HFBoards

HFBoards (http://hfboards.hockeysfuture.com/index.php)
-   By The Numbers (http://hfboards.hockeysfuture.com/forumdisplay.php?f=241)
-   -   Measuring coaching performance? (http://hfboards.hockeysfuture.com/showthread.php?t=1266813)

PSGJ 10-03-2012 03:58 PM

Measuring coaching performance?
 
I wonder if any efforts have been done to evaluate coaches through advanced stats. It seems to me there is a huge problem in isolating the performance of the coach from the performance of the players. Win/loss record and number of titles aren't really good measurements in my opinion.

billybudd 10-03-2012 06:03 PM

The place to start would be with Tippet's impact on goaltender performance.

But this has probably not been done, since many advanced stats people stubbornly refuse to admit that shot quality exists.

TheDevilMadeMe 10-03-2012 06:07 PM

Quote:

Originally Posted by billybudd (Post 54746589)
The place to start would be with Tippet's impact on goaltender performance.

But this has probably not been done, since many advanced stats people stubbornly refuse to admit that shot quality exists.

I have actually seen a study done on the impact of Pat Burns on goaltender performance, so there is some recognition out there.

billybudd 10-03-2012 06:09 PM

Quote:

Originally Posted by TheDevilMadeMe (Post 54746643)
I have actually seen a study done on the impact of Pat Burns on goaltender performance, so there is some recognition out there.

That's another good place to start.

Talks to Goalposts 10-03-2012 06:54 PM

Quote:

Originally Posted by billybudd (Post 54746589)
The place to start would be with Tippet's impact on goaltender performance.

But this has probably not been done, since many advanced stats people stubbornly refuse to admit that shot quality exists.

That's been done.

http://nhlnumbers.com/2012/8/11/the-dave-tippett-effect

MadLuke 10-03-2012 08:35 PM

It is probably very hard to do, you cannot take short time coach changes result into account because of the new coach effect by itself.

And long term coaching stats for a same team with different coach cannot exist by definition.

Probably very hard to isolate the coach factor in a team and it is probably very very minor. I have no cup with a coach should have be nominated for the Conn Smythe in mind
.

And Guy Carbonneau finished first in the east with a mediocre Canadians team and cannot found a job in the nhl because he was a pretty normal to bad coach you got lucky and had a complete season with no injury.

begbeee 10-04-2012 07:50 AM

Quote:

Originally Posted by MadLuke (Post 54749751)
It is probably very hard to do, you cannot take short time coach changes result into account because of the new coach effect by itself.

And long term coaching stats for a same team with different coach cannot exist by definition.

Probably very hard to isolate the coach factor in a team and it is probably very very minor. I have no cup with a coach should have be nominated for the Conn Smythe in mind.

And Guy Carbonneau finished first in the east with a mediocre Canadians team and cannot found a job in the nhl because he was a pretty normal to bad coach you got lucky and had a complete season with no injury.

1995 Devils.

MadLuke 10-04-2012 12:02 PM

Quote:

Originally Posted by begbeee (Post 54757487)
1995 Devils.

Good one. No Jacque Lemaire no cup for the Devils that year ?

But that is a big statement that the coaching advantage was the biggest factor in that cup win, when you play againt's Scotty Bowman.

I could search for it, but some study (no sure how good they were) showed that school teacher and team coach have very little to no effect in success, versus student and player talents.

begbeee 10-04-2012 01:07 PM

Quote:

Originally Posted by MadLuke (Post 54760555)
Good one. No Jacque Lemaire no cup for the Devils that year ?

But that is a big statement that the coaching advantage was the biggest factor in that cup win, when you play againt's Scotty Bowman.

I could search for it, but some study (no sure how good they were) showed that school teacher and team coach have very little to no effect in success, versus student and player talents.

