214 replies to this topic

[squishyx]

The problem is if you bench a guy after his luck runs out, you've used him for a bunch of games where he wasn't good and he paid the price for it.  You'll never know his luck has run out ahead of time, because luck is random.  That's one reason to use microstats to verify what you see on the ice, it can let you be more confident and quick in your decision making about who is or isn't performing.

The problem with luck is too often people just use it to explain data that doesn't fit their model. Bernier and Carter had terrible CORSI numbers last season, but they got "lucky" in the playoffs. If Gio finishes the year with 20 points and a +5 do you think Tri is going to come back and go "Gio proved me wrong, he clearly deserved and earned a spot on this team"? No he is going to say "he got lucky" or dig up some "microstat" that justifies his opinion.

Gio is 22 with little NHL experience; whose to say one way or the other what production levels should be "typical" for him. At a minimum though it means we shouldn't bench the guy who is 5th in scoring on the team just because a single statisitic suggests it.

[squishyx]

Im sure if the original post was presented like it is in bold, instead of the hyperbolic 'gionta is killing the team' or that his line 'isnt even a good fourth line', or that Gionta is 'far from an NHL player' and 'never will be', then more people would not tune out the corsi and other stats presented. You cant expect people to take the objective stats when its poorly interpreted by the presenter.

Well said.

[Triumph]

The problem with luck is too often people just use it to explain data that doesn't fit their model. Bernier and Carter had terrible CORSI numbers last season, but they got "lucky" in the playoffs. If Gio finishes the year with 20 points and a +5 do you think Tri is going to come back and go "Gio proved me wrong, he clearly deserved and earned a spot on this team"? No he is going to say "he got lucky" or dig up some "microstat" that justifies his opinion.

Gio is 22 with little NHL experience; whose to say one way or the other what production levels should be "typical" for him. At a minimum though it means we shouldn't bench the guy who is 5th in scoring on the team just because a single statisitic suggests it.

 

Stephen Gionta is 29 years old, he turns 30 is October.  If he were 22, I would not be so certain in my opinion about him.

[squishyx]

Stephen Gionta is 29 years old, he turns 30 is October.  If he were 22, I would not be so certain in my opinion about him.

You are right, I didn't mean to write that. Still though, his NHL experience his minimal which is what I was trying to get at.

[SterioDesign]

Tri always brings some interesting stat angles to the table, especially since most hockey fans still aren't that aware of them.  Tri's problem has always been bedside manners when it comes to presentation...he can be arrogant, condescending, and almost robotic at times, and it has a way of getting under some people's skin (admittedly Tri has ticked me off more than once).  I think his demeanor has turned off some fans' to the Fenwicks and Corsis and other microstats he brings to the board.  But I've also defended his viewpoints when I've agreed with them, and I don't think he's far off on Gio to this point...yeah, the numbers for him look pretty good, but beyond the obvious numbers, there's solid reasons to wonder if he can possibly keep this up.

 

This is spot on for me Colorady ( lol ) anything I posted in this whole discussion is not cause i like Gionta or anything. it's based on the fact that i'm absolutely fed up with his way of saying things, not to the point of blocking Tri of course, he's a good poster. But now every time he's shoving shooting % in a discussion to make it looks smart it's getting under my skin cause it sounds so pretentious and often it's not taking into account anything else.

 

I mean when anyone on this forum would say something like "Henrique has been playing pretty well with Zajac lately"... Tri would likely go and make a statement in opposition saying something like..."Henrique shooting % has been down 2.3% since playing with Zajac and long run Deboer will have to split them up cause they'll run Kovy to the ground since he was maintaining a constant % and bla bla bla bla bla" or something similar. I mean... to me now it always sounds like someone trying to look smarter than anyone and its ticking me off every time.

[Devils Pride 26]

This thread is so simple and is being so overblown. The premise is that Gionta and the fourth line putting up numbers is largely luck and that we have to be careful not to overuse them because that luck is going to run out very soon. That premise is undeniably true.

 

Corsi is not some detached stat line made up in a lab by a professor. It accounts for shooting attempts. You need to shoot the puck to score. It is generally better to out shoot rather than be out shot.

[Neb00rs]

Stephen Gionta has done everything asked of him and well beyond, all for a bargain basement price. I find no reason to complain about him and nor will I. What we see is what we get - to defy obvious production and not give Gionta due credit.is just plain wrong. As a matter of fact - Triumph is not in the minority - two years ago he would have been one of the myriad hockey "experts," who wrote off, "career AHLer Stephen Gionta...I for one am glad Stephen Gionta didn't listen then and I know he wouldn't listen now either.

