Around the League 2019-20 IV

[HBK27]

How is Kovy a -10 with 8 points?

Guess who is the first guy cut off at #8? Severson.

Severson is -6 on the season - tied for 29th worst overall with Wood and 8 others.

 

[New Jersey]

oh cool tampa plays at the garden tonight.

suffice it to say we'll get smoked tomorrow.

 

[BenedictGomez]

You hate to see it.

Somewhere along the way it became "cool" to say +/- is a dumb stat. On its' face, it should be intuitively obvious that that is a "stupid" belief. But much like, "dont go swimming for an hour after eating", many people now believe it, regardless of how untrue it is. Many simply say it because they think they're supposed to say it.

 

[Emperoreddy]

People say it because it’s a dumb stat and we have access to many more informative stats.

All it says is you were on the ice when a goal was scored. It says absolutely nothing about what you did while on the ice, and it doesn’t count certain special teams situations.

It’s pretty useless

 

[Blender]

Somewhere along the way it became "cool" to say +/- is a dumb stat. On its' face, it should be intuitively obvious that that is a "stupid" belief. But much like, "dont go swimming for an hour after eating", many people now believe it, regardless of how untrue it is. Many simply say it because they think they're supposed to say it.

People say it because it is a dumb stat and has been demonstrated to be so. We have much better stats in 2019 that provide a far better analysis of past performance and are far better at predicting future performance.

 

[Emperoreddy]

The reality here is Severson had a good +/- it would be argued even further that it’s an awful stat and this totally different stat that says he is terrible would be put on a pedestal.

 

[Zajacs Bowl Cut]

plus/minus is the 2nd worst stat in sports

1st in wins for a MLB pitcher

 

[BenedictGomez]

People say it because it’s a dumb stat and we have access to many more informative stats.

People say it because it is a dumb stat and has been demonstrated to be so.

The most exhaustive, mathematical study I've seen to data of "advanced stats" in hockey came to the surprised (because they didn't think it would be so) conclusion that simple goal differential was the best predictor in hockey, even in the world of teh addvahnced statz. It doesn't shock me in the least given goals are the most important & predictive thing that occurs in hockey. All +/- is is a goal differential statistic at a player level. Now obviously it's not the best stat, or the be-all-end-all stat, or even the best version of that "type" of stat (I prefer others), but to say it's completely meaningless is the verbal equivalent of The Emperor Wears No Clothes with regards to this subject. And yes, most of the time it's people trying to sound cool or smart.

 

[BenedictGomez]

The reality here is Severson had a good +/- it would be argued even further that it’s an awful stat

Except Severson has the worst +/- of any Devils defenseman on the entire team for 3 of the last 4 seasons now (one year he was second to last, by -8 to -9), which makes perfect "eye test" sense to me.

 

[FinnishDevil]

Rico has been on one of his hot streaks early this season. Hopefully he'll have a good year. Ducks dont seem that boring this year so far either.

 

[Ripshot 43]

The league feels upside down so far this season. Not the entire league but definitely some surprises out there.

 

[Blender]

The most exhaustive, mathematical study I've seen to data of "advanced stats" in hockey came to the surprised (because they didn't think it would be so) conclusion that simple goal differential was the best predictor in hockey, even in the world of teh addvahnced statz. It doesn't shock me in the least given goals are the most important & predictive thing that occurs in hockey. All +/- is is a goal differential statistic at a player level. Now obviously it's not the best stat, or the be-all-end-all stat, or even the best version of that "type" of stat (I prefer others), but to say it's completely meaningless is the verbal equivalent of The Emperor Wears No Clothes with regards to this subject. And yes, most of the time it's people trying to sound cool or smart.

[citation needed]

I would love to read this supposed study. Plus/minus has been analyzed to death and shown to be meaningless, and I have posted links before and even in the last week or so here.

