Calling all "advanced" stat believers

[Daniel]

Maybe I'm unfairly lumping all the puck possession, Corsi believers together, but oh well.

Anyway, the upcoming season is your time to shine and convince dinosaurs and fuddy duddies such as myself that your stats are what matters, or at least matters most.

We're told goaltending is what killed us last year. Well, now you've got the guy with the magical save percentage, who ought to be playing at least 50 games, and probably more if Brodeur costs the Devils as many games as people think.

Even strength is where games are won, I'm told, and even strength ought not to suffer all that much without Kovy, so sayeth the spreadsheet. If special teams is shown to be the difference next year, I'll cut you some slack, but otherwise it's time to put up or shut up so far as the point column in the standings is concerned. Blaming results that aren't consistent with your predictions on "luck" (except for injuries) is not going to cut it. Not anymore.

Sent from my iPad using Tapatalk HD

[Triumph]

Maybe I'm unfairly lumping all the puck possession, Corsi believers together, but oh well. Anyway, the upcoming season is your time to shine and convince dinosaurs and fuddy duddies such as myself that your stats are what matters, or at least matters most. We're told goaltending is what killed us last year. Well, now you've got the guy with the magical save percentage, who ought to be playing at least 50 games, and probably more if Brodeur costs the Devils as many games as people think. Even strength is where games are won, I'm told, and even strength ought not to suffer all that much without Kovy, so sayeth the spreadsheet. If special teams is shown to be the difference next year, I'll cut you some slack, but otherwise it's time to put up or shut up so far as the point column in the standings is concerned. Blaming results that aren't consistent with your predictions on "luck" (except for injuries) is not going to cut it. Not anymore. Sent from my iPad using Tapatalk HD

 

Again, you don't understand how chance works.  But yes, the truth is that Ilya Kovalchuk was not a very good player 5 on 5 for NJ - a break-even player, probably.  NJ is going to struggle mightily to score goals even if they get an RW to replace some of his minutes (and indeed they were going to struggle a lot anyway), but if the goaltending's there and they don't get killed injury-wise at forward, they should squeeze into the playoffs.

Edited July 11, 2013 by Triumph

[SterioDesign]

Maybe I'm unfairly lumping all the puck possession, Corsi believers together, but oh well. Anyway, the upcoming season is your time to shine and convince dinosaurs and fuddy duddies such as myself that your stats are what matters, or at least matters most. We're told goaltending is what killed us last year. Well, now you've got the guy with the magical save percentage, who ought to be playing at least 50 games, and probably more if Brodeur costs the Devils as many games as people think. Even strength is where games are won, I'm told, and even strength ought not to suffer all that much without Kovy, so sayeth the spreadsheet. If special teams is shown to be the difference next year, I'll cut you some slack, but otherwise it's time to put up or shut up so far as the point column in the standings is concerned. Blaming results that aren't consistent with your predictions on "luck" (except for injuries) is not going to cut it. Not anymore. Sent from my iPad using Tapatalk HD

 

it totally agree with this cause its always whats happening and if you dont agree with them you simply "dont understand it". 

 

Thats all i have to say, Tri don't reply to this i don't feel like arguing today, at all.

[Daniel]

Again, you don't understand how chance works. But yes, the truth is that Ilya Kovalchuk was not a very good player 5 on 5 for NJ - a break-even player, probably.

I understand what you mean by chance and luck. What I'm saying though is that the "advanced" stats are supposed to have some kind of predictive power, or at least they ought to if you're touting them so much. If inconsistent results are explained away by things like puck luck and chance, then you can never be proven wrong. That's the first sign of a theory that isn't really a theory at all.

Sent from my iPad using Tapatalk HD

[Triumph]

I understand what you mean by chance and luck. What I'm saying though is that the "advanced" stats are supposed to have some kind of predictive power, or at least they ought to if you're touting them so much. If inconsistent results are explained away by things like puck luck and chance, then you can never be proven wrong. That's the first sign of a theory that isn't really a theory at all. Sent from my iPad using Tapatalk HD

 

That's idiotic, even for a believer in goalie wins.  We are talking about chance events - if I believe a coin to be fair and predict 41 heads and 41 tails in 82 subsequent coin flips, and it doesn't come out that way or even close, how much do we have to question the coin? When things on a league-wide level contradict the theory, it requires fuller examination.  Indeed, it always requires examination.  But one team in one year - no.  I think we'll have more data next year and the Devils will be a good test case about how zone time shots on goal and shots on goal off the rush differ, but the Devils should still be a top possession team.  Whether they shoot 8% - eh, I wouldn't bet on that, but I would still bet on the Devils making the playoffs.

