Alex Edler - Part II

[Bleach Clean]

Corsi works because it takes out the biggest variable (shooting percentage) in tracking performance with "regular" stats. Otherwise... why didn't we just stick with +/-?

Tracking gf/ga brings shooting percentage back in as a variable.

That might make sense over a larger sample... but not 200 minutes.

I misinterpreted "over 200 min". You were referring to that sample as being too low, and I can see the logic in that.

However, conversion rates for and against, are sometimes necessary to add context to shot differential. Just like Corsi added a layer to traditional shooting percentages.

It's in that murky middle ground of treating all shots as the same vs. the growing information about shot quality. Iceberg Slim suspects that Edler has a lower differential against because he's bleeding a higher GA against. I think there's merit to that thinking. Even so, I do understand the point about sample size being too small. So grain of salt.

 

[Proto]

There still isn't all that much data to suggest there's a meaningful difference between shot quality/shot volume, except at the margins. It's certainly possible for specific players, I guess, but it would also be attributing a lot of value to one player on the ice.

The sort of differential that's being suggested by those limited numbers would essentially mean Edler is so bad that you might as well be playing short-handed. That seems unlikely...

 

[Scurr]

I misinterpreted "over 200 min". You were referring to that sample as being too low, and I can see the logic in that.

However, conversion rates for and against, are sometimes necessary to add context to shot differential. Just like Corsi added a layer to traditional shooting percentages.

It's in that murky middle ground of treating all shots as the same vs. the growing information about shot quality. Iceberg Slim suspects that Edler has a lower differential against because he's bleeding a higher GA against. I think there's merit to that thinking. Even so, I do understand the point about sample size being too small. So grain of salt.

There might be an argument to be made against Edler there... but drawing that line to the Edler/Tanev pairing because of a 200 minute sample (which is what IS did) is wrong.

 

[MS]

There still isn't all that much data to suggest there's a meaningful difference between shot quality/shot volume, except at the margins. It's certainly possible for specific players, I guess, but it would also be attributing a lot of value to one player on the ice.

The sort of differential that's being suggested by those limited numbers would essentially mean Edler is so bad that you might as well be playing short-handed. That seems unlikely...

... that's pretty much what it looked like for long stretches last year.

And yeah, not all shots are created equal. Edler was bleeding an absolutely absurd amount of odd-man rushes last year - I don't think it would be entirely out-of-line to say he was causing close to as many as the rest of our defense combined. The fact that he blasts a lot of low-percentage shots into opponent shinpads doesn't fix this.

 

[Bleach Clean]

There might be an argument to be made against Edler there... but drawing that line to the Edler/Tanev pairing because of a 200 minute sample (which is what he did) is wrong.

Fair enough. I can't argue that his sample is sufficient because it isn't. That there "may be an argument" is all I think the post should have conveyed.

I've been guilty of using that info to draw attention to the Edler-Tanev pairing. To move away from thinking that Tanev will automatically help Edler correct his play. It's not statistically significant, but it does draw my attention (mostly because it's all we have). I'm not uber positive about their 'chemistry'. Guess we'll see.

 

[Proto]

Fair enough. I can't argue that his sample is sufficient because it isn't. That there "may be an argument" is all I think the post should have conveyed.

I've been guilty of using that info to draw attention to the Edler-Tanev pairing. To move away from thinking that Tanev will automatically help Edler correct his play. It's not statistically significant, but it does draw my attention (mostly because it's all we have). I'm not uber positive about their 'chemistry'. Guess we'll see.

It's not. There's shot data that suggests they're good together. The sample is also too small, but it's definitely more significant than any chance/goal data imo.

 

[arsmaster*]

... that's pretty much what it looked like for long stretches last year.

And yeah, not all shots are created equal. Edler was bleeding an absolutely absurd amount of odd-man rushes last year - I don't think it would be entirely out-of-line to say he was causing close to as many as the rest of our defense combined. The fact that he blasts a lot of low-percentage shots into opponent shinpads doesn't fix this.

Yeah, Edler hammering shin pads and giving up 3 foot shots against is definitely not reflected in the data, but I'm gonna back away from the math here.

 

[Proto]

... that's pretty much what it looked like for long stretches last year.