Edited by Neb00rs, 21 February 2013 - 09:41 AM.

[capo]

This thread is so simple and is being so overblown. The premise is that Gionta and the fourth line putting up numbers is largely luck and that we have to be careful not to overuse them because that luck is going to run out very soon. That premise is undeniably true.

 

Corsi is not some detached stat line made up in a lab by a professor. It accounts for shooting attempts. You need to shoot the puck to score. It is generally better to out shoot rather than be out shot.

 

We definitely can't  continue to see them putting up numbers against unfavorable matchups.  However, they should be able to kick the sh!t out of opposing teams 4th lines on a nightly basis.  At this stage in the season and with all the holes in the lineup you can't squarely match up against other teams lines.  Sometimes you just have to roll them.  I'm not worried about over playing these guys.  It's not going to lead to the Devils missing the playoffs.  What I am worried about is giving Kovy 26 mins a game.  Especially since he's been playing hockey since the start of the KHL season.  What's going to happen is he's going to burn out like he did last year.  If only advanced stats could predict burnout.  Who gives a sh!t if we burn out Gionta.

[thefiestygoat]

This thread is so simple and is being so overblown. The premise is that Gionta and the fourth line putting up numbers is largely luck and that we have to be careful not to overuse them because that luck is going to run out very soon. That premise is undeniably true.

 

Corsi is not some detached stat line made up in a lab by a professor. It accounts for shooting attempts. You need to shoot the puck to score. It is generally better to out shoot rather than be out shot.

Agreed, though I will say there have been some solid posts back and forth.

 

In baseball I'm all about using sabremetrics since the game has so many situations that repeat themselves over and over. In NHL hockey I like to balance microstats and my own observations since it is a much more fluid, random game about 50/50. With that said, I don't see how Gionta continues to be this lucky and put up points at the rate he has. I love his story and his work ethic, but years of watching NHL hockey and the microstats make me skeptical it will last.

[SMantzas]

thefiestygoat, on 20 Feb 2013 - 19:24, said:
Agreed, though I will say there have been some solid posts back and forth.

In baseball I'm all about using sabremetrics since the game has so many situations that repeat themselves over and over. In NHL hockey I like to balance microstats and my own observations since it is a much more fluid, random game about 50/50. With that said, I don't see how Gionta continues to be this lucky and put up points at the rate he has. I love his story and his work ethic, but years of watching NHL hockey and the microstats make me skeptical it will last.

Agreed with pretty much everything you said, Feisty. Advanced stats are helpful because they give you hard, concrete data. To me, Gionta just doesn't pass the eye test. He's frequently scrambling around in the defensive zone and a poor penalty killer. On the PK, teams literally pass around the guy because he uses a smaller stick and doesn't have the range most players have. I blame Deboer for matching him against top players frequently and for playing him so often. To me, he should play roughly 8 mins ES with no special teams time. If Deboer uses him in strictly a 4th line role, he won't kill the team (which will inevitably happen because he has gotten extremely lucky thus far). I realize Coach hasn't had a real 3rd line, but some of his choices could be better. But thats for another thread

Edited by SMantzas, 21 February 2013 - 12:14 AM.

[EdgeControl]

Kovalchuk is no fenwick star himself, and possibly his shifting on with gio is exacerbating the problem. both of them together in the D zone is a bit scary!

Lucky or not, goal differential will trump shot differences in the short term. so until Gio's "GOAL HART+" starts to slip significantly he's staying on the line!  JJ should take note of that!  with the limited scoring this year you cant bench a guy for being lucky. now lets see how carters injury effects the mojo!

[mrthemike]

The numbers are nice and show a story. A comparison to equal players / lines would be nice. 

[squishyx]

This thread is so simple and is being so overblown. The premise is that Gionta and the fourth line putting up numbers is largely luck and that we have to be careful not to overuse them because that luck is going to run out very soon. That premise is undeniably true.
 
Corsi is not some detached stat line made up in a lab by a professor. It accounts for shooting attempts. You need to shoot the puck to score. It is generally better to out shoot rather than be out shot.

That's not really the premise. It was to "bench Gionta" because he is "killing this team". A more reasonable premise probably would have seen this thread die out days ago because everyone more or less agrees that Gio is getting lucky in some degree, but wildly disagree on whether or not it makes sense to continue to play him and ride whatever luck he has left.