 

Kovalchuk's 5 on 5 numbers from his NHL career suck any way you try to slice them.  There's just no way of making them look good.

[Daniel]

I understand what you mean by chance and luck. What I'm saying though is that the "advanced" stats are supposed to have some kind of predictive power, or at least they ought to if you're touting them so much. If inconsistent results are explained away by things like puck luck and chance, then you can never be proven wrong. That's the first sign of a theory that isn't really a theory at all. Sent from my iPad using Tapatalk HD

That's idiotic, even for a believer in goalie wins. We are talking about chance events - if I believe a coin to be fair and predict 41 heads and 41 tails in 82 subsequent coin flips, and it doesn't come out that way or even close, how much do we have to question the coin? When things on a league-wide level contradict the theory, it requires fuller examination. Indeed, it always requires examination. But one team in one year - no. I think we'll have more data next year and the Devils will be a good test case about how zone time shots on goal and shots on goal off the rush differ, but the Devils should still be a top possession team. Whether they shoot 8% - eh, I wouldn't bet on that, but I would still bet on the Devils making the playoffs.

Kovalchuk's 5 on 5 numbers from his NHL career suck any way you try to slice them. There's just no way of making them look good.

Well unfortunately for you, we're not talking about coin flips or roulette where you can test the probabilities over a long period of time and control the variables. If you conceded these limitations and just said things like puck possession, corsi, save percentage or whatever are things that get overlooked and can give you a little more insight than the traditional stats and what your eyeballs tell you, then I wouldn't really think much of it one way or the other.

But here's a great opportunity to show that there's something more to it than just an interesting observation. While predicting that the Devils should squeeze into the playoffs is a bit in the realm of cold reading, I'll accept it as a concrete prediction.

Sent from my iPad using Tapatalk HD

[Triumph]

Well unfortunately for you, we're not talking about coin flips or roulette where you can test the probabilities over a long period of time and control the variables. If you conceded these limitations and just said things like puck possession, corsi, save percentage or whatever are things that get overlooked and can give you a little more insight than the traditional stats and what your eyeballs tell you, then I wouldn't really think much of it one way or the other. But here's a great opportunity to show that there's something more to it than just an interesting observation. While predicting that the Devils should squeeze into the playoffs is a bit in the realm of cold reading, I'll accept it as a concrete prediction. Sent from my iPad using Tapatalk HD

 

Wow, I get a sample size of one instead of a sample size of the entire rest of the league throughout the last 6 years.  I don't really get why this is something that needs testing unless you're just an eternal skeptic who can't see evidence when it's right in front of his face.  Shooting percentage regresses towards the mean.  Save percentages don't significantly change when players change teams.  Defensemen cannot control their goalie save percentage.  These things are so easy to see in the past.  There is no reason why they need testing in the future.  It's just a question of degree now - should I expect NJ to shoot 8% next year?  No, probably not, but it's hard to expect them to shoot significantly lower than 7%.  And indeed, they could do better - there are large degrees of chance involved.

Edited July 12, 2013 by Triumph

[Pepperkorn]

Oddly enough, Daniel, you made me feel a lot better about things.  I didn't really think Ilya was contributing much.  So you've convinced me this is a pretty good thing in reality.  It's just a psychological blow - which is very real and is also very unpredictable.  I put it to you that it's not luck or chance -- it's mindset.  It's how the team responds to the changes.  I think it's not difficult to translate it all into a good thing for some of our young guys even. 

 

personally I found Ilya's style of hard work borderline draining.  He was a hard worker - but he made it look like work that's for sure.

[Daniel]

Well unfortunately for you, we're not talking about coin flips or roulette where you can test the probabilities over a long period of time and control the variables. If you conceded these limitations and just said things like puck possession, corsi, save percentage or whatever are things that get overlooked and can give you a little more insight than the traditional stats and what your eyeballs tell you, then I wouldn't really think much of it one way or the other. But here's a great opportunity to show that there's something more to it than just an interesting observation. While predicting that the Devils should squeeze into the playoffs is a bit in the realm of cold reading, I'll accept it as a concrete prediction. Sent from my iPad using Tapatalk HD

Wow, I get a sample size of one instead of a sample size of the entire rest of the league throughout the last 6 years. I don't really get why this is something that needs testing unless you're just an eternal skeptic who can't see evidence when it's right in front of his face. Shooting percentage regresses towards the mean. Save percentages don't significantly change when players change teams. These things are so easy to see in the past. There is absolutely no reason why they need testing anymore.

