Jump to content

Sunny Mehta hired by Devils as Director of Analytics


Triumph

Recommended Posts

Well something was bugging me when Lou confirmed that Mehta was hired to run the analytics department, and it bugged me again while reading James Mirtle's article on the Oiler's hiring of former blogger Tyler Dellow.

Here is what Dallas Eakins says about Dellow:

http://www.theglobeandmail.com/sports/hockey/oilers-hiring-move-deepens-leagues-interest-in-analytics/article20023848/?cmpid=rss1&click=dlvr.it

While this is what Lou said to Gulitti:

In one situation you have a coach that was active in the hiring of this person, while in the other you have owners that may have forced (certainly encouraged) a department for analytics. In one situation you have a coach ready to use this person frequently and seek out advice, while in the other you have a GM that is overseeing the data that the coach receives and may or may not share everything with a coach.

I don't like that at all. We know Lou is very hands on, but I think the Edmonton situation is a lot better. At least we know their coach wants his opinion and beliefs challenged. Maybe Pete does too, but he'll have to go through Lou first.

Dude, you really need to lay off caffeine. Basically, you're bullish on an organization that has been the epitome of futility and incompetence for the past decade because the coach says he wants to listen to an analytics guy. On the other, you have the one of the most successful GMs to ever be part of the game, that is just saying that he's not going to have someone who for all we know has never put on ice skates in his entire life tell the coach what to do, but will use his imput accordingly. I mean, it might have been one thing if people were right about Lou giving Marty another year or two, and making Schneider disgruntled, but, of course that never came to pass.

Analytics appears to be a nice thing, and it's certainly an improvement over some of the old fuddy duddies (which Lou never has been). But it's just a little bit strange that within the course of a summer, it seems that some folks believe you're destined to win the Cup if you do everything an analytics person says, while you'll be a bottom feeder forever if you essentially don't cede control of the team to him. The teams that have won the past three Cups have done so on being in the right place at the right time with top ten draft picks, a luxury Lou has had once in the past 20 years, not because of their analytics departments.

Link to comment
Share on other sites

Dude, you really need to lay off caffeine. Basically, you're bullish on an organization that has been the epitome of futility and incompetence for the past decade because the coach says he wants to listen to an analytics guy. On the other, you have the one of the most successful GMs to ever be part of the game, that is just saying that he's not going to have someone who for all we know has never put on ice skates in his entire life tell the coach what to do, but will use his imput accordingly. I mean, it might have been one thing if people were right about Lou giving Marty another year or two, and making Schneider disgruntled, but, of course that never came to pass.

Analytics appears to be a nice thing, and it's certainly an improvement over some of the old fuddy duddies (which Lou never has been). But it's just a little bit strange that within the course of a summer, it seems that some folks believe you're destined to win the Cup if you do everything an analytics person says, while you'll be a bottom feeder forever if you essentially don't cede control of the team to him. The teams that have won the past three Cups have done so on being in the right place at the right time with top ten draft picks, a luxury Lou has had once in the past 20 years, not because of their analytics departments.

 

All I am saying is I like Eakins approach. I understand what Lou is thinking, but his approach might need some changing. If Deboer wants to go over certain things with Sunny and get some numbers from him, that would be great as well. Right now I'm not sold that Lou has fully embraced this movement. That's all. I think incorporating analytics is best if the coach is heavily involved. It isn't about someone who "has never put on ice skates in his entire life tell the coach what to do", but rather a coach looking at players and situations from a different angle and getting that information."

 

Dude, we can debate the fuddy-duddy comment. There is definitely a good argument that Lou is that. We aren't talking winning the Stanley Cup right now either for this team, but how about just improving and getting back to being a playoff team. That's what the Devils should be looking to do. That's what Edmonton is trying to do and same with the Leafs. That's a nice start.

Link to comment
Share on other sites

All I am saying is I like Eakins approach. I understand what Lou is thinking, but his approach might need some changing. If Deboer wants to go over certain things with Sunny and get some numbers from him, that would be great as well. Right now I'm not sold that Lou has fully embraced this movement. That's all. I think incorporating analytics is best if the coach is heavily involved. It isn't about someone who "has never put on ice skates in his entire life tell the coach what to do", but rather a coach looking at players and situations from a different angle and getting that information."

 

Dude, we can debate the fuddy-duddy comment. There is definitely a good argument that Lou is that. We aren't talking winning the Stanley Cup right now either for this team, but how about just improving and getting back to being a playoff team. That's what the Devils should be looking to do. That's what Edmonton is trying to do and same with the Leafs. That's a nice start.

