You don't need a degree in fancy stats to read this.


- Mr.Bobby

that's last year, where he played well (with Greene) and schneider had a fantastic season. bring up the stats from 2012-2013, 2013-2014. I'm far from convinced he's anything other than a middle pairing defenseman who benefited from playing in front of a terrific (top 3 in the league) goaltender and with an excellent partner.

since 2012? he played 37 nhl games in 2012-2013, 26 nhl games in 2013-2014. Are you including his ahl stats because apparently he was so brilliant the devils didn't play him in the nhl for more than half a season.

I'm curious to see what advanced stats you've twisted to convince yourself that a milbury type trade was actually to the benefit of the oilers.

didn't you also say last summer that griffin reinhart was nhl ready when I told you he wasn't?
- Isles_since_6

No... I didn't say that about Reinhart. I was away from this site for about 3 years...

Here's some info on Larsson that I manipulated...

http://edmontonjournal.co...ntion-or-a-very-lucky-man

http://edmontonjournal.co...lt-of-hockey-adam-larsson

that's last year, where he played well (with Greene) and schneider had a fantastic season. bring up the stats from 2012-2013, 2013-2014. I'm far from convinced he's anything other than a middle pairing defenseman who benefited from playing in front of a terrific (top 3 in the league) goaltender and with an excellent partner.
- Isles_since_6

Can someone hunt down that article where it showed that Greene benefited more from that partnering than Larsson ever did? Actually, I do believe Larsson actually performed better when he was not partnered with Greene. It was posted on ere somewhere a while back...

Isles fan, it's evident that you have your mind made up, and that's fine... We'll see how Larsson does this year.

that's last year, where he played well (with Greene) and schneider had a fantastic season. bring up the stats from 2012-2013, 2013-2014. I'm far from convinced he's anything other than a middle pairing defenseman who benefited from playing in front of a terrific (top 3 in the league) goaltender and with an excellent partner.
- Isles_since_6

See the numbers, particularly the one all the way to the right where it is relative to team. Schneider was there for every other Devils defenseman yet Larssons GA per 60 relative to team was best in the league (including Greene). If there were an award for best defensive defenseman in the league, the conversation last year would have started and ended with Larsson.

Another thing to note, Larsson has had the highest even strength primary points per 60 on the Devils over the last two seasons. Granted the Devils defense isn't loaded with offensive talent, but the point remains.

Larsson is going to help the Oilers a great deal. After all they have finished bottom 3 in the league 3 years running with Taylor Hall.

You don't need a degree in fancy stats to read this.


- Mr.Bobby

Something tells me the fact that both Larsson and Greene are right up there in the top 5 of the league has more to do with Schneider and the NJ system than it does with them. I'm not saying they aren't good defensive players, I'm just saying, until you've isolated for a few other variables, lists like these can be fairly misleading.

For example some variables that need to be removed: the teams system (i.e., high event or low event, although more involved than that), common line-mates, goalies, quality of competition, zone starts, PP/PK time,... etc.

That's just off the top of my head, the actual list goes much deeper. This is the same problem with corsi stats except then you are even a step removed and you need to consider shot quality, efficiency of using line-mates and a whole lot more.

The objective of using these stats is to generate an unbiased low variance metric that highly correlates with wins (or generally goal differential). This is a very difficult thing to accomplish, so anyone who pretends their favorite statistic is the written word of god is typically a naive fool.

Don't tell any of this to Tanner tho. He likes to look at those additional variables and qualitatively state any one in particular has a very small impact, based mostly on his own speculation. Or he will use his typical cheap cop-out and say his opinion is commonly accepted by the stats community, which is actually quite hilarious considering he always over-states how much of an intellectual non-conformist he is.

But the important thing is that, in aggregate, all of these unaccounted variables combined can create a significant bias in certain cases. When these types of bias are observed the stats community Tanner is referring to generally just dismisses them as anomalies because they have a vested interest in the acceptance of their model as the best model. People like Tanner drink that kool-aid.

See the numbers, particularly the one all the way to the right where it is relative to team. Schneider was there for every other Devils defenseman yet Larssons GA per 60 relative to team was best in the league (including Greene). If there were an award for best defensive defenseman in the league, the conversation last year would have started and ended with Larsson.

Another thing to note, Larsson has had the highest even strength primary points per 60 on the Devils over the last two seasons. Granted the Devils defense isn't loaded with offensive talent, but the point remains.

Larsson is going to help the Oilers a great deal. After all they have finished bottom 3 in the league 3 years running with Taylor Hall.
- Mr.Bobby

The goaltending "argument" was so stupid too... Let's discredit every d-man who plays in front of a really good goalie

Can someone hunt down that article where it showed that Green benefited more from that partnering than Larsson ever did? It was posted on ere somewhere a while back...