Corsi is fine, the stats are not the problem here, it's the weight being attached to it. Making a big decision based on [i]any sole[i] stat is foolish, as Gio showed last game.

[thefiestygoat]

Not to start a huge fight or anything but for those of us who like Behind the Net I thought this was interesting:

 

Talk about how lucky Stephen Gionta has been at 5v5: 53.7 Off Zone Start% (2nd highest NJ FWD's w/ 15+ GP) and a 1061 PDO (highest amongst those FWD's).

Despite the favorable zone start % and the luck (high PDO), Gionta has a negative Corsi Rel (-14.0, 2nd lowest amongst NJ FWD's w/ 15+ GP) and the lowest Corsi On (-7.68) amongst those FWD's.

[Pepperkorn]

Clearly Corsis aren't doin' it

 

I'll say again - data sets in hockey are too small with too many inconsistencies.

 

Moneyball = one dude ~4 times a game coming up to bat.

 

Moneypuck = one dude getting inconsistent minutes, inconsistent shots, inconsistent linemates, inconsistent player match-ups.

Not really doable is it?

 

I'm not saying don't have the conversation - it's amusing & interesting -- but it's not reliable. 

It makes no sense to bench guys based on unproven statistical analysis, particularly when a statistician by trade would NEVER have you doing so. Now I've often said I was raised by an actuarial scientist on this board - thus I've never really broached the subject - it's kind of ridiculous to take it too seriously.  But what the heck right?  Why not start bringing the real field of risk analysis into the equation here, right? 

[thefiestygoat]

I would never directly compare baseball sabremetrics to hockey "microstats" but there is value in both, moreso in baseball imo because of the nature of the game. If NHL teams are hiring stat guys using things like Corsi and Fenwick, then I see no reason for the public to not embrace it. I wouldn't use hockey "microstats" in place of what my eye sees or vice versa. But in this case, a guy like Gionta who has had consistent minutes and has had consistent linemates, clearly both my eye and the numbers tell me he sucks.

 

Also sabremetrics are so much more than at bats. So much more.

[squishyx]

I would never directly compare baseball sabremetrics to hockey "microstats" but there is value in both, moreso in baseball imo because of the nature of the game. If NHL teams are hiring stat guys using things like Corsi and Fenwick, then I see no reason for the public to not embrace it. I wouldn't use hockey "microstats" in place of what my eye sees or vice versa. But in this case, a guy like Gionta who has had consistent minutes and has had consistent linemates, clearly both my eye and the numbers tell me he sucks.
 
Also sabremetrics are so much more than at bats. So much more.

The Devils generally had a high corsi number tonight, the Leafs did not. The Leafs won.
Zidlicky and Carter were some of the worst "corsi" players for the Devils, they combined for the only 2 goals.

If you looked at just the corsi for tonight's game, you would be pretty stunned to see the outcome and shocked to see who actually scored for the Devils.

Gio was a +2 today for whatever that's worth, and with only 8 minutes on the ice. From what I saw he made a few nice plays, a few bad ones, mostly average though. Once again, fine for a 4th liner.

[Triumph]

The Devils generally had a high corsi number tonight, the Leafs did not. The Leafs won.

 

Zidlicky and Carter were some of the worst "corsi" players for the Devils, they combined for the only 2 goals.

 

Cool.
 

If you looked at just the corsi for tonight's game, you would be pretty stunned to see the outcome and shocked to see who actually scored for the Devils.

 

Actually if a person knew how to interpret microstats (and indeed had the ability to think stochastically), they wouldn't have any idea who won based on Corsi.  But were they to make that judgment on Corsi alone, which is beyond moronic since who does that, they'd probably think that the Leafs won, but got incredible goaltending to do so, because it's rather rare that a team gets outattempted by that much while losing the game.  Usually teams get outattempted like that while leading in the 3rd period, but that's not really how this game went.
 

Gio was a +2 today for whatever that's worth, and with only 8 minutes on the ice. From what I saw he made a few nice plays, a few bad ones, mostly average though. Once again, fine for a 4th liner.

 

It's worth nothing, just like plus/minus is worth over anything less than an enormous sample (and even then requires giant caveats to interpret properly)

Edited by Triumph, 05 March 2013 - 12:26 AM.

[thefiestygoat]

The Devils generally had a high corsi number tonight, the Leafs did not. The Leafs won.
Zidlicky and Carter were some of the worst "corsi" players for the Devils, they combined for the only 2 goals.

