Arizona GM Chayka on Long Analytics Road

[Tulipunaruusu*]

I would like to learn since what I have read from public analysts of the 'Corsi' (anything to do with shot related possession) generation has been a waste of both our times. Personally I favour the Midtjylland/Brentford model of putting teams and players in a continent-wide league table so I am not against analytics, just how it is done. That Midtjylland branded one good is aid if you want to inspect how players could position in the big picture. It would just help if there at least some idea behind otherwise it is more of a marketing tool to generate hype.

I do not think that someone being hired is a very strong testimony of their usefulness since humankind has through the early days of recorded civilization employed all kinds of people who could perform similar tricks as these analysts pull nowadays on the ice hockey people. Some of the least competent ice hockey workers in Finland have for example been hired by NHL teams or the Finnish national team which should stand as the pinnacle of the sport. I am not talking of Jarmo Kekäläinen. Yet, these are the kind who make you waste your second round pick or underperform in the World Cup.

Just because something happens behind closed doors in some oddly worshipped institute would not make it necessary more refined in terms of resources used than what I for example did in highlighting Hartnell-Wennberg-Dano connection, Puljujärvi evasion in the draft or Nutivaara's rise. Those cases were founded on judging the core tenets of play and using the Eurasian ice hockey pyramid of the Midtjylland-Brentford form. I doubt if any of the 'Corsi' geezers can display a similarly succesful portfolio for the after lockout period.

 

[Doctor No]

If you're basing your negative impressions on what NHL teams are doing here on what you know of Corsi, I've got good news for you.

 

[Tulipunaruusu*]

Speak freely.

If you got anything else than Scotty Bowman junior & analysts taking credit for the Blackhawks success I would be more than open to take the opportunity to learn something new. For example those already mentioned heatmaps to highlight player movement and pass chains to visualize the whole ordeal would be really a really welcome tool, even if I would be the last one in the hockey world to know of these. Or anything in this direction, ready to be utilized, would in my mind be helpful.

If this is about some eyes wide shut parties... I do not think it is a negative but more reserved approach.

 

[fiveonfive]

I would like to learn since what I have read from public analysts of the 'Corsi' (anything to do with shot related possession) generation has been a waste of both our times. Personally I favour the Midtjylland/Brentford model of putting teams and players in a continent-wide league table so I am not against analytics, just how it is done. That Midtjylland branded one good is aid if you want to inspect how players could position in the big picture. It would just help if there at least some idea behind otherwise it is more of a marketing tool to generate hype.

I do not think that someone being hired is a very strong testimony of their usefulness since humankind has through the early days of recorded civilization employed all kinds of people who could perform similar tricks as these analysts pull nowadays on the ice hockey people. Some of the least competent ice hockey workers in Finland have for example been hired by NHL teams or the Finnish national team which should stand as the pinnacle of the sport. I am not talking of Jarmo Kekäläinen. Yet, these are the kind who make you waste your second round pick or underperform in the World Cup.

Just because something happens behind closed doors in some oddly worshipped institute would not make it necessary more refined in terms of resources used than what I for example did in highlighting Hartnell-Wennberg-Dano connection, Puljujärvi evasion in the draft or Nutivaara's rise. Those cases were founded on judging the core tenets of play and using the Eurasian ice hockey pyramid of the Midtjylland-Brentford form. I doubt if any of the 'Corsi' geezers can display a similarly succesful portfolio for the after lockout period.

You can say that about every single hockey executive ever, in fact it probably applies better to a regular eye test hockey executive than it does to a data-driven analyst. As far as the rest of your comment, I have a hard time understanding what your problem is with "Corsi" geezers. If anything, I would say the problem would be traditional hockey people who would not permit deep stats driven involvement in their teams. There can't be a successful portfolio of corsi-driven work if those who are in charge don't even entertain the idea.

But as people have pointed out already, we know for a fact teams have stats departments and we also know that we have no idea how integrated their work is with decision making or day to day hockey operations. It is likely there is proprietary work that has undeniable results that teams have kept secret just to maintain an advantage.