 

I don't agree, but only because Mehta and Dellow do totally different things.  Tyler's site has since vanished into the ether, but basically over the last year the biggest thing he did was marry video analysis to statistical analysis.  He'd look at certain data, then try to figure out from video why the numbers were as they were.  That's stuff that a coach can use.  Mehta's a quant - he's a guy who's worked with poker and finance, and he's much more of a numbers-only sort of guy.  That's not to say that the Devils' analytics department won't do video analysis, but Dellow is very specifically not a quant guy - he's an intelligent person who understands how randomness functions, but I don't think you'd hire him on that basis alone.  His video work was the real gem, and that work was more for coaches than GMs.

Edited by Triumph
Link to comment
Share on other sites

All I am saying is I like Eakins approach. I understand what Lou is thinking, but his approach might need some changing. If Deboer wants to go over certain things with Sunny and get some numbers from him, that would be great as well. Right now I'm not sold that Lou has fully embraced this movement. That's all. I think incorporating analytics is best if the coach is heavily involved. It isn't about someone who "has never put on ice skates in his entire life tell the coach what to do", but rather a coach looking at players and situations from a different angle and getting that information."

Dude, we can debate the fuddy-duddy comment. There is definitely a good argument that Lou is that. We aren't talking winning the Stanley Cup right now either for this team, but how about just improving and getting back to being a playoff team. That's what the Devils should be looking to do. That's what Edmonton is trying to do and same with the Leafs. That's a nice start.

The irony of course is that what DeBoer ends up getting out of his team is high possession numbers, something virtually every analytics person will tell you to look for and try to maximize.

Link to comment
Share on other sites

The irony of course is that what DeBoer ends up getting out of his team is high possession numbers, something virtually every analytics person will tell you to look for and try to maximize.

 

Three things can kill the advantage of high possession numbers, as we've seen:

 

1) Not creating or finishing off the opportunities that those numbers should lead to.

2) Getting nearly 2300 minutes of .901 save% goaltending (2013-14) and a full season of sub-.900 puck-stopping (2013). 

3) Losing every single shootout (2013-2014) or too many of them (2013). 

 

I expect a mild improvement in #1, a significant one in #2, and as for #3...it can only get better, though probably not insanely so...if the Devils win 30-40% of their shootouts next season, I'll happily take it

Link to comment
Share on other sites

Three things can kill the advantage of high possession numbers, as we've seen:

1) Not creating or finishing off the opportunities that those numbers should lead to.

2) Getting nearly 2300 minutes of .901 save% goaltending (2013-14) and a full season of sub-.900 puck-stopping (2013).

3) Losing every single shootout (2013-2014) or too many of them (2013).

I expect a mild improvement in #1, a significant one in #2, and as for #3...it can only get better, though probably not insanely so...if the Devils win 30-40% of their shootouts next season, I'll happily take it

1. I believe they'll tell you there's no phenomenon of "finishing" better than others, except for a few special players like Stamkos.

2. Not an issue anymore, and it was a unique situation while it lasted.

3. Will also tell you coach and GM has no control over it other than to do the obvious and not have Salvador be in your top 3.

DeBoer's big issue was giving too many starts to Marty, but again it isn't relevant anymore and Marty is the only one that would get away with it.

So again, the analytics people ought to be saying that DeBoer is doing what he should. Not sure if you're really disagree teeing with that though.

Link to comment
Share on other sites

1. I believe they'll tell you there's no phenomenon of "finishing" better than others, except for a few special players like Stamkos.

2. Not an issue anymore, and it was a unique situation while it lasted.

3. Will also tell you coach and GM has no control over it other than to do the obvious and not have Salvador be in your top 3.

DeBoer's big issue was giving too many starts to Marty, but again it isn't relevant anymore and Marty is the only one that would get away with it.

So again, the analytics people ought to be saying that DeBoer is doing what he should. Not sure if you're really disagreeing with that though.

 

Re:  the bolded...I'm not.  Like I said, #2 will improve significantly.  The goaltending will be much better now that a large chunk of the minutes won't be played at near-bottom level. 

 

Re:  #1, Lou did what he could to improve the offense.  If the Devils finish somewhere in around 20th in GF, I'll take it.  Last year they were 27th. 

 

Re:  #3, I've been over the shootout thing before...it's one of those things a GM can't really do much to improve on, you just hope your goalie and some of your skaters happen to be good in them. 

