I knew that Stranges clip before I even played it. I watched that exact moment last night and said to myself "look at him go" in the same way I would if I was watching my half wit dog set loose in a dog park for the first time to unleash his ultimate spaz out. Stranges is entertainment
I read the same thread you did, and I'll go further than you regarding different catch-all numbers. I think Dom's GSVA is a tier below (lol) Micah's sG and Evolving-Hockey's xGAR as far as putting a number to a player's *talent*. In fact, if I were giving them tiers, I'd put sG in 1A, xGAR in 1B, GAR in 2A and GSVA in 2B.
The reason is really simple, GSVA is very heavily points-dependent. If I understand it correctly, each player has a game score for each game* and Dom adds up the game score for each player on each team for the full season, then feeds it into a regression for goal differential to get how much each player's game score contributed to the whole, then he scales it for the number of goals scored for the season. Heiskanen's point totals dropped last year, which hurts him in GSVA especially, but I'd point to the Stars power play trotting out Benn and Pavelski all year long as the reason for that.**
There's a little cognitive dissonance here, because Dom's team projections are consistently the best in the public sphere and the only model that consistently beats betting odds, but I think GSVA is picking up some sort of team effect and not necessarily assigning value proportionally to talent.
** sidenote: I feel like Robertson and Hintz were similarly undervalued in the exercise because their point totals dropped last year, but that was mostly a consequence of Pavelski aging.
Great stuff Aaron. I feel team effects in general are understudied. I understand a lot of works goes into isolating player impact, but the level of isolation needed to truly measure singular impact seems...lacking. Show me a good team, and I'll show you a good GAR, GSVA, or WAR chart. Micah seems to be the only one working hard to constantly refine his model.
I agree! And as you like to say, matchups make fights, and I think that’s the next big piece for understanding the game in the public sphere.
I know you’re aware of his work, but for anyone who isn’t, I think Louis Boulet’s stuff integrating the A3Z data with play-by-play is really exciting and a big step towards isolating individual impact.
This write up on Miro is so great. What he does is subtle if you don’t watch every game. It’s like he solves problems before they happen and doesn’t get the credit because the problem never happened. When he’s injured you see the impact right away. Maybe a bit less so now that Harley is up to speed. Keep up this great work.
Miro is such a quality defender, but his lack of flash and playing in hockey flyover country will keep him off the radar until it is impossible to deny it. Swap teams with him and McAvoy and there would be no debate about Miro (Plus only true hockey fans would know who that Charlie guy was).
Forget individual hardware, win the cup Miro. By doing that you'd end up positioning yourself for the Norris.
Is Stranges coachable, and if so, can directed coaching teach him when and when not to go into Chaos mode? Experience tells me probably not but man, if you could teach him to use those skills to drag attention to himself and then dish the puck off to wide open linemates....
Depends on what is meant by coachable. I think he listens to his coaches, and takes his job seriously. He is, in addition to being talented, also a hard worker, and while no Patrice Bergeron -- defensively responsible. So a coach can teach him the Xs and Os of hockey, and I think he's capable of learning. But what Stranges needs is a mentor, and a skills coach: somebody who doesn't care whether he helps a team win, but only about how Stranges can play comfortably within himself, and develop into Super Stranges.
He falls into the “dipsy-doodles” category… those guys usually embarrass the wrong guy, and end up getting smoked. He needs to integrate his moves into passes with his teammates… who rarely get the puck back once they give it to him. Trying an old Give and Go, would be the simplest idea. Guessing it’s been mentioned to him…
Let me be crystal clear here wouldn’t swap Miro for Makar ever . Miro has played his whole career on his off wing , secondly he is not handed partners like Toews for his full career who cover his poor defensive play . Lastly if we had Makar instead of Miro we wouldn’t have been to the last two Western finals . Let’s get Makar to play with Suter as his partner and see what happens .
So my thoughts on that also have changed… where I thought we’d made a mistake after watching Macar get the cup… now I think what I continually argue with myself about is Miro’s lack of physicality, for his size… one side of the coin -a fault letting guys glide by him on the Rare occasions his stick work fails… or the other side of the coin which I’m trying to appreciate More, that exact quality will also make him less likely to get injured and allow him to stay on the ice… where we need him.
