HF Habs: 2020 Montreal Canadiens Off-Season Thread part 3

[mynamejeff420]

Now do Mete and Kulak. If they have better numbers it’ll be proof positive that metrics aren’t all mighty tool they’re made out to be just yet.

There's aren't "metrics", these are just a reflection of what happened when a player was on the ice. All it's showing you is where the shots are coming from when a given player is on the ice. It's hard to claim that a player is "good at clearing the front of the net" if they allow a high rate of net front shots.

Maybe Mete and Kulak have low amounts of net front shots because they're actually good defensively, better than you think they are.

Also just for transparencies sake, their isolated maps (ie adjusted for context) show the same thing: they're good at preventing net front shots.

 

[mynamejeff420]

Nah. Chiarot is not just some bottom pairing guy. If you think that about him, we are not going to be on the same page with Edmundson. Let me guess, you like Ghost type D man?

Are you sure you are not Dubas?

I don't think Gostisbehere is very good either tbh. I don't think there's one good type of defenseman. There's good puck moving defensemen, who excel despite a lack of physical strength (Quinn Hughes for example), and I think there's good physical defensemen (Shea Weber for example). Just because I don't think Chiarot is very good doesn't mean I don't think all physical defensemen are bad, there's plenty of good top 4 defensemen who play a physical game, Chiarot just isn't one of them (imo).

 

[DAChampion]

Fair enough, you're absolutely right, context is important.

View attachment 369905

This is Edmundson's isolated impact (so contextual factors are taken into account), and as you can see, he still had a problem with preventing net front shots.

After allegedly accounting for teammates, competition, and coaching, ... We get maps that look like a combination of noise and of negating the original argument.

Now it looks like Edmundson is a good PKer, but not really, because these maps have no uncertainties and no explanation of methodology.

 

[Habs Halifax]

I don't think Gostisbehere is very good either tbh. I don't think there's one good type of defenseman. There's good puck moving defensemen, who excel despite a lack of physical strength (Quinn Hughes for example), and I think there's good physical defensemen (Shea Weber for example). Just because I don't think Chiarot is very good doesn't mean I don't think all physical defensemen are bad, there's plenty of good top 4 defensemen who play a physical game, Chiarot just isn't one of them (imo).

Both Chiarot and Edmundson at $3.5M are great 4/5 guys that can be effective playing with guys like Weber and Petry. You just don't like the ones who are physical and don't put up points. It's clear

 

[mynamejeff420]

After allegedly accounting for teammates, competition, and coaching, ... We get maps that look like a combination of noise and of negating the original argument.

Now it looks like Edmundson is a good PKer, but not really, because these maps have no uncertainties and no explanation of methodology.

It appears as though I did underrate his ability to play on the PK, so my mistake there. His play defensively at even strength though is still bad. That take held up after accounting for contextual factors.

If you're curious about the methodology, there's a long writeup that goes in to great detail about the methodology, I'll link it here so you can't read it. The model makers (both this one on HockeyViz and the one on Evolving-Hockey) are very open about the methodology.

 

[DAChampion]

It's not like I'm digging in the bottom of the barrel to find obscure, cherry picked stats to prove a point, I'm looking at pretty basic ones, like Goals and Expected Goals (both for and against). If a player is on the ice for a lot of Goals For (GF), they're probably doing something right. If a player is on the ice for a lot of Goals Against (GA), they're probably doing something wrong. Same goes for Expected Goals (xG). From there, you'd want to adjust for context, so you can look at something like RAPM. The simplistic explanation of RAPM is that it takes a metric (like GF) and "isolates" it to a player's individual impact, taking things like teammates, competition, usage, zone starts, team strength, etc. into account (ie contextual factors). The full write-up on it is here if you're curious. Feel free to ask about any other stats or terms I've used that you feel I haven't explained adequately enough, or ones you don't understand.




I think Pavelski's playoff production this year had a little bit more to do with him nearly doubling his on-ice shooting percentage (7.27% in the regular season 13.13% in the playoffs), though that's not to say that he isn't a good player (or even that he's someone who elevates their game in the playoffs). I think he was a little unlucky in the regular season, shooting percentage wise. I also don't think Byron is a better player than Joe Pavelski, I just wanted to point out that Joe "top 6 forward" Pavelski and Paul "4th liner" Byron produced at similar rates at 5v5 over the last 3 years, so maybe people are underrating Byron (who, by the way, had back to back 20 goal seasons in 16/17 and 17/18, and was on pace for 22 goals in 18/19 before the injury). In my opinion, a reliable 20 goal scorer (despite not getting PP time) is a valuable asset to a team's top 6, especially when said team has issues with scoring goals. That's all I'm trying to say. Not that Paul Byron is better than Joe Pavelski, but that people are underrating the guy that scores at a higher rate than Joe Pavelski.

As for Danault calling playing with Byron and Lehkonen a "nightmare situation", I don't think that's really accurate. He expressed disinterest in playing in a "shutdown" role, which it appears the 3rd line would be, a shutdown line. I think it has a lot more to do with the usage of the line than the wingers sucking offensively.