 

Basically, if the Devils keep it up with the strong possession numbers, they should make the playoffs...#2 is probably enough to get them there by itself.

Link to comment
Share on other sites

 

I don't agree, but only because Mehta and Dellow do totally different things.  Tyler's site has since vanished into the ether, but basically over the last year the biggest thing he did was marry video analysis to statistical analysis.  He'd look at certain data, then try to figure out from video why the numbers were as they were.  That's stuff that a coach can use.  Mehta's a quant - he's a guy who's worked with poker and finance, and he's much more of a numbers-only sort of guy.  That's not to say that the Devils' analytics department won't do video analysis, but Dellow is very specifically not a quant guy - he's an intelligent person who understands how randomness functions, but I don't think you'd hire him on that basis alone.  His video work was the real gem, and that work was more for coaches than GMs.

 

Lets not put a cap on what Sunny can and can't do. If there is a department that Sunny runs, who is to say that they won't be doing something similar? 
 

I don't mind the information going through Lou. Heck, it might plant the seed in his mind regarding who is a useful player and who isn't, which might affect him in trades and FA.

 

Can't it be both. Someone in the department can prepare reports to Lou for the things you mentioned, and someone can have reports going to Deboer. We've learned how the scouting operates with Conte taking in all information and doing as he pleases, and it looks like it'll be the same here. Not a big fan out of that methodology. 

Link to comment
Share on other sites

Dude, you really need to lay off caffeine. Basically, you're bullish on an organization that has been the epitome of futility and incompetence for the past decade because the coach says he wants to listen to an analytics guy. On the other, you have the one of the most successful GMs to ever be part of the game, that is just saying that he's not going to have someone who for all we know has never put on ice skates in his entire life tell the coach what to do, but will use his imput accordingly. I mean, it might have been one thing if people were right about Lou giving Marty another year or two, and making Schneider disgruntled, but, of course that never came to pass.

Analytics appears to be a nice thing, and it's certainly an improvement over some of the old fuddy duddies (which Lou never has been). But it's just a little bit strange that within the course of a summer, it seems that some folks believe you're destined to win the Cup if you do everything an analytics person says, while you'll be a bottom feeder forever if you essentially don't cede control of the team to him. The teams that have won the past three Cups have done so on being in the right place at the right time with top ten draft picks, a luxury Lou has had once in the past 20 years, not because of their analytics departments.

 

and how is that more fair to compare a GM who's been there for 2 years and had nothing to do with the decades of suckness before he got there to Lou who's been there forever? 

Link to comment
Share on other sites

Lets not put a cap on what Sunny can and can't do. If there is a department that Sunny runs, who is to say that they won't be doing something similar?

 

If they do something similar, it should end up with the coaching staff.  My point is that that's not where his expertise lies.

 

 

 

 

Can't it be both. Someone in the department can prepare reports to Lou for the things you mentioned, and someone can have reports going to Deboer. We've learned how the scouting operates with Conte taking in all information and doing as he pleases, and it looks like it'll be the same here. Not a big fan out of that methodology.

 

Edited by Triumph
Link to comment
Share on other sites

and how is that more fair to compare a GM who's been there for 2 years and had nothing to do with the decades of suckness before he got there to Lou who's been there forever? 

 

It's a GM that hasn't done anything anywhere, a coach that hasn't done anything anywhere, and despite being handed an organization that had three number one picks in the past four or five years, it is arguably the worst team in hockey.    I'm not even really meaning to knock on them so much, other than to just say that people ought to cool their jets a little bit over the analytics craze.  And that's especially the case when Lou actually hired one, but just wants to manage it somewhat differently than the Oilers apparently will, and for reasons that we don't really know. 

Link to comment
Share on other sites

It's a GM that hasn't done anything anywhere, a coach that hasn't done anything anywhere, and despite being handed an organization that had three number one picks in the past four or five years, it is arguably the worst team in hockey.    I'm not even really meaning to knock on them so much, other than to just say that people ought to cool their jets a little bit over the analytics craze.  And that's especially the case when Lou actually hired one, but just wants to manage it somewhat differently than the Oilers apparently will, and for reasons that we don't really know. 

 

oh i agree with all that. Just don't think it was a fair comparison bringing the "decades of suckness" as a background when the "accused" had nothing to do with it directly

Link to comment
Share on other sites

I don't think there should be reports going to DeBoer because I think the value of analytics is largely in valuing other players instead of your own.  You've skirted what I said which is that one guy's work deals very clearly with tactics and the other guy's work very much does not.  Eric Tu.lsky is also working for some team, and I don't think what he does should end up with the coach first either.

