Coach Discussion: All Purpose Coaching Thread Part III

[Dayofthedogs]

That isn't the point. He says he wants to get them in a game. He says he doesn't like to leave players sit too long. I call BS. He could put them in at any time for Sbisa. He very often allows players to sit in the PB far too long. It is among his worst faults. Yet he frequently says that he doesn't like doing it. Bull!

Not wanting to change a lineup while you are winning can be carried to ridiculous extremes. With Mau it appears to be like a superstition.

The extreme being you win all the games and never change the lineup?

 

[Gm0ney]

You know what is interesting? The 3 games we won this week against Florida,Tampa and Nashville we got out shot and I am going to guess they all had more "high danger" chances then us. But in the game we lost we out shot Colorado. I wonder if we had more "high danger" chances then them?

I believe the first thing they teach you at Fancystats U is that teams playing from behind often outshoot their opponents. They call it "score effects".

 

[WPGChief]

I see we are in the particular phase where advanced stats clearly aren’t good anymore because we are winning despite how bad some of them are, whereas before when they were (still) bad and we were losing what advanced stats were telling us was good.

Some people need to knock off the trope that just because we are positive in Corsi For % or Expected Goals For % that it must mean the team should win and vice versa, but if not then that what the hell do we even use them for. We’ve already discussed ad nauseum that these statistics are more reliable and better predictors of future performance than others - note, “reliable” and “better” does not mean perfect. The fact that we won a game where all hell broke loose against San Jose should have obviously told you that.

A better way of thinking of it is that being better in these areas increases your chances of the coin flipping heads instead of tails. So, take for example this recent 4-game road trip. I’m making up the numbers in my head right now, but say the Jets had the respective chances of winning each game: 52%, 47%, 49%, and 42%. Go throw that in a random generator and click submit a few times and I’ll bet you’re bound to get a sequence where the coin lands heads 4 times in a row.

It’s not impossible. It’s just improbable to continue. There’s a chance that we could go 0-7-1 in the next 8 games as easily as we’ve gone 7-1-1. If you’re a betting person, you’re opting for the former instead of the latter right now. Doesn’t mean you’ll win.

 

[BoneDocUK]

That isn't the point. He says he wants to get them in a game. He says he doesn't like to leave players sit too long. I call BS. He could put them in at any time for Sbisa. He very often allows players to sit in the PB far too long. It is among his worst faults. Yet he frequently says that he doesn't like doing it. Bull!

Not wanting to change a lineup while you are winning can be carried to ridiculous extremes. With Mau it appears to be like a superstition.

We have heard versions of this PB line about various players for years — including this year, with reference to Heinola and Gus (I’d actually failed to note that Potato was back from injury).

Given that Sbisa has barely practiced or played at the NHL level in the past 2 years and has apparently welded down his spot, maybe the PB is less harmful to gritty vets than youngsters?

I get the logic of not messing with a winning lineup — but that shouldn’t be mistaken for a related homily — if it ain’t broke, don’t fix it — IMO.

 

[Mortimer Snerd]

It's definitely a superstition. Both could be true - he wants to get them in because he doesn't like them sitting too long, but he's also not wanting to make changes to a lineup that's winning, despite who it is. If the team wasn't winning, they wouldn't be sitting too long.

You realize that superstition is stupidity, right?

It makes sense to not do a massive shakeup of a winning lineup, but refusing to swap out one borderline #8/9 D man for another borderline #8/9 D man who might actually be slightly better is just silly. He could be rotating those 3 Dmen every game, regardless of who wins or loses. If one screws up worse than the other 2, leave him out of the rotation. If one starts playing like a legitimate NHL starter you can end the rotation. But to say that you want to get the other 2 in but can't because - superstition? Or even just because the current lineup is winning and you don't want to mess with it when we are talking about the bottom of the roster Dman who isn't actually playing particularly well. Not knocking Sbisa's play here. He has not been too bad, considering. But Bitetto was at least as good when he was playing. Dahlstrom, I'm not so sure of.

But we don't dare to touch anything unless it messes with our luck.

 

[tbcwpg]

You realize that superstition is stupidity, right?

It makes sense to not do a massive shakeup of a winning lineup, but refusing to swap out one borderline #8/9 D man for another borderline #8/9 D man who might actually be slightly better is just silly. He could be rotating those 3 Dmen every game, regardless of who wins or loses. If one screws up worse than the other 2, leave him out of the rotation. If one starts playing like a legitimate NHL starter you can end the rotation. But to say that you want to get the other 2 in but can't because - superstition? Or even just because the current lineup is winning and you don't want to mess with it when we are talking about the bottom of the roster Dman who isn't actually playing particularly well. Not knocking Sbisa's play here. He has not been too bad, considering. But Bitetto was at least as good when he was playing. Dahlstrom, I'm not so sure of.

