Confirmed with Link: Duclair/Clendening to Chicago for Panik/Dauphin

[Jakey53]

I mean, it really is that simple. If I'm the GM and I want raw analytics on my team/players and that of every other NHL team/player so I can use that data to make educated decisions using either simple or complex calculations, I go out and hire a team or a firm to put that together. Then I take the data that matters to me and make moves that fit my image of a cup contending team.

The way that GMs read the data may be different from team to team, but most will have similar data based on how good the people compiling the data are.

Some GMs will rely more heavily on the data than the eye test but I can't imagine any team has significantly less access because their GM is less driven by spreadsheets.

You could be right, I don't have a good enough understanding how this data is collected.

 

[Jakey53]

Craig Morgan mentioned recently in a tweet or an article that NHL teams don’t use the same advanced stats available to fans. They mostly have proprietary versions and they tend to be pretty dismissive of the ones available to fans. I believe Morgan stated that WAR and GAR are particularly scoffed at by Chayka and Co.

It’s probably just nerds splitting hairs and it’s basically the same stuff. But I thought it was interesting.

I totally believe this. I read or heard that Chayka's advanced stats differ from Dubas of the Leafs.

 

[Jakey53]

Not just Chayka, other GMs are not WAR or VAR fans either. Every team uses or looks at some form of advanced stats. In hockey, there are just too many variables with other players that affect outcomes to rely on stats to make a decision. Coaches and GMs rely on the eye test, they have trained eyes, unlike most of us, and they watch tape. This is how they make decisions. They then also use advanced stats to confirm what they are thinking. If the advanced stats don't confirm their thinking, they take a harder look at the player and find out they missed something or they wind up ignoring the advanced stats.

I think you are wrong. You are saying it all comes down to the eye test when they analysis a player. I think the eye test holds a certain value in the equation of analysing a player, and the same holds true with different analytics.

 

[MIGs Dog]

The NHL is working on getting puck and player tracking up and running in time for the playoffs. Will bring a whole new element to the data analytics arena.

https://www.espn.com/nhl/story/_/id/27544907/nhl-replaces-partner-player-puck-tracking

 

[Jakey53]

Didn’t Chayka just revamp our amateur scouting department including hiring folks whose sole job is to code plays to create the raw data for their analytics. The key I. The process is the coding. Otherwise it’s garbage in /garbage out.

I agree. That's why I said it's not as simple as some think. It would be interesting to find out how different teams collect this data, is it all in house, or is a third party collecting it, and how they implement it, and the value's they put on the data collected.

 

[MIGs Dog]

The NHL is working on getting puck and player tracking up and running in time for the playoffs. Will bring a whole new element to the data analytics arena.

https://www.espn.com/nhl/story/_/id/27544907/nhl-replaces-partner-player-puck-tracking

Will be cool to know just how fast Grabner is compared to other players.

 

[BUX7PHX]

The fact is, all data can tell you something - it is the pieces that you correctly identify or mis-identify that showcase hits or misses.

If we were to take Kessel's -18 as the only stat that we were to use, we would trade him right away. But, there may be other components that relay the opposite message (effective zone entrances/exits, etc.).

There are some stats that explain a measurable that someone either worries or doesn't worry about. Using Kessel as an example again, let's say that his zone entry percentage is great, but his shooting percentage is floundering. Well, that is less of a worry for me, because he is helping to get the offense into the zone, but his shots are being stopped. Could mean he simply needs to get into better angles for shots. That's an example of two stats working together to shape the story that while Kessel has not been good, that is more attributed to a lack of production in goals only, and not because of some of the other potential reasons to help the team maintain time in the offensive zone.

I also think that Chayka's metrics and analytics are far more based on things like location on ice, positioning, and reaction. If you lose the puck, do you turn on a dime, float in to the turn, turn the wrong direction, etc.? If you show the proper foundational techniques, then we have a little more to work with and we can have an idea of how that player will back-check and what sort of involvement they find themselves in. Obviously, there is also time to understand the proper reads, outlets, etc.

I look at it like a poker hand - sometimes, you make the poor play from preflop through river but still wind up winning. I would be more concerned with finding the player who makes the right decisions through all avenues and just gets unlucky on the river. Those players are going to consistently find ways to be a stronger player, even if the result doesn't become a positive. In other words, don't show a 5 second snippet of awesome goals in the O zone only. Show me the entire 55 second shift and what that player did to get to that goal. Did he position himself properly defensively and maintain effective spacing on offense? Did he find the soft spot of the defense? Did he make the play after doing something poorly, and can that poor decision be fixed effectively? Don't just show all 5 cards and the end of the play - show me the action pre-flop, flop, turn, and river to get there.

 

[XX]

Didn’t Chayka just revamp our amateur scouting department including hiring folks whose sole job is to code plays to create the raw data for their analytics. The key I. The process is the coding. Otherwise it’s garbage in /garbage out.