To some extent. Let Canadian national team (Olympic) play without coach and they still win almost everything on pure talent advantage. Any Scott Bowman or Lemaire wouldn't make them much more better, in fact they could even hurt particular players by their system. Does it work with bunch of 2nd or 3rd tier players? Hell, no! Good coach make them much more better than no coach. Recent example is Slovak national team. Terrible results with weak coaches like Glen Hanlon and immediate success with Vujtek and his system. And everyone would swear that Vujtek has weaker players on the paper compared to past.

EDIT: Recent Devils trilogy came into my mind too. Same team, worse in first half of the season (MacLean), best in second half (Lemaire) and SC finalist next year (DeBoer). There is a good reason to say MacLean is bad coach. Harder question is..are Lemaire and DeBoer such great coaches or any average coach would lead them to the same (or similar) results.

EDIT2: Canada - Czech semifinal in Nagano. Good example of outcoaching.

MadLuke 10-04-2012 01:39 PM

Quote:

Originally Posted by begbeee (Post 54761927)
EDIT: Recent Devils trilogy came into my mind too. Same team, worse in first half of the season (MacLean), best in second half (Lemaire) and SC finalist next year (DeBoer). There is a good reason to say MacLean is bad coach. Harder question is..are Lemaire and DeBoer such great coaches or any average coach would lead them to the same (or similar) results.

How much Kowalchuck is responsable of that vs Lemaire (and Parize return ?)

I will agree that bad coaching can make a team going down no doubt, but how much outcoaching can you have for an edge vs other professional coached team.

IT will be very hard to have the coaching factor with good sample size, would like to see study about coach influence, but they are probably hard to do.

billybudd 10-05-2012 02:52 PM

Quote:

Originally Posted by Talks to Goalposts (Post 54747601)

Interesting, thanks.

The Bob Cole 10-05-2012 08:19 PM

Quote:

Originally Posted by MadLuke (Post 54762609)
How much Kowalchuck is responsable of that vs Lemaire (and Parize return ?)

I will agree that bad coaching can make a team going down no doubt, but how much outcoaching can you have for an edge vs other professional coached team.

IT will be very hard to have the coaching factor with good sample size, would like to see study about coach influence, but they are probably hard to do.

A coaches influence and perhaps most importantly, contribution to a team, often happens outside of 'the game'. It happens in practice and in the real world interactions with the players. It's nearly impossible to quantify those types of relationships in a study, but you do hear stories of the great leaders in every sport who could bring together players to become a better team.

When we hear of strategically 'outcoaching' in sports (hockey), it's more often that the grass is greener on the other side than not.

Iain Fyffe 10-05-2012 10:30 PM

Quote:

Originally Posted by billybudd (Post 54746589)
The place to start would be with Tippet's impact on goaltender performance.

But this has probably not been done, since many advanced stats people stubbornly refuse to admit that shot quality exists.

Try again. The "advanced stats people" were the first to try to quantify the effect of shot quality systematically. Many of them now believe that the effect of shot quality on save percentage is relatively small and only has a significant effect in extreme cases. This conclusion is based on the data.

Which is a far cry, of course, from stubbornly refusing to admit it exists. It's also a more nuanced, less inflammatory description of the situation.

PSGJ 10-08-2012 05:32 PM

Here's an idea. Calculate expected performance for all the players who have played for a coach. Do this by looking at their careers before and after. Then look at how many under performed and over performed with the coach in question. Basically do the Mike Smith thing, but with all the players. Now, obviously for any given player there is a lot of chance involved, but with enough players it should regress towards the mean.

The problem with this is of course that a massive amount of calculations have to be made, but I could see it done in a professional setting. For instance an NHL team doing due diligence before hiring.

There's also a problem in how you count expected performance of course, but it seems to me that could be worked out.


All times are GMT -5. The time now is 11:47 AM.

vBulletin Copyright ©2000 - 2016, Jelsoft Enterprises Ltd.
HFBoards.com, A property of CraveOnline, a division of AtomicOnline LLC ©2009 CraveOnline Media, LLC. All Rights Reserved.