Isles fan, it's evident that you have your mind made up, and that's fine... We'll see how Larsson does this year.
- laughs2907

See the chart I posted. Larsson posted better defensive numbers than Greene.

Or you can read this, Larsson is not the next Mark Fayne.
http://www.theoilersrig.c...-faynelarsson-comparison/

Btw does anybody know an easy site where you can see the list of moves a team has made under a particular management? Even a chronological list of moves would be good enough, I can figure out the management part on my own.
- RonPielep

http://www.nhltradetracker.com/user/GM_list

Something tells me the fact that both Larsson and Greene are right up there in the top 5 of the league has more to do with Schneider and the NJ system than it does with them. I'm not saying they aren't good defensive players, I'm just saying, until you've isolated for a few other variables, lists like these can be fairly misleading.

For example some variables that need to be removed: the teams system (i.e., high event or low event, although more involved than that), common line-mates, goalies, quality of competition, zone starts, PP/PK time,... etc.

That's just off the top of my head, the actual list goes much deeper. This is the same problem with corsi stats except then you are even a step removed and you need to consider shot quality, efficiency of using line-mates and a whole lot more.

The objective of using these stats is to generate an unbiased low variance metric that highly correlates with wins (or generally goal differential). This is a very difficult thing to accomplish, so anyone who pretends their favorite statistic is the written word of god is typically a naive fool.

Don't tell any of this to Tanner tho. He likes to look at those additional variables and qualitatively state any one in particular has a very small impact, based mostly on his own speculation. Or he will use his typical cheap cop-out and say his opinion is commonly accepted by the stats community, which is actually quite hilarious considering he always over-states how much of an intellectual non-conformist he is.

But the important thing is that, in aggregate, all of these unaccounted variables combined can create a significant bias in certain cases. When these types of bias are observed the stats community Tanner is referring to generally just dismisses them as anomalies because they have a vested interest in the acceptance of their model as the best model. People like Tanner drink that kool-aid.
- RonPielep

Again, the last column is relative to team. Schneider was the goalie for the rest of the Devils as well, yet Larsson out performed his team by a wider margin than any other defenseman in the league.

The goaltending "argument" was so stupid too... Let's discredit every d-man who plays in front of a really good goalie
- laughs2907

You don't need to discredit them, that is silly. You just need to account for it in your statistical analysis before making very bold claims about said d-man.

The goaltending "argument" was so stupid too... Let's discredit every d-man who plays in front of a really good goalie
- laughs2907

Don't even need to do that. Believe it or not Schneider wasn't Larssons personal goalie, he stayed on the ice when the other Devils defenders were out there too.

You don't need to discredit them, that is silly. You just need to account for it in your statistical analysis before making very bold claims about said d-man.
- RonPielep

They do. That's what relative to team means.

See the chart I posted. Larsson posted better defensive numbers than Greene.

Or you can read this, Larsson is not the next Mark Fayne.
http://www.theoilersrig.c...-faynelarsson-comparison/
- Mr.Bobby

I don't think that was the article I was talking about, but you've been a good soldier... I'll talk to Hillary and get you taken off of her list.

Again, the last column is relative to team. Schneider was the goalie for the rest of the Devils as well, yet Larsson out performed his team by a wider margin than any other defenseman in the league.
- Mr.Bobby

Yes, I saw that afterwards. But the point remains. Until you go about looking into how exactly they accounted for 'team' then a healthy dose of speculation should be applied. Accounting for something like the goal-tending (let alone team) is not nearly as simple as one would imagine. There are a lot more intricacies going on in the real-world system than can simply be accounted for by SV%.

Do you think Larsson will lead the league or even be top 3 next year in GA60RelTM?

Because I would be willing to bet money that playing in a different system, in front of a different net-minder and with different line-mates will change his GA60RelTM numbers significantly (even tho those things are 'accounted for'). I say this because I have a strong suspicion that their RelTM term is too simplistic and their model is flawed.

http://www.nhltradetracker.com/user/GM_list
- Tumbleweed

Thank you good sir.

Yes, I saw that afterwards. But the point remains. Until you go about looking into how exactly they accounted for 'team' then a healthy dose of speculation should be applied. Accounting for something like the goal-tending (let alone team) is not nearly as simple as one would imagine. There are a lot more intricacies going on in the real-world system than can simply be accounted for by SV%.

Do you think Larsson will lead the league or even be top 3 next year in GA60RelTM?

