Is it possible to affect quality of shots against and mitigate the effects of Corsi?

[SnowblindNYR]

So this is the case of defenseman Kevin Klein on the Rangers. The advanced stats proponents on the Rangers board don't like him because he has bad Corsi numbers. The rest say "watch him" and things like "he's steady in his zone". The assumption made by the latter guys is that he may spend more time in his zone but he makes up for it by being positionally sound and preventing quality shots against. In fact, that's the #1 argument I've seen against Corsi. It's the "quality of shots".

Anyway, for my money the best poster on the Rangers board, 31, started a thread showing save percentage for Lundqvist while playing behind all of the defensemen from either last season or the season before. The save % was completely random. The guys that everyone considers the best like McDonagh and Staal were not necessarily the best in terms of Lundqvist save %. Eminger, if I remember correctly had a really high save % when he was on the ice (though he didn't play all season, maybe he played in a particularly good stretch). Unless luck was a huge issue, the guys that had a high save % should correlate with the worst chances against and vice versa. The randomness REALLY struck me. Maybe the better defensemen looked bad because they played against the best players, I guess quality of competition is a factor not looked into, but we've been a top team in the league against the best players for years and McDonagh, Girardi, and Staal have done well in that role for years. In general those numbers were all over the place. But on the other hand, I just can't fathom certain defensemen not being better than others in keeping shots to the outside, just watching games seems to point that this is not the case. So can players affect quality of shots against to a significant degree and if so is that a factor not accounted for in Corsi/Fenwick?

Click to read more...

 

[wgknestrick]

I would think there is a talent to be found there, but no means to currently track it.

Largely, shot quality is tied to shot location.

Shot location FOR is tracked here for every player:
http://www.sportingcharts.com/nhl/icetrack/?r1Id=296786&r1Year=2013&r1eventType=%27SHOT%27%2C+%27GOAL%27&r1type=1&r1strength=&r2Id=268144&r2Year=2013&r2eventType=%27SHOT%27%2C+%27GOAL%27&r2type=1&r2strength=

But they don't track shot locations AGAINST.

If you can adjust for goaltender SV%, you could use on ice SV% to help, but the spread of defensmen is about the same as NHL goaltenders. This indicates to me that there isn't much of a difference.

Why not just jump to the chase and look at GA/20min (it takes BOTH important factors in, is tracked, and easily sortable)
http://stats.hockeyanalysis.com/rat...1500&teamid=0&type=goals&sort=A20&sortdir=ASC

 

[wgknestrick]

http://stats.hockeyanalysis.com/rat...400&teamid=20&type=goals&sort=A20&sortdir=ASC


You don't have much of a case for Klein IMO. He's just not very good (if I had to make a stat based judgement).

 

[SnowblindNYR]

I would think there is a talent to be found there, but no means to currently track it.

Largely, shot quality is tied to shot location.

Shot location FOR is tracked here for every player:
http://www.sportingcharts.com/nhl/icetrack/?r1Id=296786&r1Year=2013&r1eventType=%27SHOT%27%2C+%27GOAL%27&r1type=1&r1strength=&r2Id=268144&r2Year=2013&r2eventType=%27SHOT%27%2C+%27GOAL%27&r2type=1&r2strength=

But they don't track shot locations AGAINST.

If you can adjust for goaltender SV%, you could use on ice SV% to help, but the spread of defensmen is about the same as NHL goaltenders. This indicates to me that there isn't much of a difference.

Why not just jump to the chase and look at GA/20min (it takes BOTH important factors in, is tracked, and easily sortable)
http://stats.hockeyanalysis.com/rat...1500&teamid=0&type=goals&sort=A20&sortdir=ASC

I didn't look at GA/20, because that takes into consideration how long you're in your own zone. So a defenseman with a high Corsi if he's as adapt at keeping shots to the outside would have fewer goals go in against him than one with low Corsi. I would be interested if you were to just take the defensive zone and just analyze play in that zone. Thanks though. Looks like you can't really track that yet.

 

[SnowblindNYR]

http://stats.hockeyanalysis.com/rat...400&teamid=20&type=goals&sort=A20&sortdir=ASC


You don't have much of a case for Klein IMO. He's just not very good (if I had to make a stat based judgement).

I wasn't trying to make the case though, I was open minded. It's interesting, he gets the easiest minutes with John Moore but still allows the 2nd most goals per 60 minutes. That also shows how much better Stralman is than him.

 

[SnowblindNYR]

Does the "Hard" actually say that John Moore is the Rangers' best defensive defenseman? That's laughable.

 

[KaseMeOutside]

all corsi is, to me, is 1 part of measuring puck possession, but i think if teams have really low shooting %'s but high corsi's that they are then taking alot of low % shots, and are not working the higher quality areas enough

 

[SnowblindNYR]

all corsi is, to me, is 1 part of measuring puck possession, but i think if teams have really low shooting %'s but high corsi's that they are then taking alot of low % shots, and are not working the higher quality areas enough

I also think some teams are just better at finishing than others. This is not proof of anything and just one example but if you saw game 4 of the Rangers-Flyers series McDonagh gets a great chance on Mason, Mason makes a great save, but who has a better shot at finishing that, McD (who is a great well rounded D-man don't get me wrong) or Karlsson or even Weber? The Rangers are the type of team you're talking about, they get their chances but have no guys who can really consistently finish. Nash and MSL are the best at that but outside of that no one.

 

[Talks to Goalposts]

Does the "Hard" actually say that John Moore is the Rangers' best defensive defenseman? That's laughable.

I don't think anyone but the guy who came up with it use HARD or the others like it, the correction factors for difficulty of icetime don't seem to be nearly strong enough to work in a sensible manner.