But we don't dare to touch anything unless it messes with our luck.

Yes, superstition isn't based on anything logical. I just think being critical of this particular lineup choice is nitpicking for nitpicking's sake.

 

[Mortimer Snerd]

The extreme being you win all the games and never change the lineup?

If that was the potential payoff for never changing anything, I would be all for it. Stupid superstition or not.

But if it makes sense to never change anything while you are winning, it also makes sense to change everything as soon as you lose 1 game. Of course neither makes any sense at all. Make each lineup decision on its merits. Not on some silly fear of change. I wonder if Mau is still wearing the same underwear as he had on for the first of these 3 wins? Makes about as much sense.

 

[Mortimer Snerd]

We have heard versions of this PB line about various players for years — including this year, with reference to Heinola and Gus (I’d actually failed to note that Potato was back from injury).

Given that Sbisa has barely practiced or played at the NHL level in the past 2 years and has apparently welded down his spot, maybe the PB is less harmful to gritty vets than youngsters?

I get the logic of not messing with a winning lineup — but that shouldn’t be mistaken for a related homily — if it ain’t broke, don’t fix it — IMO.

I don't particularly care if he plays Sbisa or Bitetto. Neither is good, although I think Bitetto was less bad, but maybe I'm wrong there. They are close enough to each other that it just doesn't matter.

My point is about him saying one thing and doing another - on very flimsy grounds.

 

[Mortimer Snerd]

Yes, superstition isn't based on anything logical. I just think being critical of this particular lineup choice is nitpicking for nitpicking's sake.

As I just posted in reply to another, I don't care in the least which one he plays. They are interchangeable sub-replacement level players. It is the BS of what he says that I am taking issue with. Its why I quit listening to him years ago. He is full of BS.

 

[Howard Chuck]

I see we are in the particular phase where advanced stats clearly aren’t good anymore because we are winning despite how bad some of them are, whereas before when they were (still) bad and we were losing what advanced stats were telling us was good.

Some people need to knock off the trope that just because we are positive in Corsi For % or Expected Goals For % that it must mean the team should win and vice versa, but if not then that what the hell do we even use them for. We’ve already discussed ad nauseum that these statistics are more reliable and better predictors of future performance than others - note, “reliable” and “better” does not mean perfect. The fact that we won a game where all hell broke loose against San Jose should have obviously told you that.

A better way of thinking of it is that being better in these areas increases your chances of the coin flipping heads instead of tails. So, take for example this recent 4-game road trip. I’m making up the numbers in my head right now, but say the Jets had the respective chances of winning each game: 52%, 47%, 49%, and 42%. Go throw that in a random generator and click submit a few times and I’ll bet you’re bound to get a sequence where the coin lands heads 4 times in a row.

It’s not impossible. It’s just improbable to continue. There’s a chance that we could go 0-7-1 in the next 8 games as easily as we’ve gone 7-1-1. If you’re a betting person, you’re opting for the former instead of the latter right now. Doesn’t mean you’ll win.

Fair enough. Since this is the coaching thread, my thought would be how would you use these numbers to make your team better? It's one thing to keep track of these stats, but it's completely another to be able to make changes to make the numbers (and the team) better.

I think this should be the ultimate use of stats.

 

[ffh]

As I just posted in reply to another, I don't care in the least which one he plays. They are interchangeable sub-replacement level players. It is the BS of what he says that I am taking issue with. Its why I quit listening to him years ago. He is full of BS.

what are you railing against. Team is playing well he doesnt want to mess with it . If you quit listening to him years ago how do you know he said what he said.

 

[Dayofthedogs]

If that was the potential payoff for never changing anything, I would be all for it. Stupid superstition or not.

But if it makes sense to never change anything while you are winning, it also makes sense to change everything as soon as you lose 1 game. Of course neither makes any sense at all. Make each lineup decision on its merits. Not on some silly fear of change. I wonder if Mau is still wearing the same underwear as he had on for the first of these 3 wins? Makes about as much sense.

I know. I was just having fun with it. I'll work on better material.

 

[Dayofthedogs]

I see we are in the particular phase where advanced stats clearly aren’t good anymore because we are winning despite how bad some of them are, whereas before when they were (still) bad and we were losing what advanced stats were telling us was good.