If all you're saying is that shooting percentage tends to regress towards the mean and that save percentage doesn't "significantly" change between teams (and I'll cut you some slack with the weasely use of the word "significantly" -- I'm an attorney, I know weasel words when I see them) then fine, I wouldn't care all that much. But you made categorical assertions about how awful the goaltending was last year and that it was what kept and otherwise good team from getting further. You have also said that Kovalchuk doesn't help all that much at even strength. These are predictions that I care about, and am interested to see if it plays out that way.

Sent from my iPad using Tapatalk HD

[Pepperkorn]

Are you two dating?  You should be.

[Triumph]

If all you're saying is that shooting percentage tends to regress towards the mean and that save percentage doesn't "significantly" change between teams (and I'll cut you some slack with the weasely use of the word "significantly" -- I'm an attorney, I know weasel words when I see them) then fine, I wouldn't care all that much. But you made categorical assertions about how awful the goaltending was last year and that it was what kept and otherwise good team from getting further. You have also said that Kovalchuk doesn't help all that much at even strength. These are predictions that I care about, and am interested to see if it plays out that way. Sent from my iPad using Tapatalk HD

 

The trouble is, again, if the Devils don't do well at even strength, it may have nothing at all to do with Kovalchuk.  And if the goaltending is awful again, it may have nothing at all to do with the defense or offense.

 

They're not weasel words.  They are allowing for the possibility of chance.  Eric Tulsky has already analyzed save percentage between teams, at a link I cannot find right now, but I'll find it eventually, and it just doesn't change significantly.  Here is The Contrarian Goaltender saying same:  http://blog.philbirnbaum.com/2011/12/are-there-good-team-goalies-and-bad.html#c4394415888881657115

[Jas0nMacIsaac]

Im still going to toss 50 on the devils as soon as vegas reacts to the doom and gloom scenerio. They are still a playoff team imo. If schnieder goes on a 940 streak its hard to beat.

[Daniel]

Im still going to toss 50 on the devils as soon as vegas reacts to the doom and gloom scenerio. They are still a playoff team imo. If schnieder goes on a 940 streak its hard to beat.

Now that's a prediction I'm interested in. Or I like it when people put their money where their mouth is.

Edit: and your tweet to the same effect and your presidents trophy predicition is what prompted me to start the thread.

Sent from my iPad using Tapatalk HD

Edited July 12, 2013 by Daniel

[Triumph]

Now that's a prediction I'm interested in. Or I like it when people put their money where their mouth is. Edit: and your tweet to the same effect and your presidents trophy predicition is what prompted me to start the thread. Sent from my iPad using Tapatalk HD

 

If gambling laws in the US weren't totally screwed, I'd put money on the Devils at whatever odds they're offering in 2 weeks.

[sundstrom]

i'm willing to bet the devils vegas odds don't change that dramatically because there's very little action on it in the first place and hockey isn't something vegas pays that much attention to.

[DH26]

i'm willing to bet the devils vegas odds don't change that dramatically because there's very little action on it in the first place and hockey isn't something vegas pays that much attention to.

 

after they got Schneider they jumped from 50-1 to 25-1 so there's that 

[Triumph]

BTW, here's the piece on goalies changing teams, wasn't written by Eric T. after all:  http://vhockey.blogspot.com/2009/07/shot-quality-fantasy.html

[Satans Hockey]

I'm not into the advanced stats but I do find them to be helpful. I'm just not motivated into following them and would much rather read Triumphs cliff notes version haha

[AEWHistory]

......Defensemen cannot control their goalie save percentage.  These things are so easy to see in the past.......

Normally I really like and agree with your posts Tri, but I've never disagreed with something you have written so much. The above is just not true. How do you mean defensemen cannot control save percentage? There are many factors, but defense is a key element.

[Devils731]

Normally I really like and agree with your posts Tri, but I've never disagreed with something you have written so much. The above is just not true. How do you mean defensemen cannot control save percentage? There are many factors, but defense is a key element.

Good defense tends to cut down the number of shots but the ratio of good to bad shots stays pretty static. So good defense doesn't change save percentage but does cut down a goalies goals against average.