While I won't try and pretend that Danault is some secret PP god, him "not being able to play on the PP" isn't really accurate imo. The only forward on the Habs with a higher primary assist rate on the powerplay over the last 2 seasons is Jonathan Drouin. I think the coaching staff feels that they're better served by having Danault save his energy for tough 5v5 matchups and PK time, so they don't play him on the PP. If he got regular PP time I'm sure he'd put up the point totals people expect out of a #1C, because he puts up #1C caliber point totals at 5v5 (he was 9th among Cs in 5v5 points this year).

A player playing on the first line is doing so at 5v5 (that's where lines 1/2/3/4 are deployed. So a "first line" player doesn't necessarily have to play on the powerplay, as long as they're playing up to snuff at 5v5. Danault not being a PP option doesn't mean he sucks at 5v5 (in fact, he's great at 5v5).

Nick Suzuki played with a variety of linemates this season, but was mainly playing with players who many would consider to be "top 6 forwards". When he was on the ice, the Habs scored 1.91 GF/60 at 5v5, and got 43.42% of the goals, or had a GF% of 43.42. When Phillip Danault was on the ice, the Habs scored 3.59 GF/60 and had a GF% of 60.21. If you're concerned about trying to win games 2-1 when you're behind 2-1 or 3-1, then Danault is a much better option for the 1C spot than Suzuki. Danault produced at 5v5 this year, Suzuki didn't. Danault was the stellar two-way C this year, Suzuki was the defense only C. The Habs top line was one of the best lines in the league this year, in part thanks to Danault's contributions. Having him play on the top line is not an issue, or else the line wouldn't be putting up a near 60 GF% with them on the ice.

You use dozens of different stats (poor methodology), and whenever you introduce new ones you fail to define them.

It's technically true that 80% of the game is played at even strength. But your stating of that is a demonstration of a poor instinct of statistics. You're not weighing those minutes. Is even strength where 80% of goals are scored or allowed? No. Is it where 80% of energy is expenses of where 80% injuries happen? Probably not. So no it's not sensible for a defenseman to be a 5on5 specialist. All that does is add pressure and fatigue on the other defensemen -- which you would consequently misinterpret as weaker effectiveness at 5on5.

 

[mynamejeff420]

You use dozens of different stats (poor methodology), and whenever you introduce new ones you fail to define them.

It's technically true that 80% of the game is played at even strength. But your stating of that is a demonstration of a poor instinct of statistics. You're not weighing those minutes. Is even strength where 80% of goals are scored or allowed? No. Is it where 80% of energy is expenses of where 80% injuries happen? Probably not. So no it's not sensible for a defenseman to be a 5on5 specialist. All that does is add pressure and fatigue on the other defensemen -- which you would consequently misinterpret as weaker effectiveness at 5on5.

Tell me which stats I've failed to define. Goals are pretty self explanatory, I didn't realize I needed to define those. If you're curious, a "goal" is when the puck enters the net.

While it is true that ~80% of the game is played at 5v5, the goal share is lower, with ~70% of the goals coming at 5v5. Why is it more sensible to have a player that excels in 30% of the game and stinks for 70% of the game than a player who excels in 70% of the game and stinks in 30% of the game? Last time I checked, 70 is a bigger number than 30.

 

[mynamejeff420]

Both Chiarot and Edmundson at $3.5M are great 4/5 guys that can be effective playing with guys like Weber and Petry. You just don't like the ones who are physical and don't put up points. It's clear

That's a total strawman argument. I've never claimed that all physical defensemen who put up points are bad. I don't hate physical defensemen who don't put up points, and I can name a ton of good ones off the top of my head. "Good defensemen" isn't a title exclusive to non-physical, puck moving defensemen who put up points. There's lots of ways to be an effective defenseman. Just because Chiarot and Edmundson aren't that good (though lumping them together is a little unfair, Chiarot is mediore and Edmundson is bad imo) doesn't mean that all physical defensemen who don't put up points are bad. Just those two.

 

[Habs Halifax]

That's a total strawman argument. I've never claimed that all physical defensemen who put up points are bad. I don't hate physical defensemen who don't put up points, and I can name a ton of good ones off the top of my head. "Good defensemen" isn't a title exclusive to non-physical, puck moving defensemen who put up points. There's lots of ways to be an effective defenseman. Just because Chiarot and Edmundson aren't that good (though lumping them together is a little unfair, Chiarot is mediore and Edmundson is bad imo) doesn't mean that all physical defensemen who don't put up points are bad. Just those two.

So the only 2 guys who play physical and don't put up points who are "bad at everything else" are Chiarot and Edumndson in the NHL? .

 

[DAChampion]

Tell me which stats I've failed to define. Goals are pretty self explanatory, I didn't realize I needed to define those. If you're curious, a "goal" is when the puck enters the net.