Players like Duclair, Strome, and Domi have strong analytics. Chayka traded them for other reasons.

Chayka's approach is paying off when it comes to scouting and the draft. The Coyotes are doing much better outside of the first two rounds it seems like, if only because they're now paying attention to league weighted production. Maccelli was the best USHL player in his draft year not in the NTDP for example, which made him a no brainer value where he went.

 

[BUX7PHX]

Players like Duclair, Strome, and Domi have strong analytics. Chayka traded them for other reasons.

Chayka's approach is paying off when it comes to scouting and the draft. The Coyotes are doing much better outside of the first two rounds it seems like, if only because they're now paying attention to league weighted production. Maccelli was the best USHL player in his draft year not in the NTDP for example, which made him a no brainer value where he went.

Depends on the analytic that is being used. What if the measurement that all of those players were strongest at are actually not viewed as integral or premium analytics for Chayka? When Schmaltz was traded to Arizona from Chicago, his CF% Rel was -0.4 and his FF% Rel was -2.0. Strome with Arizona was at 2.3 and 1.1, respectively. This year, Strome is at 1.2 and 0.9 (after being -2.6 and -2.4 post-trade). Schmaltz is at 5.9 and 6.0, respectively

 

[hbk]

Players like Duclair, Strome, and Domi have strong analytics. Chayka traded them for other reasons.

Chayka's approach is paying off when it comes to scouting and the draft. The Coyotes are doing much better outside of the first two rounds it seems like, if only because they're now paying attention to league weighted production. Maccelli was the best USHL player in his draft year not in the NTDP for example, which made him a no brainer value where he went.

It’s tough to call Maccelli a no brainer when he goes in the 4th round. I get your point about offensive production but for a team to take Maccelli they have to believe he has enough offensive upside that he can develop into a player who can either still be a top 6 scorer at the NHL level or he is going to adapt his game (as Garland has) to contribute as a bottom 6 player. If people were sold he’d be a top 6 scoring player then he goes earlier. AZ was fortunate to get him where they did.

 

[MIGs Dog]

It’s tough to call Maccelli a no brainer when he goes in the 4th round. I get your point about offensive production but for a team to take Maccelli they have to believe he has enough offensive upside that he can develop into a player who can either still be a top 6 scorer at the NHL level or he is going to adapt his game (as Garland has) to contribute as a bottom 6 player. If people were sold he’d be a top 6 scoring player then he goes earlier. AZ was fortunate to get him where they did.

Maybe he meant "no brainer" for the 4th rd. I think Garland was a no brainer for the 5th rd. "Hey look, the kid that won the Jean Beliveau trophy twice is still on the board" Last player to do that was Crosby. Definitely worth a 5th rd gamble.

 

[XX]

It’s tough to call Maccelli a no brainer when he goes in the 4th round. I get your point about offensive production but for a team to take Maccelli they have to believe he has enough offensive upside that he can develop into a player who can either still be a top 6 scorer at the NHL level or he is going to adapt his game (as Garland has) to contribute as a bottom 6 player. If people were sold he’d be a top 6 scoring player then he goes earlier. AZ was fortunate to get him where they did.

The old Coyotes would have taken an NCAA project or thrown the pick away on someone that did not have impressive production but had one or two traits scouts liked. Chayka is more production focused. Maccelli in his draft year had a pretty unremarkable skillset but excellent production. By my count, eight non NTDP players were taken ahead of him, all but one with worse production. His scoring at his age in his league became hard to pass up. Scouts circled back and decided he had no major red flags or flaws rather than tacitly endorsing him.

They are using analytics to filter out noise that clouds the judgment of traditional scouts and causes certain players to get overlooked. That's how you go from picking Jedd Soleway types in later rounds to guys like Kirk, Garland, and Bergkvist.

 

[hbk]

Maybe he meant "no brainer" for the 4th rd. I think Garland was a no brainer for the 5th rd. "Hey look, the kid that won the Jean Beliveau trophy twice is still on the board" Last player to do that was Crosby. Definitely worth a 5th rd gamble.

Conner Garland won the CHL top scorer award in 2015 (his first of two). We selected him in the 5th round. The 8th player we chose that draft. Was he a “no-brainer” as well?

 

[XX]

Conner Garland won the CHL top scorer award in 2015 (his first of two). We selected him in the 5th round. The 8th player we chose that draft. Was he a “no-brainer” as well?

At that point, that late in the draft, yes. Raw production begins to outweigh concerns at a certain point. That was the argument. Garland had an incredible amount of work to do. His production is what got him noticed. There were a thousand players just like him but without the numbers his draft year.