And yeah, not all shots are created equal. Edler was bleeding an absolutely absurd amount of odd-man rushes last year - I don't think it would be entirely out-of-line to say he was causing close to as many as the rest of our defense combined. The fact that he blasts a lot of low-percentage shots into opponent shinpads doesn't fix this.

I do think Edler was pretty bad last year (that Edler/Garrison pairing was just brutal), but I can't see anyway they'd be worse with him on the ice than they would be playing 4 v 5 hockey. That just seems ludicrous.

That said, I do think Tortorella's weird system for his D was a nightmare, and Edler seemed to perform the worst in it. Edler is typically not a very good defender in my opinion: he makes slow/poor choices at the opposition blueline once the zone has been gained, has poor shot selection, and plays 2 on 1s worse than anyone I can think of, but he's generally not anywhere near as nightmarishly awful as he was for parts of last year.

I would think a coherent system will at least have Edler looking more typically Edlerian this year. The unfortunate thing is the trade of Garrison and relative dearth of prospects at that position mean that moving Edler (even for a great package) is almost impossibly unlikely at this point.

 

[Bleach Clean]

It's not. There's shot data that suggests they're good together. The sample is also too small, but it's definitely more significant than any chance/goal data imo.

Are you speaking about their 55CF% when on the ice together?

 

[Proto]

Are you speaking about their 55CF% when on the ice together?

Yes. I'm just saying there's not really enough data to draw any conclusions, but if you're going to look for anything, that's probably more likely to be relevant. Usually people want at least 400 minutes with pairings before they start to draw even preliminary conclusions, though...

I think Bieksa and Edler are both actually 50+% CF players with every partner they've ever played 400+ minutes with except for each other. That's why the Edler/Bieksa pairing Tortorella kept going back to during the year was so damned funny.

 

[Bleach Clean]

Yes. I'm just saying there's not really enough data to draw any conclusions, but if you're going to look for anything, that's probably more likely to be relevant. Usually people want at least 400 minutes with pairings before they start to draw even preliminary conclusions, though...

Ok, but that CF% may be 'inflated' in this case due to a higher (highest amongst pairings) GA ratio. So it's hard to take it at face value on it's own.

 

[Ernie]

... that's pretty much what it looked like for long stretches last year.

And yeah, not all shots are created equal. Edler was bleeding an absolutely absurd amount of odd-man rushes last year - I don't think it would be entirely out-of-line to say he was causing close to as many as the rest of our defense combined. The fact that he blasts a lot of low-percentage shots into opponent shinpads doesn't fix this.

Except the numbers don't back this up - not even advanced stats. He didn't "bleed goals" at a considerably higher rate than the rest of the defense. If he was allowing that many odd man rushes, they weren't ending up in the net.

The Canucks just couldn't score when he was on the ice. A valid criticism might be that he was making dumb plays in the offensive zone that stifled offence. But none of the people who seem to be making "eye test" criticisms have made any complaints about that. If the Canucks had scored when he was on the ice at a similar rate to the other defensemen, he would have been about a -13. Not very good, but not the disaster people are making out.

What the numbers do suggest is that he peaks out at about the level of a 3rd or 4th defenseman, even going back several years when he was "good". This was concealed by the fact that he was playing with elite possession and point players in Daniel and Henrik Sedin.

What we saw last year was a result of him being played in a role by Tortorella for which he was not suited, by bad luck, and by him just not being that good of a player in the first place.

He needs to have the Cody Hodgson treatment this season - give him favourable zone starts with the Sedins and first unit PP time, and flip him when his "value" increases.

 

[MS]

Except the numbers don't back this up - not even advanced stats. He didn't "bleed goals" at any considerably higher rate than the rest of the defense. If he was allowing that many odd man rushes, they weren't ending up in the net.

The Canucks just couldn't score when he was on the ice. A valid criticism might be that he was making dumb plays in the offensive zone that stifled offence. But none of the people who seem to be making "eye test" criticisms have made any complaints about that.

What the numbers do suggest is that he peaks out at about the level of a 3rd or 4th defenseman, even going back several years when he was "good". This was concealed by the fact that he was playing with elite possession and point players in Daniel and Henrik Sedin.

What we saw last year was a result of him being played in a role by Tortorella for which he was not suited, by bad luck, and by him just not being that good of a player in the first place.

