Sign in to follow this  
Followers 0
Z-Man

Devils joining the hockey analytics movement

19 posts in this topic

http://insider.espn.go.com/blog/craig-custance/post/_/id/1793/the-future-of-hockeys-analytics-movement

 

Great ESPN article on advance stats & the Devils.  Sorry it's only insider if you subscribe.  Here's a teaser/snipet:

 

 

In January, the New Jersey Devils publicly posted a job opening for a position called Director of Hockey Analytics that was fascinating for a number of reasons. For one, the Devils aren’t often seen as a franchise on the forefront of utilizing advanced stats and hockey analytics. Secondly, if Lou Lamoriello was, in fact, joining the advanced stats movement in hockey, it was stunning that we’d find out about it publicly. Nobody runs a tighter ship than Lamoriello when it comes to containing information, and teams tend to keep their analytics use very private.

The result of the public posting, however, was a barrage of applicants.

“We’ve had... I don’t know how many applicants,” Lamoriello said when we chatted on the topic last week.

 

Share this post


Link to post
Share on other sites

Wonder if Triumph applied for the position?

 

Background in Fenwick, Corsi, Fensi, Forsi, Corwick, Fencor, and Corsiwick a must. 

1 person likes this

Share this post


Link to post
Share on other sites

I like Lomardi's quotes that Lou is about the least "old school" GM that you could find. 

 

 

“Never have I ever considered him old school in his management,” said Kings GM Dean Lombardi, a Lamoriello disciple. “He gave [ilya] Kovalchuk 15 years. Was that old school when he got Scott Stevens in a arbitration hearing? That’s progressive thinking, the way he pulled that off. Go back and read that brief. And look at how aggressive he is in building that team in free agency.”

Lombardi was just getting started.

“Lou has never been an old-school thinker. This is a guy who started his own conference. He starts Hockey East and look what it turned into. This guy ain’t progressive? Give me a break. He hired Rick Pitino out of nowhere,” Lombardi said. “When I saw that posting [for the analytics job], I wasn’t surprised. Actually, I’d go the other way and say, ‘What took so long?’”

 

Still, I wouldn't expect much to change.  Conte has said in the past that they do look at analytics.  I imagine they just want someone that knows it better than Lou/Conte do, and also putting modern tracking equipment into the arena. 

 

We'll see how much they've embraced the movement when it comes time to decide to bring Brodeur back and on what terms.  But it's hard to imagine that analytics didn't influence the decision to trade for Schneider.

Share this post


Link to post
Share on other sites

I like Schneider but I don't know if Analytics would have said that the difference between Schneider and the available alternatives was worth the 9th overall pick.

Share this post


Link to post
Share on other sites

I like Schneider but I don't know if Analytics would have said that the difference between Schneider and the available alternatives was worth the 9th overall pick.

 

I think it was more that the Devils absolutely could not rely on Marty to be the number 1.  The analytics would also tell you that Schneider has been among the best goalies in the league.  So you add that up with the fact that Lou probably knew Kovy was leaving at that point, and didn't know whether he would get Jagr, I think he realized he needed something more than .914 save percentage kind of guy.

 

Otherwise, I haven't seen any analytics evaluation of the trade when it went down. 

Share this post


Link to post
Share on other sites

If they are indeed serious about this, then whoever the director is will advise Lou to ditch Gionta and Salvador (who we have been begging them to drop for the last two years). Still, it's a good sign that NJ has been looking into analytics. You see STL and SJ who have their scouts look at Corsi and Fenwick for CHL players and how well they have drafted recently. I dont think this is due to luck

Share this post


Link to post
Share on other sites

Wonder if Triumph applied for the position?

 

List of applicants for the hockey analytics position:

 

Applicant 1: Triumph Smith

 

Applicant 2: Triumph Jones

 

Applicant 3: Tryumph Jackson

 

Applicant 4 Tri Umphchuk

 

Applicant 5: Bobby Triumph

 

Applicant 6: Traiumph Fennywick

 

Applicant 7: Triumphant Corsi4ever

 

Applicant 8: Bobby Holik

 

Applicant 9: ............

