NHL 2014 Off-Season Thread

[DrJones]

Re: NHL 2014 Off-Season Thread

Glenn Healy and P.J. Stock have been hired by Sportsnet. God help us all.

[Kobalt]

Re: NHL 2014 Off-Season Thread

That joke has been made every summer for years and you use it in the summer in which it isn't applicable? lol

With all the advanced stat guys the Leafs have brought in this year, I think they'll give Shane O'Brien a contract. He was the Avs best possession player in his last season with the team. http://stats.hockeyanalysis.com/rati...p&sortdir=DESC

Spoiler

Another reason why Corsi/Fenwick is a horrible stat

[AC]

Re: NHL 2014 Off-Season Thread

With all the advanced stat guys the Leafs have brought in this year, I think they'll give Shane O'Brien a contract. He was the Avs best possession player in his last season with the team. http://stats.hockeyanalysis.com/rati...p&sortdir=DESC

Spoiler

Another reason why Corsi/Fenwick is a horrible stat

One player does not make or break a stat

[Kobalt]

Re: NHL 2014 Off-Season Thread

One player does not make or break a stat

There are a lot of examples to why Corsi is a terrible stat, not just using Shane O'Brien as an example. There are some people that use past year's corsi to predict how the team will do in the future years, when there's no correlation between the two at all. It's hilarious.

Leafs will start signing guys like Benoit Pouliots (5yr 20M) of the world and when they don't have the stats to justify those contact, they'll be calling for his head. They already have one of those players in David Clarkson, who lead the Devils in Corsi in 12-13. I look forward to see how the Leafs will be in couple of years.

[AC]

Re: NHL 2014 Off-Season Thread

There are a lot of examples to why Corsi is a terrible stat, not just using Shane O'Brien as an example. There are some people that use past year's corsi to predict how the team will do in the future years, when there's no correlation between the two at all. It's hilarious.

Leafs will start signing guys like Benoit Pouliots (5yr 20M) of the world and when they don't have the stats to justify those contact, they'll be calling for his head. They already have one of those players in David Clarkson, who lead the Devils in Corsi in 12-13. I look forward to see how the Leafs will be in couple of years.

The ideology of the stat (and more or less the calculation) is to out shoot your opponents and that good things will come of that. I'm sorry, I just don't see how that's a broken line of thought.

[DrJones]

Re: NHL 2014 Off-Season Thread

With all the advanced stat guys the Leafs have brought in this year, I think they'll give Shane O'Brien a contract. He was the Avs best possession player in his last season with the team. http://stats.hockeyanalysis.com/rati...p&sortdir=DESC

Spoiler

Another reason why Corsi/Fenwick is a horrible stat

28-game sample of the Roxy's favourite son, who ranked 9th in 5-on-5 TOI among Avs' D-man in a shortened season? Cherry-pick much?

This wouldn't have anything to do with advanced stats types thinking your Avs will regress, does it?

[Kobalt]

Re: NHL 2014 Off-Season Thread

The ideology of the stat (and more or less the calculation) is to out shoot your opponents and that good things will come of that. I'm sorry, I just don't see how that's a broken line of thought.

Do you consider batting average a good stat in baseball? A single and a home run is the same in BA, just like shooting from the blue line with no screen at all, and shooting it in the slot is the same in Corsi/fenwick. It's a flawed stat, only time it's "useful" when there's context to it (like looking at linemates/quality of competition/off&def zone %/etc.) but not a lot of people do. People like to talk about Leafs' freefall in the 2nd half(post-olympic break? Can't quite remember the time) because they were a terrible Corsi team. Canucks on the other hand had great Corsi and still had their freefall.

Also using my favorite team, Joe Sacco's Avs had a solid Corsi because all they did was fire it to the net once they got into the offensive zone from the blueline, those teams weren't very good.

Once there are zone entries, zone exits, zone time, actual time of possession, etc. Advance stats in NHL are flawed. I'm sure some teams already have people working in their front office that keeps track of this.

I'm interested to see these stats when it's over. http://www.tsn.ca/nhl/story/?id=459985

28-game sample of the Roxy's favourite son, who ranked 9th in 5-on-5 TOI among Avs' D-man in a shortened season? Cherry-pick much?