 

[hbk]

At that point, that late in the draft, yes. Raw production begins to outweigh concerns at a certain point. That was the argument. Garland had an incredible amount of work to do. His production is what got him noticed. There were a thousand players just like him but without the numbers his draft year.

So to summarize -“If you are going to take a home run swing late in the draft then draft a guy who has displayed offense at a lower level.”

I think that is too simplistic of a theory. Garland has to have the inner drive to change his game. He’s provided spot offense but what has separated him is his drive and competitive nature; his character. I think you are looking for players with offensive and personal drive. Not just those hot shots (this was me as a player btw) who put up great numbers but wouldn’t recognize their goaltender if they ran into them in the street.

 

[XX]

So to summarize -“If you are going to take a home run swing late in the draft then draft a guy who has displayed offense at a lower level.”

Yes. Because statistically, they are hanging with or even beating their peers already. It's far more likely they continue to do so than someone suddenly manifesting scoring ability out of nowhere.

Bergkvist, for example, had more U19 D PPG than Broberg and F.Johansson his draft year in the Allsvenskan. Those guys were first rounders. The guys above him - Cronholm and Norlinder - were taken as 4th and 3rd rounders, although a year apart. If your traditional qualifiers (is he a good person, teammate, coachable, basic skating etc...) come back positive, why not roll the dice on a guy like that? Sure beats picking someone like David Westlund.

There is always projection involved but they are using statistical analysis to pick guys in appropriate slots now. Instead of throwing darts in the dark, they've got the lights on. It's already noticeably better.

 

[hbk]

Yes. Because statistically, they are hanging with or even beating their peers already. It's far more likely they continue to do so than someone suddenly manifesting scoring ability out of nowhere.

Bergkvist, for example, had more U19 D PPG than Broberg and F.Johansson his draft year in the Allsvenskan. Those guys were first rounders. The guys above him - Cronholm and Norlinder - were taken as 4th and 3rd rounders, although a year apart. If your traditional qualifiers (is he a good person, teammate, coachable, basic skating etc...) come back positive, why not roll the dice on a guy like that? Sure beats picking someone like David Westlund.

There is always projection involved but they are using statistical analysis to pick guys in appropriate slots now. Instead of throwing darts in the dark, they've got the lights on. It's already noticeably better.

Traditionally as u get late in the draft scouts are highlighting individual attributes like skating, toughness, etc to lobby for a player to get picked. Doing that but letting their offensive production eliminate potential candidates sounds like a better way to do things. Ie so what if the guy can skate if he can’t produce offense in the Swedish second division.

 

[cobra427]

Yes. Because statistically, they are hanging with or even beating their peers already. It's far more likely they continue to do so than someone suddenly manifesting scoring ability out of nowhere.

Bergkvist, for example, had more U19 D PPG than Broberg and F.Johansson his draft year in the Allsvenskan. Those guys were first rounders. The guys above him - Cronholm and Norlinder - were taken as 4th and 3rd rounders, although a year apart. If your traditional qualifiers (is he a good person, teammate, coachable, basic skating etc...) come back positive, why not roll the dice on a guy like that? Sure beats picking someone like David Westlund.

There is always projection involved but they are using statistical analysis to pick guys in appropriate slots now. Instead of throwing darts in the dark, they've got the lights on. It's already noticeably better.

They are still throwing darts, percentages of success are still very low. We are lucky we got Garland that late based on who he is today not who he was when we drafted him. With smarter scouting, chances of later round success, second or later, are now 6% instead of 4%ish.

 

[BUX7PHX]

The old Coyotes would have taken an NCAA project or thrown the pick away on someone that did not have impressive production but had one or two traits scouts liked. Chayka is more production focused. Maccelli in his draft year had a pretty unremarkable skillset but excellent production. By my count, eight non NTDP players were taken ahead of him, all but one with worse production. His scoring at his age in his league became hard to pass up. Scouts circled back and decided he had no major red flags or flaws rather than tacitly endorsing him.

They are using analytics to filter out noise that clouds the judgment of traditional scouts and causes certain players to get overlooked. That's how you go from picking Jedd Soleway types in later rounds to guys like Kirk, Garland, and Bergkvist.

I believe you are on the right track from certain production standpoints.

However, I think you are being too generic with the production values, say, P per 60 mins played. Remember that Chayka founded Stathletes as a way to help the player understand certain skills the player needed to improve. So I believe that a player like Maccelli, Raty, or whomever these late round gems are getting looked at because the wide-angle lens is showing other skills, such as reaction to the puck, lack of wasted motion, etc. Quickness and acceleration are more important than outright speed type of things.

His entire message has been to find areas that help players, which in turn, developed a science behind scouting decisions. We have his services, but clearly other teams had used Stathletes and courted Chayka too. I just believe that people are selling short just how much time is put into finding the players with the "fixable" problems, as opposed to finding skill players who will struggle to make it, even if those areas are corrected.