Some people need to knock off the trope that just because we are positive in Corsi For % or Expected Goals For % that it must mean the team should win and vice versa, but if not then that what the hell do we even use them for. We’ve already discussed ad nauseum that these statistics are more reliable and better predictors of future performance than others - note, “reliable” and “better” does not mean perfect. The fact that we won a game where all hell broke loose against San Jose should have obviously told you that.

A better way of thinking of it is that being better in these areas increases your chances of the coin flipping heads instead of tails. So, take for example this recent 4-game road trip. I’m making up the numbers in my head right now, but say the Jets had the respective chances of winning each game: 52%, 47%, 49%, and 42%. Go throw that in a random generator and click submit a few times and I’ll bet you’re bound to get a sequence where the coin lands heads 4 times in a row.

It’s not impossible. It’s just improbable to continue. There’s a chance that we could go 0-7-1 in the next 8 games as easily as we’ve gone 7-1-1. If you’re a betting person, you’re opting for the former instead of the latter right now. Doesn’t mean you’ll win.

Advance stats are great. Certainly good as both a descriptor and predictor. I think some of us are just okay with being an outlier for now. I doubt very many of us believe this is a recipe for long term success but the roster has holes.

 

[KingBogo]

Plenty of players have not earned what they have received under Moe.

I am happy for Laine's much improved play but I feel that plenty of other players have been gifted roster spots due to Maurice's biases and not actual performance.

What players are you thinking of?

 

[LowLefty]

Sbisa is no Bobby Orr, but when the team is 7-1-1 in November, most coaches are hesitant to make changes unless it's an impact player returning from injury.

Not saying Sbisa is the reason they're winning in the slightest, but Bitetto and Dahlstrom are no impact players (at least positive impact players).

This makes sense to me -
You have a team that is winning and Mo wants to give his PB players a little ice time - but he doesn't want to mess with success.

I don't have an issue with how he's handling his bottom 6. He's sticking with the players that are winning the games rather than a perpetual blender in an effort to work PB players into the line up when the current bunch seem to be getting it done.

If he did rotate this bunch and we started a slide, where do you think the conversation would go next? How often do you see coaches messing with a line up when they are winning?

 

[Mortimer Snerd]

I know. I was just having fun with it. I'll work on better material.

I got it. I had toyed with the same thought but didn't do anything with it. Like running with the hot goalie until he loses. What if he never loses? That would never do. Your backup goalie would get stale.

 

[Mbraunm]

First half of last season the Jets were winning despite consistently getting out chanced and out shot. Everyone was happy but the underlying numbers suggested we were due for a crash down to earth.

This season feels similar. We've won a lot of games in extra time, and have won a pile of 1 goal games. It wont take much for the tide to turn against us, even a small dip in Helly's performance and our win loss record could easily be reversed.

I fear that by the end of the season we will likely crash back down to earth, which may then mean the end of the Maurice era.

I sincerely hope so!

 

[Guffman]

So, which alternative head coach would have done a better job at coaching up our AHL defence? Where should the Jets be in the standings with your superior head coach?

If you were to speculate which head coaches would be available this offseason, that it would be plausible for us to sign, who would you want? And what basis do you have to think he would do better with this team?

 

[surixon]

What players are you thinking of?

Tanev when he first started here. Hendricks, Chariot, Psia, Bourque, Letestu, Thor etc.... He has no issues playing his type of grinders without them earning anything.

I have no issue with players earning things, but the criteria is subjective and a coaches biases will play into those decisions.

Many businesses these days are shrinking their hierarchies, I think sports teams need to do the same. Ultimately you should be trying optimize the lineup, not get involved in office politics and build based on seniority.

 

[WPGChief]

Fair enough. Since this is the coaching thread, my thought would be how would you use these numbers to make your team better? It's one thing to keep track of these stats, but it's completely another to be able to make changes to make the numbers (and the team) better.

I think this should be the ultimate use of stats.

One thing they do really well (in this context) is challenge your biases. It's been said many, many times before but as humans we absolutely suck in terms of basing our opinions on something or someone on an anchor point - for defensemen, it's often referred to as "the big mistake". In that sense, our perspective on a player can be skewed based on what might have been 5 seconds in a game, and even into the season. In terms of your "eye test", it's easy for scouts and coaches to scour over video, but even then you're probably looking more at clips instead of the overall game - nevertheless, going over video is likely the best method to help coach the individual to point the specifics of what can be done differently, or better.