 

[bigd]

The team hunted down bad-contracts of other teams to get additional assets in return for cap room and kept Mike Smith in goal. They are bad by design. Also, having a GM that is analytically savy doesn't magically transform the team's players into good players. It will take much more time and work than that.

So they sux by design, that's a first. Personally I just think the Arizona fans are just grasping at straws. I wonder how long, "suxing by design" will go over before they turn on him? Two years, three, or four, how long will they be patient?

 

[mouser]

So they sux by design, that's a first. Personally I just think the Arizona fans are just grasping at straws. I wonder how long, "suxing by design" will go over before they turn on him? Two years, three, or four, how long will they be patient?

Not getting where all your vitriol is coming from? So far Chayka has overseen several decisions that project well for the team in the future. Such as drafting Keller, the draft move for Chychrun, and the trade for DeAngelo who's now looking impressive in his NHL time. At this point the team fans are far and away more dissatisfied with the coaching staff then the GM over the team's performance/ice time of younger players, and rightfully so.

 

[fiveonfive]

So they sux by design, that's a first. Personally I just think the Arizona fans are just grasping at straws. I wonder how long, "suxing by design" will go over before they turn on him? Two years, three, or four, how long will they be patient?

Suxing by design is only a first if you exclude all the teams who have done it before, like Buffalo or Toronto. Realizing that your team has a poor chance of succeeding in it's current form and focusing on medium and long term outcomes isn't a controversial strategy. If you think they traded for those bad contracts because they genuinely though it would help their play on the ice, you are gonna have a tough time understanding what is going on. Teams want to contend for the cup, improving from bottom 3 to bottom 10 wouldn't do them any good. It's more logical to take a step back now, so that you can set up a plan for multiple steps forward in the future.

Mouser's post above goes into more detain in terms of actual decisions or actions taken. I don't think you can look at those and think "Analytics GM makes moves thinking his team will win the cup this year". It's definitely more of a "Gm makes moves to build a team, understanding that it's a process and not an overnight outcome".

 

[bigd]

Suxing by design is only a first if you exclude all the teams who have done it before, like Buffalo or Toronto. Realizing that your team has a poor chance of succeeding in it's current form and focusing on medium and long term outcomes isn't a controversial strategy. If you think they traded for those bad contracts because they genuinely though it would help their play on the ice, you are gonna have a tough time understanding what is going on. Teams want to contend for the cup, improving from bottom 3 to bottom 10 wouldn't do them any good. It's more logical to take a step back now, so that you can set up a plan for multiple steps forward in the future.

Mouser's post above goes into more detain in terms of actual decisions or actions taken. I don't think you can look at those and think "Analytics GM makes moves thinking his team will win the cup this year". It's definitely more of a "Gm makes moves to build a team, understanding that it's a process and not an overnight outcome".

I didn't expect them to be a Stanley cup contender but I did expect improvement. As for taking over bad contracts to improve your team later, that's not a new concept. The Islanders did that a few years ago and they still suck. We shall see what the future holds. I would expect to see some improvement by next year at least.

 

[fiveonfive]

I didn't expect them to be a Stanley cup contender but I did expect improvement. As for taking over bad contracts to improve your team later, that's not a new concept. The Islanders did that a few years ago and they still suck. We shall see what the future holds. I would expect to see some improvement by next year at least.

Who said the concept was new? The point is that it doesn’t exactly send a message of “we are trying to be better, right this momentâ€, so reading it that way is hardly sensible.

If by next year they are starting to sign legitimate NHLs instead of trading for cap hits of retired players, it will be clear they are trying to improve. Then we will have grounds to comment on how good their reconstruction decisions are panning out. For now, new management group wasn’t anticipating to win, so to say that the new analytics GM is doing a poor jobs makes little sense.

 

[Doctor No]

I didn't expect them to be a Stanley cup contender but I did expect improvement. As for taking over bad contracts to improve your team later, that's not a new concept. The Islanders did that a few years ago and they still suck. We shall see what the future holds. I would expect to see some improvement by next year at least.

Improvement isn't always borne out as linear on the ice.

And as for "not a new concept", the point is to eventually win a Stanley Cup, not to be coy.

 

[Tulipunaruusu*]

And as for "not a new concept", the point is to eventually win a Stanley Cup, not to be coy.