This wouldn't have anything to do with advanced stats types thinking your Avs will regress, does it?

Not sure what that has to do with this. They went from 2nd to last in the NHL to 2nd best in West, I'll be shocked if there isn't some type of regression. Devils were a top 2/3 Corsi team in the past 2 years, I'm sure their front office were happy to miss the playoffs since they were a great Corsi team.

Edit: Also, Avs were supposed to regress since November last year, same with Anaheim because of their high PDO, they regressed all the way to 1st place in the Western Conference! It doesn't bother me when people say Avs will regress.

[AC]

Re: NHL 2014 Off-Season Thread

IIRC there's been plenty of work done that depicts shot quality as a non-skill. Could be wrong. There's also been work done that shows Corsi has a strong link to winning. I think you're also underestimating the amount of randomness at play. The Canucks' free fall isn't necessarily a representation of their true talent, as an example.

[DrJones]

Re: NHL 2014 Off-Season Thread

Do you consider batting average a good stat in baseball? A single and a home run is the same in BA, just like shooting from the blue line with no screen at all, and shooting it in the slot is the same in Corsi/fenwick. It's a flawed stat, only time it's "useful" when there's context to it (like looking at linemates/quality of competition/off&def zone %/etc.) but not a lot of people do. People like to talk about Leafs' freefall in the 2nd half(post-olympic break? Can't quite remember the time) because they were a terrible Corsi team. Canucks on the other hand had great Corsi and still had their freefall.

Of course it's flawed. All stats require context. Most people I see using Corsi and Fenwick account for this. And bloggers like the ones you threw shade on are the very ones who are detailing zone entries, zone exits, et al. There's a large difference between a stat being flawed but useful and dismissing it out of hand.

Not sure what that has to do with this. They went from 2nd to last in the NHL to 2nd best in West, I'll be shocked if there isn't some type of regression. Devils were a top 2/3 Corsi team in the past 2 years, I'm sure their front office were happy to miss the playoffs since they were a great Corsi team.

Wouldn't the lesson to be gleaned from this be that goaltending and luck tend to play a big role when teams consistently over- or underperform relative to their shot totals? (Whether using Corsi, Fenwick, plain ol' SOG.) The impression I got from the 2013-14 Devils is that they were a solid, playoff-worthy team that was extremely unlucky (0-13 in shootouts) and extremely unwise to give 39 starts to the corpse of Martin Brodeur.

I'm not a huge follower of advanced stats, BTW, but to me they tend to roughly correlate to what I see when I watch games. The Leafs winning games despite being horrendously outshot didn't seem sustainable. The Canucks' strategy of running Kesler and the Sedins into the ice (all 3 were in the top 5 in TOI among forwards, which is way, way more than in previous seasons) didn't seem sustainable, particularly in a stacked division. Do I think Vancouver would've made the playoffs last year if they played in the East? Yes, I do.

[DrJones]

Re: NHL 2014 Off-Season Thread

Leafs will start signing guys like Benoit Pouliots (5yr 20M) of the world and when they don't have the stats to justify those contact, they'll be calling for his head. They already have one of those players in David Clarkson, who lead the Devils in Corsi in 12-13. I look forward to see how the Leafs will be in couple of years.

Pouliot will be an interesting experiment. As far as Clarkson goes, the advanced stats guys were appalled at his signing, and the anti-stats guys were overjoyed ("culture in the room!"), which would imply that the statheads didn't go by the raw Corsi stats (especially in a shortened season), but rather used context (and common sense) to determine that he was a fluky product of playing with Elias.

[DrJones]

Re: NHL 2014 Off-Season Thread

Edit: Also, Avs were supposed to regress since November last year, same with Anaheim because of their high PDO, they regressed all the way to 1st place in the Western Conference! It doesn't bother me when people say Avs will regress.

Fair. I apologize for my snark. The timing on your post just seemed a little random, and I wrongfully assumed fanboyism might be behind it.

Plus it distracted from the larger issue: Glenn Healy sucks.