In terms of statistics (or lies, damned lies, and statistics), it helps you set benchmarks and key performance indicators (as well as be able to perform some better overall "guess-sumptions" on how nearly 700 players in the NHL are performing in the case of trades). In-game, you can be biased towards the length of time a team spends in a zone (even though we already know through data extended zone time doesn't always mean more goals for). But, for example, you can review the game and see while there was possession, it wasn't productive possession, meanwhile the opponent has multiple rush chances against you though they haven't scored. Do you keep rolling with that line as they are or do you review video with them to shift them towards a style more conducive to scoring goals?

This is only a quick example. There's a lot more applications, but benchmarking should be a key one. As with any business project, you don't want to go from 0 to 100% completion without being reliably able to say that the project will do what you want it to do at completion? Same thing in terms of using statistics or analytics in hockey.

 

[Howard Chuck]

One thing they do really well (in this context) is challenge your biases. It's been said many, many times before but as humans we absolutely suck in terms of basing our opinions on something or someone on an anchor point - for defensemen, it's often referred to as "the big mistake". In that sense, our perspective on a player can be skewed based on what might have been 5 seconds in a game, and even into the season. In terms of your "eye test", it's easy for scouts and coaches to scour over video, but even then you're probably looking more at clips instead of the overall game - nevertheless, going over video is likely the best method to help coach the individual to point the specifics of what can be done differently, or better.

In terms of statistics (or lies, damned lies, and statistics), it helps you set benchmarks and key performance indicators (as well as be able to perform some better overall "guess-sumptions" on how nearly 700 players in the NHL are performing in the case of trades). In-game, you can be biased towards the length of time a team spends in a zone (even though we already know through data extended zone time doesn't always mean more goals for). But, for example, you can review the game and see while there was possession, it wasn't productive possession, meanwhile the opponent has multiple rush chances against you though they haven't scored. Do you keep rolling with that line as they are or do you review video with them to shift them towards a style more conducive to scoring goals?

This is only a quick example. There's a lot more applications, but benchmarking should be a key one. As with any business project, you don't want to go from 0 to 100% completion without being reliably able to say that the project will do what you want it to do at completion? Same thing in terms of using statistics or analytics in hockey.

I always enjoy the thought you put into responses. Nice explanation. I still wonder if the Jets consistently do what you describe above. Or if they do, do they do it soon enough.

 

[KingBogo]

Tanev when he first started here. Hendricks, Chariot, Psia, Bourque, Letestu, Thor etc.... He has no issues playing his type of grinders without them earning anything.

I have no issue with players earning things, but the criteria is subjective and a coaches biases will play into those decisions.

Many businesses these days are shrinking their hierarchies, I think sports teams need to do the same. Ultimately you should be trying optimize the lineup, not get involved in office politics and build based on seniority.

I played pretty competitive hockey when I was younger and knew a couple guys that grinded out professional careers. They would be the guys this board would s*** on day after day late in their careers. They were never gifted a thing and spent everyday working as hard as they could to maintain a career. IMO they earned every second they spent on the ice.

 

[surixon]

I played pretty competitive hockey when I was younger and knew a couple guys that grinded out professional careers. They would be the guys this board would s*** on day after day late in their careers. They were never gifted a thing and spent everyday working as hard as they could to maintain a career. IMO they earned every second they spent on the ice.

There in lies the subjective nature and inherent biases that exist with regards to determining what is "earned". Rewarding effort more so then results has been prevalent for a long time but doesn't mean it is what should be happening.

 

[WPGChief]

I always enjoy the thought you put into responses. Nice explanation. I still wonder if the Jets consistently do what you describe above. Or if they do, do they do it soon enough.

People ask me this a lot and the answer is a confident yes.

Now, do they look at the same things I would in terms of judging whether or not my team is performing how I want them to perform, and whether or not it is garnering the results I expected? Probably not (looking at you, extended offensive zone time).

 

[BoneDocUK]

I don't particularly care if he plays Sbisa or Bitetto. Neither is good, although I think Bitetto was less bad, but maybe I'm wrong there. They are close enough to each other that it just doesn't matter.

My point is about him saying one thing and doing another - on very flimsy grounds.

I agree, of course -- but I think holding PoMo to anything he says in a press scrum, even the old chestnuts, like "earning the keys to the car" or fretting about how to get those durned PB guys into a game, is like trying to bail out a boat with a sieve.

KingBogo, in a recent comparison between Babcock and Maurice on these very boards, lauded PM as first and foremost a "communicator," and I'd agree. He's a very good communicator. And he's also a politician. I think it's important that his players trust him -- as many seem to -- but I don't, which is as it should be, perhaps, for a fan. I think there's a lot more content in a Vincent interview, but then no one pays the same kind of attention to an AHL coach. I expect you learn a lot about media management in HC school.