 

[BenedictGomez]

[citation needed]

I posted the Goal Differential study last year in the analytics thread. It's a good read. And unlike most of the stuff I read that's just tossed-out-there with literally, and I do mean literally, no explanation for method of calculation, it was well done. It's also not the same thing as +/- obviously. That said, calling +/- "meaningless" is foolish. If it's foolish stats you're looking for, I'd submit Goals Above Replacement, which reminds me of when I was in meetings with derivatives statisticians trying to explain the math behind Collateralized Debt Obligations prior to the subprime mortgage crisis.

 

[devilsblood]

The most exhaustive, mathematical study I've seen to data of "advanced stats" in hockey came to the surprised (because they didn't think it would be so) conclusion that simple goal differential was the best predictor in hockey, even in the world of teh addvahnced statz. It doesn't shock me in the least given goals are the most important & predictive thing that occurs in hockey. All +/- is is a goal differential statistic at a player level. Now obviously it's not the best stat, or the be-all-end-all stat, or even the best version of that "type" of stat (I prefer others), but to say it's completely meaningless is the verbal equivalent of The Emperor Wears No Clothes with regards to this subject. And yes, most of the time it's people trying to sound cool or smart.

I certainly wouldn't categorize it as a good stat. It jumbles up way too much stuff, while also leaving a bunch of stuff out.

Now, the fact that Sev's is consistently terrible in this statisitic is telling. Or at least alarming in regards to making one want to look into better stats. Which also don't paint a great picture for Sev's.

 

[BenedictGomez]

I certainly wouldn't categorize it as a good stat. It jumbles up way too much stuff, while also leaving a bunch of stuff out.

Now, the fact that Sev's is consistently terrible in this statisitic is telling. Or at least alarming in regards to making one want to look into better stats. Which also don't paint a great picture for Sev's.

I agree. But not according to the people who call +/- a "dumb" stat. It's completely irrelevant to them.

 

[devilsblood]

I agree. But not according to the people who call +/- a "dumb" stat. It's completely irrelevant to them.

My solution would be just to use better stats which also support the same argument. Like his 5v5 +/- which is a team worst -7.

Carrick is the only other d-man who is worse then a -1.

 

[Zajacs Bowl Cut]

looks like Cam Fowler has played very well to start this year

54.2% CF and a 13 (!!!) CF% rel to go along with 7 points in 14 games. I know he got discussed a lot last year as a potential trade target.

 

[Triumph]

I posted the Goal Differential study last year in the analytics thread. It's a good read. And unlike most of the stuff I read that's just tossed-out-there with literally, and I do mean literally, no explanation for method of calculation, it was well done. It's also not the same thing as +/- obviously. That said, calling +/- "meaningless" is foolish. If it's foolish stats you're looking for, I'd submit Goals Above Replacement, which reminds me of when I was in meetings with derivatives statisticians trying to explain the math behind Collateralized Debt Obligations prior to the subprime mortgage crisis.

You posted a study that agrees with what you already thought, and that study was a lot of obvious stuff that anyone who has spent time thinking about this stuff would realize is true.

+/- is meaningless. GAR has its problems for sure, and it's also not predictive, but I'd take it over +/- any day.

 

[devilsblood]

looks like Cam Fowler has played very well to start this year

54.2% CF and a 13 (!!!) CF% rel to go along with 7 points in 14 games. I know he got discussed a lot last year as a potential trade target.

Also looks like he is being put into some very advantageous situations.

More then 40% of his minutes are against weaker comp, and he's getting 75% OZS vs that weaker comp.

Looks like it is working, but it is nice when you play behind a guy like Lindholm who hogs all the tough minutes.

 

[SteveCangialosi123]

 

[njdevil26]

Really scary what happened to Scott Sabourin. Very classy by Backes and both teams on how they handled it. Hockey is an amazing community.

 

[Blackjack]

yyyyup...

 

[Blackjack]

Okay, went down a rabbit hole a bit here, and it seems that a week ago (Oct. 26) CJ Turturo had Pavel Zacha as one of the best players in the league on expected goal differential.