 

maybe this analytics guy can save josefson.

 

Lol...there are many different ways to view a single assist,

Edited by Neb00rs
1 person likes this

Share this post


Link to post
Share on other sites

Wouldn't analytics show that the team wins more and scores more with Marty on the ice rather than Cory?  Wouldn't they say GAA, save %, etc don't mean a damn thing if the team doesn't win?  Not trying to start another raging debate here but isn't that their purpose?  I can see 90% of the posters here whining about Cory not getting starts and spouting off his amazing stats and Lou saying "Marty starts why?" and pointing to me and I say "because the team scores more and wins more when he's on the ice."

Share this post


Link to post
Share on other sites

Wouldn't analytics show that the team wins more and scores more with Marty on the ice rather than Cory?  Wouldn't they say GAA, save %, etc don't mean a damn thing if the team doesn't win?  Not trying to start another raging debate here but isn't that their purpose?  I can see 90% of the posters here whining about Cory not getting starts and spouting off his amazing stats and Lou saying "Marty starts why?" and pointing to me and I say "because the team scores more and wins more when he's on the ice."

No.

 

The Devils don't get significantly more shots when Marty starts, which kills the theory that he creates offense. The Devils scoring more for Marty is due to blind luck, rather than Brodeur starting rushes.  

 

Cory stops the puck at a .915 rate (and it would be around .925 if it weren't for 4 bad games), which is pretty good.  Marty is literally one of the worst at stopping the puck nowadays. 

Share this post


Link to post
Share on other sites

more from the article

 

It’s not cheap to start an analytics department, and it also takes support from ownership. When the Devils were sold in August, it was reported that they were losing around $25 million per season. That’s probably not the time to ask about adding new positions to hockey operations.

But that changed when Philadelphia 76ers owner Josh Harris bought the Devils. They pledged to build a highly successful management team and provide them with the resources necessary to win. That’s already starting.

At the recent Sloan Analytics Conference, you couldn’t help but feel like the NHL was light years behind the NBA in gathering and using data to improve teams. Part of it is because the NBA has motion-tracking cameras in every arena, generating data on every single player during every single game. NBA owners have seen the benefit of having that data, and it’s safe to assume Harris and his group would like the Devils to be just as progressive in their data gathering. The challenge now is finding a way to make it happen.

“Our new owners are very much into that. They spoke to me about -- how could that help? And math is my background,” Lamoriello said. “If they felt that good and they were willing to support me putting together some people to look into some areas and the stats are getting better. Why not?”

Cost is one barrier. According to a team that examining the idea of having motion-tracking cameras gather data for their franchise, it would cost into the six figures to have the entire season tracked and the data produced. And that’s before knowing if the data is applicable.

Part of the issue is that if one team signs on and gathers data, there wouldn’t be comparative data from other teams or historical data to examine exactly what the data means. For example, the tracking system can help determine a player’s top speed and how often he’s skating within the top 20 percent of his top speed while on the ice. That’s fascinating information, but without a pool of comparison data over a long stretch of time, it’s not easy to interpret.

But even with your own data, you could discover that one of your defenseman’s average speed is five percent lower than the average defenseman on your team. You could determine pass completion percentages, number of possessions in the offensive zone, offensive zone possession time compared to the average player and the total number of touches just to start. In a story that launched FiveThirtyEight.com, Nate Silver suggested that this data could measure the perseverance of an NHL player, removing the bias of coaches and scouts.

“It’s got tons of potential,” concluded one exec.

