Every team can use a guy (or two) like Zanon.  Don’t get too concerned about his corsi ratings because that isn’t his job.  His job is to stop the other team from scoring and he does that well.

Posted by HockeyAnalysis on 07/15/11 at 06:04 PM ET

Avatar

got to agree that zanon is far better than his corsi. nice stat pulling.

Posted by callmedrw from detroit on 07/15/11 at 06:20 PM ET

PuckStopsHere's avatar

Ralph: Where are you getting that he faced weak competition? Corsi Rel QoC has his competition 2nd among Wild D (behind Schultz, who’s a complete beast), and coupled with that zone start he doesn’t look too different from guys like Brewer, Ohlund, Weaver, and Polak, other similar guys who are made to play mostly defense with little offense at 5v5. It’s playing to his comparative advantage, which makes perfect sense for (more-or-less) one-dimensional players.


PSH:  When talking about quality of competition, I usually am tsalking about the QualComp category on behind the net where Zanon is a -0.008.  Thats third among defencemen on the Wild behind Schultz and Burns.  I tend not to look at Corsi based quality of competition numbers because they implicitly fold in zone start numbers with them.  Zanon frequently starts in his own zone.  In that we know he plays a defensive role.  When he starts shifts in his own zone he starts then against players who start shifts in the offensive zone.  These offensive zone starts show up in the Corsi ratings and thus in Corsi quality of competition numbers,  QualComp decouples the zone starts from the quality of opposing players better.  Zanon’s Corsi related quality of opposition numbers are high due to his zone starts more than the actual quality of players he plays against, which is quite average.

Posted by PuckStopsHere on 07/15/11 at 08:31 PM ET

Avatar

These articles are AWFUL.
You are supposed to be investigating whether or not these methods of statistical analysis have any value; instead, you are giving a flat-out explanation of the numbers without the critical reflection.
It reads a bit like: “Bad teams get scored on often”.
Could you try to get to the part where you find some stats that actually illuminate a situation?
Or perhaps, at the end of the dryest article I have ever read, you say “and that’s why I give these numbers no credence”.
Or perhaps, you apply it to the AHL so you are not stating the obvious (Minnesota Wild players have a poor plus/minus, no matter how you mess with the numbers), and maybe give us some players to watch for…...

Posted by tuxedoTshirt on 07/15/11 at 08:35 PM ET

PuckStopsHere's avatar

Tuxedo

That has been done by me over the years and by many others as well.  Instead of complaining this post doesn’t have what you are looking for, look around in my archives and other places and you will find plenty of what you are looking for.

This post is a look at the player with the worst Corsi, why he has it and an attempt to give his approximate value to an NHL team.

Posted by PuckStopsHere on 07/15/11 at 09:38 PM ET

cainer4wingsglory's avatar

You don’t need Corsi numbers to know Greg Zanon sucks. The simple eye test suffices. Nice try though TPSH…

Posted by cainer4wingsglory on 07/15/11 at 11:30 PM ET

Avatar

QualComp doesn’t decouple zone start numbers. Moreover, you’re assuming that Zanon is playing against players who get as privileged zone starts as his are difficult, which is partly true but not entirely, as the vast majority of players are between 45 and 55.

Posted by Ralph on 07/16/11 at 12:44 AM ET

Avatar

,We do realize, moreover, that Corsi Rel QoC is exactly the same as QualComp, except in that it’s based on Corsi Rel instead of relative plus minus aka rating, right? It’s a common point of confusion.

Posted by Ralph on 07/16/11 at 12:47 AM ET

Avatar

I should explain my reasoning. You say the opposition gets better Corsi with better zone start. Better Corsi should lead to more shots and more goals, and I wouldn’t be surprised if top offensive players like the Sedins see an uptick in shot quality (there is a weak positive correlation between zone start and shooting percentage).