Because I would be willing to bet money that playing in a different system, in front of a different net-minder and with different line-mates will change his GA60RelTM numbers significantly (even tho those things are 'accounted for'). I say this because I have a strong suspicion that their RelTM term is too simplistic and their model is flawed.
- RonPielep

There's a very good chance he will.

And the way it's measured is the team with Larsson on ice vs. without. What the stat says is that the Devils give up an additional goal for every 60 minutes that Larsson is not on the ice, with or without Schneider. Again this stat is only relative to the rest of the Devils, who also had Schneider playing behind them.

There's a very good chance he will.

And the way it's measured is the team with Larsson on ice vs. without. What the stat says is that the Devils give up an additional goal for every 60 minutes that Larsson is not on the ice, with or without Schneider. Again this stat is only relative to the rest of the Devils, who also had Schneider playing behind them.
- Mr.Bobby

Slap my ass and call me Sally, that's an awesome stat to have.

Yep Edmonton had the 4th worst GA per game last year and 18th ranked PK last year.

Add a defenseman who had the 2nd lowest GA per 60 (to Stralman), yet had the highest defensive zone starting percentage in the league, with next to no PP time and played on the top PK unit for the 8th best PK in the league. Managed to be a +15 on a team that had -20 goal differential.

But he is terrible because he didn't have 40 points while playing on the lowest scoring team in the league. Adam Larsson is going to make a lot of people look silly this year, but one of them won't be PC.
- Mr.Bobby

Damn, I hope you're right, but that jury is out for now.

There's a very good chance he will.

And the way it's measured is the team with Larsson on ice vs. without. What the stat says is that the Devils give up an additional goal for every 60 minutes that Larsson is not on the ice, with or without Schneider. Again this stat is only relative to the rest of the Devils, who also had Schneider playing behind them.
- Mr.Bobby

Ooooh I see. Just relative to his own team. My bad.

This is going to sound like I'm nit-picking but if you want to accurately determine that then you need to normalize for a lot of factors in order to make sure the times he was on the ice are equivalent (in terms of any other variables that can effect GA) to the times he was not on the ice. This is an impossible thing to accomplish, but there are ways to attempt it in hopes of reducing bias (although sometimes you end up adding more).

Here is a theoretical example of where these things can get really complicated. Say Larsson gets injured and Schneider has a bit of a slump during that period. Well, some could say that Schneider didn't play as well because Larsson wasn't there to help limit the tough opposition. That is an entirely reasonable conclusion but it is far from fact. There are many other variables that could have caused a Schneider slump, some that may not have had anything to do with Larsson. Maybe they were facing tougher teams, maybe team chemistry/morale as a whole was in the poopter or maybe Schneider just wasn't feeling it. The ultimate effect is Larsson ends up with much better GA60RelTM numbers because of Schneider's slump during his absence. But whether or not Larsson's absence actually contributed to that slump is entirely unknown (the slump could have started one or two games prior).

The Yak isn't going anywhere. George Costanza is going to give him '1 more year' to realise his potential & stature of a #1 pick (again). (just like being at the bar & having 1 more for the road ).

Anyways, sorry about ranting, if anyone is interested I would be willing to bet that Larsson isn't in the top 10 in the league in the simple GA60 stat next year.

I don't want to bet on the RelTM stat because I don't doubt that Larsson will be the best defensive D-man on the Oil next year. The bar isn't set very high.

Anyways, sorry about ranting, if anyone is interested I would be willing to bet that Larsson isn't in the top 10 in the league in the simple GA60 stat next year.

I don't want to bet on the RelTM stat because I don't doubt that Larsson will be the best defensive D-man on the Oil next year. The bar isn't set very high.
- RonPielep

Away from Greene/ Schneider, Larssons goals against per 60 is likely to look worse, But I think he will lead the Oilers, and probably by about as large a margin as he did with the Devils.

And wasn't that the point of the trade? To improve the Oilers defense?

So the habs would have three right-wingers that play top six? Does that mean they would shift Radulov to left-wing complicating his adjustment to the nhl and therefore increasing his possibility of failure. Maybe third line of DD, Shaw, Yaks? Maybe Bergevin will move Gallagher for top flight left side moving defenceman. Bold move!!!! Lmfao!

Away from Greene/ Schneider, Larssons goals against per 60 is likely to look worse, But I think he will lead the Oilers, and probably by about as large a margin as he did with the Devils.

And wasn't that the point of the trade? To improve the Oilers defense?
- Mr.Bobby

You're (frank)ing right, Bobby... You're (frank)ing right.

You're (frank)ing right, Bobby... You're (frank)ing right.
- laughs2907