He needs to have the Cody Hodgson treatment this season - give him favourable zone starts with the Sedins and first unit PP time, and flip him when his "value" increases.

Edler's GA/60 was amongst the few very worst in the NHL, especially amongst players not playing for bottom-5 defensive teams (if memory serves only two teenaged rookies - Jones and Trouba - were worse).

Goals were absolutely flying in with him on the ice.

 

[Ernie]

Edler's GA/60 was amongst the few very worst in the NHL, especially amongst players not playing for bottom-5 defensive teams (if memory serves only two teenaged rookies - Jones and Trouba - were worse).

Goals were absolutely flying in with him on the ice.

The Canucks weren't a very good defensive team at 5v5 - 12th from the bottom. Their overall GA was saved by their top 10 PK.

And there were three forwards on the Canucks who had worse GA/60. Higgins, Kesler, and Sestito.

The only good defensive forwards (according to GA) on the Canucks were the Sedins and Booth.

 

[CanucksSayEh]

The most favorable zone start for Edler, to the Canucks, is the bench. Better yet, the press box.

 

[Scurr]

Fair enough. I can't argue that his sample is sufficient because it isn't. That there "may be an argument" is all I think the post should have conveyed.

I've been guilty of using that info to draw attention to the Edler-Tanev pairing. To move away from thinking that Tanev will automatically help Edler correct his play. It's not statistically significant, but it does draw my attention (mostly because it's all we have). I'm not uber positive about their 'chemistry'. Guess we'll see.

It's not. There's shot data that suggests they're good together. The sample is also too small, but it's definitely more significant than any chance/goal data imo.

Exactly.

Yeah, Edler hammering shin pads and giving up 3 foot shots against is definitely not reflected in the data, but I'm gonna back away from the math here.

They track this data... Edler gets less shots blocked and more on net then anyone else on our blueline iirc.

Ok, but that CF% may be 'inflated' in this case due to a higher (highest amongst pairings) GA ratio. So it's hard to take it at face value on it's own.

You're using the noise in the less reliable data to make an argument that isn't there imo. I can buy Edler being a drain on GA for his partner, but it just doesn't make sense that Tanev would make anyone worse defensively.

 

[arsmaster*]

Do they track how many times his bad reads lead to point blank chances?

 

[Scurr]

Do they track how many times his bad reads lead to point blank chances?

Sort of. A player that makes bad reads that lead to point blank chances... also makes bad reads that lead to easier shots... so they just track shots. They used some fancy charts to convince me it was science.

 

[Addison Rae]

The most favorable zone start for Edler, to the Canucks, is the bench. Better yet, the press box.

Good one dude!

 

[Verviticus]

...he's good

 

[deckercky]

The Canucks just couldn't score when he was on the ice. A valid criticism might be that he was making dumb plays in the offensive zone that stifled offence.

Does this theory even make sense though? If he was making dumb plays in the offensive zone, you'd expect them not to result in shots at all. To make dumb plays which consistently result in team-mates generating low percentage shots seems like something difficult to plan even if trying to do it.

He coughs up chances against the Canucks - accept that about his game. But it's difficult to look objectively at the factors behind his offensive totals dropping last year as anything but bad luck.

 

[Addison Rae]

He's been a stud this year, he hasn't been as good offensively as he was in years past but he's playing some of the best hockey of his life (he was sheltered in his best offensive years, this year he's playing against elite competition)

 

[Scurr]

He's been a stud this year, he hasn't been as good offensively as he was in years past but he's playing some of the best hockey of his life (he was sheltered in his best offensive years, this year he's playing against elite competition)

Yup. I wanted to keep him because I thought he could get back to treading water and putting up some points. This is much, much better.

I still think he puts up 40pts.

 

[bossram]

He's been a stud this year, he hasn't been as good offensively as he was in years past but he's playing some of the best hockey of his life (he was sheltered in his best offensive years, this year he's playing against elite competition)

Edler's good. Not a surprise to people who don't think Edler's mere presence on the ice caused the team to shoot at 3.8% last year, resulting in his -39.

 

[Addison Rae]

Edler's good. Not a surprise to people who don't think Edler's mere presence on the ice caused the team to shoot at 3.8% last year, resulting in his -39.

Yeah, preach. Edler was never as bad as his -39 indicating, though he looks like a completely different player this year compared to last. (the Chris Tanev effect)