 

Agree 100%. DeBoer should not be involved in the day-to-day of what Sunny Mehta is doing.

 

What Sunny Mehta is doing should inform what players we trade for, sign, or let go. And what Sunny Mehta is doing should inform whether DeBoer is the right coach at the end of the season. DeBoer should otherwise be doing the same job he's been doing.

 

I view it like an HR person at a company. That person's job isn't to consult with every department about the job they're doing every week. The good HR person will lead to the right hires that make the company better down the road, it's a process.

Edited by DJ Eco
Link to comment
Share on other sites

Agree 100%. DeBoer should not be involved in the day-to-day of what Sunny Mehta is doing.

 

What Sunny Mehta is doing should inform what players we trade for, sign, or let go. And what Sunny Mehta is doing should inform whether DeBoer is the right coach at the end of the season. DeBoer should otherwise be doing the same job he's been doing.

 

I view it like an HR person at a company. That person's job isn't to consult with every department about the job they're doing every week. The good HR person will lead to the right hires that make the company better down the road, it's a process.

 

Are you serious about the HR person comparison?  You are willing to give a guy who has zero hockey experience that much power including when it comes to whether DeBoer is the right person for the job just by looking at a bunch of numbers on a sheet?

 

Sunny should be a consultant but not a decision maker.

Link to comment
Share on other sites

Are you serious about the HR person comparison?  You are willing to give a guy who has zero hockey experience that much power including when it comes to whether DeBoer is the right person for the job just by looking at a bunch of numbers on a sheet?

 

Sunny should be a consultant but not a decision maker.

 

Exactly. You can NEVER strictly look at only numbers to make decisions and that's all this guy will be doing. Thats fvcking retarded. For every stats you have to look at the situation and a bunch of other factors and that guy is not qualify to judge all that.

 

Deboer will use him to back up some stuff he sees on the ice and other way around he may bring to Deboer informations that may have went under the radar. It's something you have to work hand in hand absolutely.

 

Link to comment
Share on other sites

Are you serious about the HR person comparison?  You are willing to give a guy who has zero hockey experience that much power including when it comes to whether DeBoer is the right person for the job just by looking at a bunch of numbers on a sheet?

 

Sunny should be a consultant but not a decision maker.

 

I guess I didn't mean it too literally, but yeah, Lou would be the HR manager, and Sunny his assistant/advisor haha

Link to comment
Share on other sites

I would have hired someone with a stats background, but whatever. 

 

Masters in music, what?

Music is math.  Additionally: performing artists are focused on high quality production. Highly disciplined, EXTREME work ethic, creative thinkers.  If you ever have the chance to hire performing artists do it.  It's a lesson learned time and time again by the corporate world.  The only problem is when it appears failure is eminent, they might bale out to seek out a more successful endeavor - or worse yet just burn out and hang on in a semi-vegetative state merely collecting a paycheck until someone makes their decision to move on for them.

Link to comment
Share on other sites

Are you serious about the HR person comparison? 

 

Sunny should be a consultant but not a decision maker.

HR isn't a decision maker.  Strategic HR (not benefits mgmt, payroll employee relations etc) offers performance metrics for management to use as part of their personnel decision making process.  Would it be wrong of me to say "duh?" at this point?

 

EDIT: Oh sh!t... I didn't even see who wrote that post -- great now it's going to be all about how I hate him blah blah blah.  :argh:

Edited by Pepperkorn
Link to comment
Share on other sites

Exactly. You can NEVER strictly look at only numbers to make decisions and that's all this guy will be doing. Thats fvcking retarded. For every stats you have to look at the situation and a bunch of other factors and that guy is not qualify to judge all that.

 

Deboer will use him to back up some stuff he sees on the ice and other way around he may bring to Deboer informations that may have went under the radar. It's something you have to work hand in hand absolutely.

You are responding to a response not the actual post.  Bad habit -- a way to get yourself all fired up about nothing -- just sayin'

 

Me, if I'm DeBoer -- I dont talk to this guy.  It's not my problem.  I talk to Lou.  Lou talks to this guy to see if there is any correlation between his metrics and the players performance as observed by DeBoer -- then discuss this with the coaching staff to aid them in generating that road map for the individual as well as the team. 