If you looked at just the corsi for tonight's game, you would be pretty stunned to see the outcome and shocked to see who actually scored for the Devils.

Gio was a +2 today for whatever that's worth, and with only 8 minutes on the ice. From what I saw he made a few nice plays, a few bad ones, mostly average though. Once again, fine for a 4th liner.

I don't think Corsi numbers are that valuable unless they are used in a decent sample size. Over a period of a time a team with a high Corsi% should win more often then lose but there are going to be games a team drops despite being better in possession. There are going to be times when players get beat in possession in a single game but are fortunate enough to convert their few offensive chances. Even though tonight Zidlicky +2 and Carter -1 were on the low end of Corsi for Devils players, they still had nearly even or plus numbers.

 

Corsi is valuable as a tool for long term analysis/trends. I'm not saying its the end all be all but I think it has its place in hockey. Despite losing tonight, if the Devils play a ton of games like this in terms of possession, then they should win the majority of them. Hitting 2 posts and Elias being robbed by Reimer completely changed the game along with Moose's poor goaltending.

 

Gionta was on for both Devils goals but did he really do that much to contribute to them? Coming into tonight Gionta's PDO was tied for 14th highest in the league and he benefited from the players around him again tonight. At some point it's due to regress as history shows.

 

I hope Josefson regains his form in Albany. I think you can make the arguement that Carter and Bernier would be even more effective playing on a line together away from Gionta. Both of them have a higher Corsi For% without Gionta, Carter (55.9) and Bernier (55.6) then with, Carter (41.4) and Bernier (45.2).

We're all probably debating this issue too much but I find it fun nonetheless.

[Pepperkorn]

Feisty, I totally agree they need to be used in conjunction.  Then I think, on top of that, we need to look at the roll the player is filling and consider (since even the coach who is with them day in day out can't be certain) the chemistry he is contributing to.

 

Clearly CBGB is filling a need.  Clearly that need just can't be the big time minutes and line match-ups etc.  I mean it's all been said here. It's not sustainable. it is in fact not actually working now -- but it makes fans and probably the team feel a little more confident so...  so it's been working more than anything else lately.  I dont see it as being as effective now... and sure, you kind of want to hold out for Carter's return and sure Bernier is really the go-to guy.  BUT maybe Bernier doesn't need that pressure?  Maybe Gionta takes that off of him even if he's not physically contributing  the the performance of his lines-mates.

 

I do think we create our own luck.  Out of what?  Well -- I think the microstats ( why can't we just call them teeny datasets which is what they are  we gotta make 'em sexy to sports writers I guess) are trying to get to that place - that predictor.  There are pretty decent algorithms out there which can really help predict outcomes with teeny amounts of data - I mean with science you NEED that. I think i still have issues with the quality even if you can justify the sparse data. I think the variables in hockey are too great for the stats,  ehem - microstats, we have now.  But I do think they're valuable.

You just HAVE TO look at the intangibles though when it comes to CBGB-like situations.

 

I'm a big fan of blowing things apart to get to the raw exposed wound that needs fixing.  and in all truth I'm kind of almost there -- BUT then I think the problem isn't that CBGB is kind of sort of producing - the problem is that the other guys are not.  and now no one is resting on our 4th lines production as they may have been when this thread was started.

 

So I dont think that the stats are responsible for the earth shattering prediction "CBGB can't sustain the teams winning record".  All responses agreed that it's a no brainer but just because you can drum up some hard numbers doesn't mean it should be tampered with.  No, in fact, we just need to be thankful and even without the stats - we were bracing for the crash.

 

but because it may be fun   what do you think the stats old and new forewarned us about or are forwarning us about - that isn't self-evident now?

 

and what combined stats are going to get a closer to the algorithm for chemistry.  Math is the language of God - we MUST have a way to mathematically predict chemistry    or not?  This is my internal debate for ALL OF LIFE - not just hockey.  Richard Feynman would tell me shut the fvck up and relax or some combination thereof...

 

(I think the reaction in this thread wasn't so much "microstats are bullsh!t and CBGB is fabulous" as it was "NO sh!t SHERLOCK! So what's the brilliant plan - destroy the one thing that may not be WORKING but puts some bacon on the table?"  Corsis and Fenwicks weren't necessary at all to make the threads obvious point.  Reaction is so huge because microstats offer no solution except finger pointing -- NOT pin pointing as it hides behind in this infant stage microstats are in.

 

I write too much at one time.  Oh well.  Too much going on and not enough time to be patient.