While it is true that ~80% of the game is played at 5v5, the goal share is lower, with ~70% of the goals coming at 5v5. Why is it more sensible to have a player that excels in 30% of the game and stinks for 70% of the game than a player who excels in 70% of the game and stinks in 30% of the game? Last time I checked, 70 is a bigger number than 30.

We were just talking about these heat maps, which you're using without defining.

I never wrote anything about having PK or PP specialists on D. I wrote something different, that a defense can't carry a 5on5 specialist.

 

[mynamejeff420]

So the only 2 guys who play physical and don't put up points who are "bad at everything else" are Chiarot and Edumndson in the NHL? .

No there's good physical players and bad physical players. Just like there's good skill players and bad skill players. Chiarot and Edmundson aren't bad because they're physical, they're bad for other reasons (namely positioning in the defensive zone, and an inability to get the puck out of the defensive zone once possession is regained).

 

[mynamejeff420]

We were just talking about these heat maps, which you're using without defining.

I never wrote anything about having PK or PP specialists on D. I wrote something different, that a defense can't carry a 5on5 specialist.

Why can a defense carry a PK specialist but not a 5v5 specialist?

 

[DAChampion]

Why can a defense carry a PK specialist but not a 5v5 specialist?

I haven't seen anybody suggest that a defense should carry a special teams specialist.

 

[mynamejeff420]

I haven't seen anybody suggest that a defense should carry a special teams specialist.

That's what Edmundson is though. Good on the PK, bad at 5v5.

 

[Grate n Colorful Oz]

Gaston Therrien on why we should extend Danault:

“He’s un québécois, and he can speak to the public after the game”.....end quote



it wasn’t talent, it wasn’t faceoffs, it was about language again......

Gaston doesn't even hide his self-serving nature.

 

[Sterling Archer]

There's aren't "metrics", these are just a reflection of what happened when a player was on the ice. All it's showing you is where the shots are coming from when a given player is on the ice. It's hard to claim that a player is "good at clearing the front of the net" if they allow a high rate of net front shots.

View attachment 369906

View attachment 369908

Maybe Mete and Kulak have low amounts of net front shots because they're actually good defensively, better than you think they are.

Also just for transparencies sake, their isolated maps (ie adjusted for context) show the same thing: they're good at preventing net front shots.
View attachment 369909

View attachment 369910

Or maybe it's because they play against weaker opposition and maybe it's because your charts just aren't very good.

I have literally seen Mete and Kulak run around, get bodied and not able to handle players pushing them out of the way to get to the net. There's a reason they're not trusted to close out games and play the PK unless there is literally no other option available. Both players can barely stay in the lineup on a team with no D depth. I haven't seen it to anywhere near that extent with Edmundson and in fact, when I did watch him, I was almost always left impressed with his coverage and play.

So the day I trust a chart telling me something that's diametrically opposed to what I'm literally seeing with my own eyes, is the day I stop thinking for myself and yield completely to chat board posts and stop thinking for myself. If that happens, pull the plug on me.

 

[DAChampion]

That's what Edmundson is though. Good on the PK, bad at 5v5.

You changed your mind on Edmundson's PK skills pretty quick following a scintilla of counterargumentation.

Regarding his even strength play there have already been posts on that subject. He wasn't as effective in Carolina as he had been in St Louis due to system changes.

 

[Habs Halifax]

You changed your mind on Edmundson's PK skills pretty quick following a scintilla of counterargumentation.

Regarding his even strength play there have already been posts on that subject. He wasn't as effective in Carolina as he had been in St Louis due to system changes.

Basically, what he is saying is you should have the top of your line-up as stars and taking up 90% of the saalry and guys like Chiarot and Edmundson should make $1M only. He doesn't like middle core type players and paying them $3.5M.

 

[Adam Michaels]

I don't know why, but I feel like the Gallagher extension will be announced today or this weekend.

 

[417]

I don't know why, but I feel like the Gallagher extension will be announced today or this weekend.

You're probably not too far from the truth.

 

[DAChampion]

Do you guys agree with the policy of extending players as early as possible?

I think it's better than Gainey's policy of virtually never negotiating during the season.

 

[CHfan1]

Do you guys agree with the policy of extending players as early as possible?

I think it's better than Gainey's policy of virtually never negotiating during the season.

Depends on the player. If they’ve been consistently good, season after season, and the organization doesn’t think they’ll decline, I like to see the player extended as early as possible.

If they just had their best season I would want to see the Habs wait and see if the player can replicate that season before extending him (Domi for example).

This season is also unique with the flat cap. A team may be able to get a bargain extending players.

 

[DramaticGloveSave]

I don't know why, but I feel like the Gallagher extension will be announced today or this weekend.

Seems that’s the focus right now. Could see a trade soon after that- though might depend on the draft and who’s available at 16 as well.

 

[salbutera]

Do you guys agree with the policy of extending players as early as possible?

I think it's better than Gainey's policy of virtually never negotiating during the season.

Yup - this era of millennials in general are not suited to handling being told to wait for anything.

 

[DAChampion]

Yup - this era of millennials in general are not suited to handling being told to wait for anything.

Ok boomer.