First priority is always to have a job.

So did these geezers of Chayka employment use the next generation of 'statistics' or what? You cannot be serious that despite the NHL being a fully professional Arena league there are not for example player movement related heatmaps available for the public eye? Sportingcharts for example has very primitive tools.

Would think that the advantage would come from those rather than investing more powder through shot focusment.

 

[Doctor No]

I can't tell if your response is intended to be a non-sequitur or not.

With that said, it's not really related to what the Coyotes are doing with respect to analytics, nor is it related to their success (or lack thereof) on the ice as a result.

Care to clarify?

 

[Tulipunaruusu*]

Care to clarify?

What I suspect is that the whole approach is flawed be it then because baseball or American football, with heavy stats tracking, intermingled with ice hockey hence the misunderstanding, or whatever. Therefore it is apparently, from what I gather, impossible to produce meaningful stats tracking as the whole approach leaves you wanting for proper, individual movement highlighting heatmaps and visualized pass chain tracking.

Why the Coyotes fail is that this is just another attempt at gamble instead of well reasoned and researched work to support the team behind it all. I would try hiring someone from the gambling business as the next general magi to lead in team building if this is the selected method.

Maybe after Chayka there will be another, more powerful... He will have to come through revolution, not evolution.

 

[Doctor No]

So you're back at the start, with the "I don't believe that this stuff could possibly work."

Which is fine, but (1) I thought we moved past that - you're clearly not going to be convinced, and I don't think there's anyone left who's passionate enough to try - and (2) it's not related to the new stuff.

"Maybe after Chayka there will be another, more powerful". Or maybe you'll start judging a GM - *any* GM - after more than five weeks of on-ice play.

 

[Tulipunaruusu*]

Final plea: does anyone know of use of visualized heatmaps or pass-chains in ice hockey?

 

[XX]

Final plea: does anyone know of use of visualized heatmaps or pass-chains in ice hockey?

Chayka uses heatmaps. It was shown in a behind the scenes video.

 

[Tulipunaruusu*]

Chayka uses heatmaps.

But again they were shot-based heatmaps which tells nothing?

 

[fiveonfive]

Final plea: does anyone know of use of visualized heatmaps or pass-chains in ice hockey?

But again they were shot-based heatmaps which tells nothing?

Why would heat maps or pass-chains mean anything vs heat maps of shots meaning nothing? Do you have anything of substance to comment on those things other than "his methods wont work, but I know the real way" claims? I'd love to learn something, but so far you haven't really said anything worthwhile.

 

[Tulipunaruusu*]

Because one highlights the play, other focuses on outcome.

The Leading One's blog is a starting point if we are to exchange resources:

https://blog.paf.com/petterisihvonen/

 

[fiveonfive]

Unfortunately I do not speak any Finnish (if that is indeed the language of that link).

And yes, one is an outcome measure and the other is process, but I still don't understand how you can proclaim that one definitely doesn't work and the other is the golden ticket to success in hockey, and the NHL specifically. You haven't provided anything tangible to distinguish that.

 

[Tulipunaruusu*]

Unfortunately I do not speak any Finnish (if that is indeed the language of that link).

Try Google Translate.

 

[Freudian]

It's not a sprint. In five years we'll see how he performed.

I do think in some circles GMs who are known to use analytics are considered smart by default and GMs who are more traditional are considered dumb by default and that's obviously nonsense. Same goes for dismissing analytics and thinking the eye-test and things like compete/heart is all that works.

Chayka took over a flawed team and they are still flawed. The assets he's acquired and the players he drafted won't make a difference right now.

 

[fiveonfive]

Try Google Translate.

I have now. So far no heat maps or any data about the NHL. But I will give it a deeper dive when I have a chance.

 

[Tulipunaruusu*]

I have now. So far no heat maps or any data about the NHL. But I will give it a deeper dive when I have a chance.

The Leading One will only open your understanding of ice hockey. If we had these visualized heat-maps and pass-chains, ready to be used, surely we would not give them to the Enemy so easily.

 

[bert]

Not really working out for him so far.

Was it working for the leafs last year?