deadhead
Registered User
- Feb 26, 2014
- 49,215
- 21,617
EvolvingWild's RAPM: Reviving Regularized Adjusted Plus-Minus for Hockey
Check the D-zone and O-zone maps, seems these are impactful.
EvolvingWild's RAPM: Reviving Regularized Adjusted Plus-Minus for Hockey
That's a very passionate post.A-f*cking-men! If the Corsi-loving crowd doesn't think a player who runs players has no affect on a hockey game than they are nuts. The last playoff series the Flyers were in; the Caps number one objective was to eliminate #14. They ran him till Ovechkin broke him. When the Flyers beat the Pens way back when; it was due to the Flyers starting crap after every whistle & making Malkin get stupid. That stuff is not measured in CORSI etc. Some players fade when the game gets rough. Look at JVR; come playoff time lets see how much time he is in front of the net versus a game in Jan. His glides will start closer and closer to the red line. The game has been toned down. But playoff games are still physical to an extent and you better know who is on the ice at all times. Because guys like Wilson or Brooks Orpik in the past will take the suspension if it means knocking #28 or #14 out of the series. That element is still in the game. Controlled entries or not.....
A POV that drives me nuts and makes no sense: "Actually watch the games you nerd!"
I dont know anyone who only follows their favorite hockey team by perusing NaturalStatTrick after the game to see what the 5v5 metrics are (& doesn't actually watch the games). Can anyone explain that POV? It operates under an inane assumption that's profoundly odd.
According to twitter he's been promoted to the 2nd line since Farabee was sent down
Going from 4C in Anaheim to 2LW in Philly seems like an odd move, though.
Maybe he's going to play 3C?
Laughton-Grant-Pitlick
Raffl-Thompson-NAK
Homie, there's legitimate r sq values associated with these models that essentially quantify robustness.I check NST after games to check on what I thought I saw, because I'm watching a game, not taking notes on every shift - NST is a quick and dirty summary.
But it's not a replacement for watching games.
Given that the various models give different results, robustness is a serious issue with hockey models.
Which isn't to say they're useless, rather, the results of any one model should be taken with a grain of salt.
And the models can influence outcomes, I think some teams misunderstood Corsi and engaged in strategies to raise their shot totals, which then made Corsi less valuable.
I imagine he’ll end up a 4th line wing, but they are seeing how he looks at 3C because they like Laughton better at wing if possible.According to Charlie O, Curu is right and Grant is slotted into the 3C spot for tonight with Laughton taking the spot on Hayes' LW (which is another conversation we can have).
Look, I like Grant. I've seen him a fair bit in person and on screens.
A lot of Ducks fans out here like Grant.
He's a good 4th line option and Alex's numbers help explain why.
But this "Grant as a potential 2nd line winger" and "Grant versus Farabee" and "Grant as a 3C option" is just setting him up for failure, in my view.
Grant is popular out here because he's League minimum underdog who has fought hard and done reasonably well on the Ducks 4th line.
Grant isn't a 2nd line player, or even a 3rd line center, on the League's 4th worst team. Why would he suddenly be a 2nd, or 3rd, line guy on the 8th best team?
Homie, there's legitimate r sq values associated with these models that essentially quantify robustness.
Nothing is perfect, but models with high r sq values intrinsically aren't to be "taken with a grain of salt" (which by the way is another way to say useless) unless you don't believe in math. But, I'll be the first to admit they shouldn't be taken as gospel either. In addition to a reliable eye for the game, these models can seriously paint a great picture of what is happening and what could happen in the future. It's just more information to use to make a conclusion.
Do you look at weather radar? That's based on prediction models with some magnitude of robustness that doesn't equal 100%. Sometimes that's wrong too.
Grant isn't a 2nd line player, or even a 3rd line center, on the League's 4th worst team. Why would he suddenly be a 2nd, or 3rd, line guy on the 8th best team?
I imagine he’ll end up a 4th line wing, but they are seeing how he looks at 3C because they like Laughton better at wing if possible.
Grant has a higher pp/60 the last three seasons than either Pageau or Goodrow.
So are they 2C/3C types?
I don't think Grant is a great 3C, but most NHL 3Cs score between 20-30 ES points playing 12-13 minute at night at ES. So it's not like he'd be some black hole there offensively.
Go look at my other posts if you want the numbers. I've typed them out plenty of times
Lazy is "have you even played bro"
I only play like 3 games once per week nowadays. I can only hope to be as wise as youI've played for 24 years and still play 4-6 a week. What magical power is this supposed to imbue me with?
Does anyone else think Frost looks like a fit at 3C, or am I the only one?
Do you even want to win in the playoffs?Does anyone else think Frost looks like a fit at 3C, or am I the only one?
Does anyone else think Frost looks like a fit at 3C, or am I the only one?
I only play like 3 games once per week nowadays. I can only hope to be as wise as you