Who to pick 1st will be argued all the was past their Hall of Fame introductions, :)))).
I don't know. I go back and forth here. While I consider three-zone defensive presence more essential to the job than play finishing, being able to do so at a high level like Makar still counts for a lot. Not to be argumentative, but I'm pretty sure Makar would trade two WC Finals appearances for the Cup he earned. That kind of ability to simply jailbreak a play still counts for a lot, consistency be damned. I guess my real point is that while I loosely agree, the unfortunate reality is that Heiskanen's ability on his strongside is completely theoretical: he's spent more on as a RHD than as a LHD at this point in his career. Until he switches (and hopefully this is the year), that's the only way to evaluate him. With Lindell getting his big extension, there's no end in sight either. (I know Heiskanen-Dumba is grabbing all the headlines, but I doubt it lasts. How do you put Heiskanen on his natural side while keeping Lindell and Harley in the top four?)
Strong side shift in transition leading to extended shifts from #1-#3 in favorable situations, with #4 - #6 2nd over the boards in transition & occasionally loosing shifts, but subject to favorable zone starts based off playing style.(Dumba-Dual, Lyubuskin-D, Lundkvist- O)
Well consider if Lundquist / Bischel or another is a quality 6 th D man and PDB actually roles 6 D like he mostly does the forwards which I think he would do . At that point Lindell would be fine with say Lundquist plus playing PK minutes . On Makar he has great offensive skills but put him on his off wing with a Polak or Suter and see how he would play instead of Toews
I knew that Stranges clip before I even played it. I watched that exact moment last night and said to myself "look at him go" in the same way I would if I was watching my half wit dog set loose in a dog park for the first time to unleash his ultimate spaz out. Stranges is entertainment
Reminds me of Harlem Globetrotters type hockey player… great at showing off specific skills, but not that great at the actual purpose of the game.
Exactly lol. Super fun to watch. Maybe we can petition for a monthly highlight reel of him in the AHL being a cartoon Tasmanian Devil
I read the same thread you did, and I'll go further than you regarding different catch-all numbers. I think Dom's GSVA is a tier below (lol) Micah's sG and Evolving-Hockey's xGAR as far as putting a number to a player's *talent*. In fact, if I were giving them tiers, I'd put sG in 1A, xGAR in 1B, GAR in 2A and GSVA in 2B.
The reason is really simple, GSVA is very heavily points-dependent. If I understand it correctly, each player has a game score for each game* and Dom adds up the game score for each player on each team for the full season, then feeds it into a regression for goal differential to get how much each player's game score contributed to the whole, then he scales it for the number of goals scored for the season. Heiskanen's point totals dropped last year, which hurts him in GSVA especially, but I'd point to the Stars power play trotting out Benn and Pavelski all year long as the reason for that.**
There's a little cognitive dissonance here, because Dom's team projections are consistently the best in the public sphere and the only model that consistently beats betting odds, but I think GSVA is picking up some sort of team effect and not necessarily assigning value proportionally to talent.
* https://hockey-graphs.com/2016/07/13/measuring-single-game-productivity-an-introduction-to-game-score/ this is an old version, and he's updated it with xG and QoT/QoC adjustments, but points is still the biggest driver.
** sidenote: I feel like Robertson and Hintz were similarly undervalued in the exercise because their point totals dropped last year, but that was mostly a consequence of Pavelski aging.
Great stuff Aaron. I feel team effects in general are understudied. I understand a lot of works goes into isolating player impact, but the level of isolation needed to truly measure singular impact seems...lacking. Show me a good team, and I'll show you a good GAR, GSVA, or WAR chart. Micah seems to be the only one working hard to constantly refine his model.
I agree! And as you like to say, matchups make fights, and I think that’s the next big piece for understanding the game in the public sphere.
I know you’re aware of his work, but for anyone who isn’t, I think Louis Boulet’s stuff integrating the A3Z data with play-by-play is really exciting and a big step towards isolating individual impact.