 

This thread is driving me insane.  Just stop and think about the situation - don't second guess what you think people will do -- you all are just confounding yourself.  Hmeh -- it's the off season. I guess it's what we do -- OK -- go on then  :P

 

 

Well I'll add a little bit more -- as a coach, I'd initially value my powers of observation.  I cannot look at numbers then form my observation - it's too easy to get confused/distracted.  I cannot check my observations against the numbers on a regular basis - I'd get tangled in minutia.  What i need as a coach is for someone to find the stat that tells me WHY I'm seeing what I'm seeing on ice.  If I cannot correct something and I'm at that can't see the forest for the trees place then it's time to look at the basic skills I might be taking for granted. As a GM I'd look at the product - identify weak spots and then check with Mr Stats guy to see if they an help articulate areas of improvement for my coaching staff.  Basically you dont want your stats to destroy your creativity is all.

Edited by Pepperkorn
Link to comment
Share on other sites

I guess I didn't mean it too literally, but yeah, Lou would be the HR manager, and Sunny his assistant/advisor haha

You had it right the first time even if you went of on a bit of a tangent. 

Mehta is indeed part of human resource management. 

It's not a crazy analogy, it's a FACT.   I have no idea why this is even remotely confusing to anyone.

 

If you don't understand that you're not looking at the value of stats correctly.

 

 

(FYI - The HR department fires you not because they made that decisions  - they're doing your boss's dirty work.  Your boss fired you.  No one else did.  Your boss thought you were doing a lousy job.  No one told them you were so they went ahead and agreed. No. They thought you sucked and didn't want to bear the responsibility of telling you so.  Don't kid yourself folks.  Well -- unless facing the truth severely inhibits your ability to move on... I guess then you have to pretend the stats guy got you fired...err I mean laid off because sucking down more revenue than you generate doesn't mean you suck, right?  :unsure: ...  *cough cough*  You know, I used to really talk to people like that, particularly in high school.  Right straight to their face.  I didn't see any other possible way to communicate. I thought to not speak that plainly was to waste peoples time.  hence my soft spot for Bobby Holik I guess :P  But if you really need me to tell you why -- well I'm going to tell you why in the most basic of terms - with a little worst-case thrown in there so you can face the worst and move on.  Funny, now I kind of hate the expression "Get over it"  It sounds so hopelessly out of touch. )

Edited by Pepperkorn
Link to comment
Share on other sites

You are responding to a response not the actual post.  Bad habit -- a way to get yourself all fired up about nothing -- just sayin'

 

Me, if I'm DeBoer -- I dont talk to this guy.  It's not my problem.  I talk to Lou.  Lou talks to this guy to see if there is any correlation between his metrics and the players performance as observed by DeBoer -- then discuss this with the coaching staff to aid them in generating that road map for the individual as well as the team. 

 

This thread is driving me insane.  Just stop and think about the situation - don't second guess what you think people will do -- you all are just confounding yourself.  Hmeh -- it's the off season. I guess it's what we do -- OK -- go on then  :P

 

 

Well I'll add a little bit more -- as a coach, I'd initially value my powers of observation.  I cannot look at numbers then form my observation - it's too easy to get confused/distracted.  I cannot check my observations against the numbers on a regular basis - I'd get tangled in minutia.  What i need as a coach is for someone to find the stat that tells me WHY I'm seeing what I'm seeing on ice.  If I cannot correct something and I'm at that can't see the forest for the trees place then it's time to look at the basic skills I might be taking for granted. As a GM I'd look at the product - identify weak spots and then check with Mr Stats guy to see if they an help articulate areas of improvement for my coaching staff.  Basically you dont want your stats to destroy your creativity is all.

 

Again, a lot of analytics people will tell you DeBoer is doing the right thing with high possession numbers.  (They fault him for starting Marty in too many games last year, but that's not an issue anymore). 

 

I believe that in baseball, the players, coaches and managers do look at the analytics, and will have face to face discussions with the quants.  Baseball's a different animal than hockey I realize, so don't know if it would translate that well into in game strategy, except perhaps for line combinations and divying out ice time. 

Link to comment
Share on other sites

  • 6 months later...

ESPN did a feature on ranking professional sports teams in how receptive they are to analytics. Of course, it's hard to truly know the inner-workings of any team that isn't announcing the information.

 

According to the Devils blurb, in addition to Sunny, the Devils hired Sai Okabayashi. "who created and ran the popular website shiftchart.com"

 

http://espn.go.com/espn/feature/story/_/id/12331388/the-great-analytics-rankings

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.