[DrJones]

Re: NHL 2014 Off-Season Thread

In other news, I see that Puck Daddy's "Most Disappointing Summer" series is down to two: Toronto and Vancouver. Those should be the two longest articles in the site's history.

[AC]

Re: NHL 2014 Off-Season Thread

Edit: Also, Avs were supposed to regress since November last year, same with Anaheim because of their high PDO, they regressed all the way to 1st place in the Western Conference! It doesn't bother me when people say Avs will regress.

It's perfectly possible for a team to post seemingly unsustainable numbers on a single season. Regression isn't saying the numbers will regress in the NEAR future. It's saying that the numbers are likely to regress at some point.

[Kobalt]

Re: NHL 2014 Off-Season Thread

IIRC there's been plenty of work done that depicts shot quality as a non-skill. Could be wrong. There's also been work done that shows Corsi has a strong link to winning. I think you're also underestimating the amount of randomness at play. The Canucks' free fall isn't necessarily a representation of their true talent, as an example.

Corsi was first created to measure a Goalie's activity, now it's a strong indicator of a team winning?

It's perfectly possible for a team to post seemingly unsustainable numbers on a single season. Regression isn't saying the numbers will regress in the NEAR future. It's saying that the numbers are likely to regress at some point.

This completely ignores the fact that rosters are changing every season. If the Avs defense in 2014 is same in 2017, I'll be worried, but they won't be.

Of course it's flawed. All stats require context. Most people I see using Corsi and Fenwick account for this. And bloggers like the ones you threw shade on are the very ones who are detailing zone entries, zone exits, et al. There's a large difference between a stat being flawed but useful and dismissing it out of hand.

Wouldn't the lesson to be gleaned from this be that goaltending and luck tend to play a big role when teams consistently over- or underperform relative to their shot totals? (Whether using Corsi, Fenwick, plain ol' SOG.) The impression I got from the 2013-14 Devils is that they were a solid, playoff-worthy team that was extremely unlucky (0-13 in shootouts) and extremely unwise to give 39 starts to the corpse of Martin Brodeur.

I'm not a huge follower of advanced stats, BTW, but to me they tend to roughly correlate to what I see when I watch games. [/B]The Leafs winning games despite being horrendously outshot didn't seem sustainable. The Canucks' strategy of running Kesler and the Sedins into the ice (all 3 were in the top 5 in TOI among forwards, which is way, way more than in previous seasons) didn't seem sustainable, particularly in a stacked division. Do I think Vancouver would've made the playoffs last year if they played in the East? Yes, I do.

I enjoy looking at advanced stats and reading about them. You'll be shocked at how much people only look at Corsi and Fenwick to determine a team/player.

Like when people talk about the Avs, Roy has to use the collapsing strategy in the own zone since our defense sucks. That leads to a lot of shots from the blueline, which Varlamov pretty much saved this past year.

Spoiler

Roy's strategy ever since his coaching days in Juniors emphasized aggressive neutral zone play (Kings and 'Hawks are great at this). I'm sure he wants to go to the aggresive man coverage system but he probably realized he couldn't do that with guys like Nick Holden, Nate Guenin, Cory Sarich, Ryan Wilson, etc. on defense. Not to mention he prefers the team to carry the puck into the offensive zone instead of just chipping it and line change, and time spent in the offensive zone. He values what all advance stats' people value, get some better defenseman and their advance stats will improve (Trading for Brad Stuart though ).

I also don't understand how people compare Toronto to Colorado. Carlyle's philosophy and Roy's philosophy is completely different.

And yes, Glen Healy is horrible. With Drew Remenda joining the Oilers as well, I guess I won't be watching them much this year if he's still whiny.

[AC]

Re: NHL 2014 Off-Season Thread

Corsi was first created to measure a Goalie's activity, now it's a strong indicator of a team winning?

5v5 Close FF% has a .312 r^2 to win%, which is very good. It's a pretty decent predictor of win%. Corsi is similar (.304). And I have a feeling that that gets magnified if you correlate it to goal differential, which will obviously correlate to win% and support the shots = goals = wins ideology.