This write up on Miro is so great. What he does is subtle if you don’t watch every game. It’s like he solves problems before they happen and doesn’t get the credit because the problem never happened. When he’s injured you see the impact right away. Maybe a bit less so now that Harley is up to speed. Keep up this great work.
Miro is such a quality defender, but his lack of flash and playing in hockey flyover country will keep him off the radar until it is impossible to deny it. Swap teams with him and McAvoy and there would be no debate about Miro (Plus only true hockey fans would know who that Charlie guy was).
Forget individual hardware, win the cup Miro. By doing that you'd end up positioning yourself for the Norris.
Thanks Jeff!
the EV RAPM comparo of Makar to Heiskanen - holy crap what an eye opener
Yep. I'd expect Makar to bounce back (he really started to towards the end of the season) though.
An Henri Bergson reference in a hockey article. You, sir, are the people's champion.
Haha. Thanks Jon. I'll shoehorn them in when I can.
Is Stranges coachable, and if so, can directed coaching teach him when and when not to go into Chaos mode? Experience tells me probably not but man, if you could teach him to use those skills to drag attention to himself and then dish the puck off to wide open linemates....
Depends on what is meant by coachable. I think he listens to his coaches, and takes his job seriously. He is, in addition to being talented, also a hard worker, and while no Patrice Bergeron -- defensively responsible. So a coach can teach him the Xs and Os of hockey, and I think he's capable of learning. But what Stranges needs is a mentor, and a skills coach: somebody who doesn't care whether he helps a team win, but only about how Stranges can play comfortably within himself, and develop into Super Stranges.
He falls into the “dipsy-doodles” category… those guys usually embarrass the wrong guy, and end up getting smoked. He needs to integrate his moves into passes with his teammates… who rarely get the puck back once they give it to him. Trying an old Give and Go, would be the simplest idea. Guessing it’s been mentioned to him…
Let me be crystal clear here wouldn’t swap Miro for Makar ever . Miro has played his whole career on his off wing , secondly he is not handed partners like Toews for his full career who cover his poor defensive play . Lastly if we had Makar instead of Miro we wouldn’t have been to the last two Western finals . Let’s get Makar to play with Suter as his partner and see what happens .
So my thoughts on that also have changed… where I thought we’d made a mistake after watching Macar get the cup… now I think what I continually argue with myself about is Miro’s lack of physicality, for his size… one side of the coin -a fault letting guys glide by him on the Rare occasions his stick work fails… or the other side of the coin which I’m trying to appreciate More, that exact quality will also make him less likely to get injured and allow him to stay on the ice… where we need him.
Who to pick 1st will be argued all the was past their Hall of Fame introductions, :)))).
All the “Way”
I don't know. I go back and forth here. While I consider three-zone defensive presence more essential to the job than play finishing, being able to do so at a high level like Makar still counts for a lot. Not to be argumentative, but I'm pretty sure Makar would trade two WC Finals appearances for the Cup he earned. That kind of ability to simply jailbreak a play still counts for a lot, consistency be damned. I guess my real point is that while I loosely agree, the unfortunate reality is that Heiskanen's ability on his strongside is completely theoretical: he's spent more on as a RHD than as a LHD at this point in his career. Until he switches (and hopefully this is the year), that's the only way to evaluate him. With Lindell getting his big extension, there's no end in sight either. (I know Heiskanen-Dumba is grabbing all the headlines, but I doubt it lasts. How do you put Heiskanen on his natural side while keeping Lindell and Harley in the top four?)
#1 - #4
#2 - #5
#3 - #6
Strong side shift in transition leading to extended shifts from #1-#3 in favorable situations, with #4 - #6 2nd over the boards in transition & occasionally loosing shifts, but subject to favorable zone starts based off playing style.(Dumba-Dual, Lyubuskin-D, Lundkvist- O)
Well consider if Lundquist / Bischel or another is a quality 6 th D man and PDB actually roles 6 D like he mostly does the forwards which I think he would do . At that point Lindell would be fine with say Lundquist plus playing PK minutes . On Makar he has great offensive skills but put him on his off wing with a Polak or Suter and see how he would play instead of Toews