Posted by Ralph on 07/16/11 at 12:56 AM ET

PuckStopsHere's avatar

+/-doesn’t show the same sensitivity to zone starts that Corsi does.  Evidence is to compare the worst Corsis with the players who take the most defensive zone starts - offensive ones and then do the same with +/-.

Your guess about the Sedins is hard to accept given that Daniel is 6th in the league and Henrik 16th in their unadjusted Corsis.  They are however good examples of players Zanon plays against.  They are in the same division and have a lot of offensive zone starts.  I bet the Sedins are some of the most common opponents (if not the most common) that Zanon faced this season.

So the +/- based quaity of competition numbers better decouple zone starts.  Thus I view them as more “orthogonal” to zone starts, while Corsi numbers are more strongly mixed with them.

Posted by PuckStopsHere on 07/16/11 at 01:12 AM ET

Avatar

Are you sure that’s not just because of noise? On top of that, on the grand scale, the competition metrics all do okay, even though all too frequently in single cases they get it all wrong, RATING included.

Moreover, we’re not comparing Corsis—we’re comparing Corsi Rel QoC, aka QualComp based on the opponents’ relative Corsi ratings (instead of relative plus-minus).

For consistency’s sake, it seems better to me to use plus-minus instead of Corsi differentials if you’re going to use QualComp instead of Corsi QoC or Corsi Rel QoC.

Here I made a mistake as well: since you’re using raw Corsi, I should be looking at Corsi QoC, not Corsi Rel QoC, at least in my mind. On the other hand, Corsi Rel QoC does a much better job meshing with what we see. I’m not sure what to make of that.

What about the Sedins is hard to accept? 6th and 16th in unadjusted Corsi, 1 and 2 (ignoring Cam Janssen) in zone start ratio, 1-2 in total offensive zone faceoffs among forwards (at least, in offensive zone wins, with no one else close), and 4-5 in GFON/60. It’s rare that a player sees more than 5% of his ice time against another single player on a different team over the course of a season (assuming health, and that BtN ran its numbers correctly).

At the bottom of the relative plus-minus rankings (min 40 GP), there are plenty of high defensive zone start players: Carcillo, Powe, Betts, Marchant, Glass, McClement, Spaling among others.

Just having run a basic correlation between defensive zone starts and relative plus-minus versus defensive zone starts and Corsi versus defensive zone starts and relative Corsi (10-11 season, min 40 GP, too lazy to compile all four seasons’ worth of data), the r squared values go .002, .016, .007. So RATING actually correlates better with defensive zone starts (the raw number, not ratio) than either Corsi metric.

Posted by Ralph on 07/16/11 at 03:24 AM ET

Avatar

If you say you have seen Zanon play many times, I suggest you get your eyes checked. He may not be the best player on the ice, and he isn’t getting 50 goals a season, but that isn’t his job, his job is to stop the other team from scoring and he does it well, so how can you say he has no value to his team? That makes absolutely no sense and just shows your lack of knowledge on the player. If you are looking at numbers I can find some type of numbers to make any player look bad pretty much, they don’t mean anything, because they are only relative to what was going on at the time. If a player plays on the same line all year and scores 40 goals and you switch one of those players out the next year and he only scores 10 goals it does not mean he got worse, it means his line got worse, where are the numbers to prove otherwise?
Also if there is a line change and a player barely even has his 2nd skate on the ice yet and the opposing team scores guess what, that player gets a -1. How can you look at a players +/- knowing stupid things like this. (and this is just an example, there are others that prove most numbers are pointless.)
Again I am not saying he is the puck handler on the ice, but it’s kind of hard to get puck possession ratings when you are purposely stopping a puck with your face every time the other team is taking a shot. I wonder how many more games the Wild would have lost had he not been blocking that many shots, so to say that blocking shots correlates with losing games makes no sense either. Maybe you just can’t accept that you had nothing to write about so you decided to take a player with some random low rating and make a story about it before even watching and knowing the player…

Posted by Hansonjj from MN on 07/19/11 at 03:44 PM 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