Share this post


Link to post
Share on other sites

If they are indeed serious about this, then whoever the director is will advise Lou to ditch Gionta and Salvador (who we have been begging them to drop for the last two years). Still, it's a good sign that NJ has been looking into analytics. You see STL and SJ who have their scouts look at Corsi and Fenwick for CHL players and how well they have drafted recently. I dont think this is due to luck

 

 

I think it's largely due to luck.  If you look at STL's drafting, they have picked relatively few players from the CHL over the past 5 or 6 years, and only Pietreangelo has turned into a mainstay, and it wasn't like he was some kind of darkhorse pick.  It's a similar story with the Sharks.  You could, of course argue that they passed up on certain players because of analytics. 

Share this post


Link to post
Share on other sites

I think it's largely due to luck. If you look at STL's drafting, they have picked relatively few players from the CHL over the past 5 or 6 years, and only Pietreangelo has turned into a mainstay, and it wasn't like he was some kind of darkhorse pick. It's a similar story with the Sharks. You could, of course argue that they passed up on certain players because of analytics.

If they are tracking CHL stats, I don't think it's unreasonable to conclude they are doing it in other junior leagues or in Europe.

Share this post


Link to post
Share on other sites

No, I didn't apply for this - I don't have a data science degree or anything like that which you'd need to really look at this stuff in depth.  I still don't think this is Lamoriello's doing at all and that ownership convinced him into doing it, and that a position like this will be used to evaluate potential moves only on a negative basis.  Tyler Dellow has made the excellent point that some of the resistance among general managers to use this kind of information is that unless you look at this stuff a lot, it's hard to separate the snake oil from the legitimate insights (and it doesn't help that there are a lot of people peddling statistical snake oil).

 

BTW, Gionta is quietly having an okay season territorially.

Edited by Triumph

Share this post


Link to post
Share on other sites

No, I didn't apply for this - I don't have a data science degree or anything like that which you'd need to really look at this stuff in depth.  I still don't think this is Lamoriello's doing at all and that ownership convinced him into doing it, and that a position like this will be used to evaluate potential moves only on a negative basis.  Tyler Dellow has made the excellent point that some of the resistance among general managers to use this kind of information is that unless you look at this stuff a lot, it's hard to separate the snake oil from the legitimate insights (and it doesn't help that there are a lot of people peddling statistical snake oil).

 

BTW, Gionta is quietly having an okay season territorially.

Huh? He has the worst rel. Corsi on the team of any forward not named Tedenby and has faced pretty easy competition. Granted he starts a lot of shifts in the D zone, but the point stands

Share this post


Link to post
Share on other sites

Huh? He has the worst rel. Corsi on the team of any forward not named Tedenby and has faced pretty easy competition. Granted he starts a lot of shifts in the D zone, but the point stands

 

Corsi Rel is largely garbage.  The critical point is that Gionta's starting in the O-zone less than 40% (meanwhile the team's O-zone percentage is something like 53%) and is breaking even on shots.  Given that there's a mixture of 4th lines and top lines in that mix, I'll take it.  I don't know for much longer it can continue and how much Gionta is responsible, and it's impossible to tell from With Or WIthout You studies (because Gionta starts so much in the defensive zone as compared to his peers, his WOWYs will look bad no matter what), but it's acceptable, and it's better performance than NJ has gotten out of some recent 4th lines.

Share this post


Link to post
Share on other sites

Corsi Rel is largely garbage.  The critical point is that Gionta's starting in the O-zone less than 40% (meanwhile the team's O-zone percentage is something like 53%) and is breaking even on shots.  Given that there's a mixture of 4th lines and top lines in that mix, I'll take it.  I don't know for much longer it can continue and how much Gionta is responsible, and it's impossible to tell from With Or WIthout You studies (because Gionta starts so much in the defensive zone as compared to his peers, his WOWYs will look bad no matter what), but it's acceptable, and it's better performance than NJ has gotten out of some recent 4th lines.

Ok, but as long as he's on the team, he will be over-utilized or misused (I'm not sure what is worse to be honest). I think most people accept that he is an NHLer, though some would say he's a fringe player, but his presence on the team is negative. It really isn't his fault, but it is what it is

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!


Register a new account

Sign in

Already have an account? Sign in here.


Sign In Now
Sign in to follow this  
Followers 0