Contrarian view: puckanalytics.com.

[VirginiaMtlExpat]

A few juicy, contrarian tidbits from puckanalytics.com.

Brace yourselves. A picture is worth a thousand words, so here are a couple thousand. Team Synopsis: Montreal Canadiens – August 16, 2022 - Puck Luck Analytics

First off, they wrongly foresaw that Price would play, so they inflated the Habs outlook.

Then, they somehow came to the conclusion that Gallagher is going to have a big year. Could be a sell-high opportunity.

Click to read more...

 

[BigDaddyLurch]

...wow...when did the Season change to 120 Games, cuz that's about where we'd need it to be to reach 91 points...

 

[jrom]

At first I was thinking "ok Price really skews results" but that doesn't explain their model predicting Florida/Boston as top2 in the division (and Ottawa doing so poorly).

 

[OnTheRun]

Top4 most impactful forwards:
1. Gallagher
2. Dadonov
3. Armia
4. Drouin

Yeah... I'm out of here.

 

[ArtPeur]

...wow...when did the Season change to 120 Games, cuz that's about where we'd need it to be to reach 91 points...

I'm not sure 120 games would be enough

 

[BigDaddyLurch]

Top4 most impactful forwards:
1. Gallagher
2. Dadonov
3. Armia
4. Drouin

Yeah... I'm out of here.

...Cannabis is legal in Canada, so...that's pretty much the only explanation for this atrocity of an article...

I'm not sure 120 games would be enough

...originally I had it at 130 games, but I'm an optimist, so...

 

[holy]

I could’ve told you that without the fancy analytics. All heart no brain babyyyy.

 

[Belial]

I have no idea what kind of numbers you have to manipulate or what sort of model you have to use to be able to come to the conclusion that Pezzetta is more impactful than Suzuki!

 

[BehindTheTimes]

I’m hoping no one is paying for this analysis. Sounds like a click bait site that is desperately seeking attention.

 

[Habs Icing]

Forget about the predictions. What I want to know is who the hell is Mitchell Stephens and how did we get him and where the puck was I when we got him?

 

[GrandBison]

Top4 most impactful forwards:
1. Gallagher
2. Dadonov
3. Armia
4. Drouin

Yeah... I'm out of here.

I'm confused, if it means that their level of play will dictate the kind of season Habs are going to have, I agree. If they are considered as great offensive contributors, well...

 

[Sterling Archer]

So the last place team, in the midst of a rebuild is projected to be a below average team!? Mind blown!

 

[beowulf]

Analytics remain a fun little distraction but are far from something to make big decisions on.

 

[abo9]

Analytics remain a fun little distraction but are far from something to make big decisions on.

It can be helpful, because you can't possibly watch all players for 82 games so it helps painting an overall portrait of players, teams etc.

But, one has to validate their results with common knowledge, or at least address the obviously weird ones.

Suzuki being less inpactful than 3/4 of the roster is an obvious red flag. Whats going on there?

 

[Ghetto Sangria]

It can be helpful, because you can't possibly watch all players for 82 games so it helps painting an overall portrait of players, teams etc.

But, one has to validate their results with common knowledge, or at least address the obviously weird ones.

Suzuki being less inpactful than 3/4 of the roster is an obvious red flag. Whats going on there?

His advanced stats were a complete disaster while facing tough competition last season. Didn't help that he had to take on the Danault minutes while being expected to produce at the same time. The eye test shows this will change when he gets a good enough supporting cast to take some attention away from him

Maybe Dach or DVo (or both) help alleviate some of his defensive responsibilities 5 on 5.

 

[tazsub3]

if we had a defence then i might see it. I think we more then ok up front.
But no elite goaltending to go with a very very avg to inexperienced d, is not usually a recipe for a team to overperform

 

[Paddyjack]

I have no idea what kind of numbers you have to manipulate or what sort of model you have to use to be able to come to the conclusion that Pezzetta is more impactful than Suzuki!

I think the graph is about "replacement impact", meaning that the impact this player will have on the team this season compared to a player who had the same role last season. Suzuki being #1C last season and this season has thus a lower replacement impact than say Gallagher who had a terrible season but they expect him to be better now.

 

[Kairi Zaide]

I think the graph is about "replacement impact", meaning that the impact this player will have on the team this season compared to a player who had the same role last season. Suzuki being #1C last season and this season has thus a lower replacement impact than say Gallagher who had a terrible season but they expect him to be better now.

I've never heard of the website in OP, nor of the author of said website. I don't know what he uses or how he models his stuff, but I haven't seen any validation he may have made posted on his website, and he only has one season of track record. His stuff doesn't seem, at first glance, aligned with other analytics people's models.

In the analytic community, the term "replacement player" refers to the idea of a player who is readily available to play. Think of it as a 13th forward or a 7th d-man caliber player.

Basically, "G/60 Impact Above Replacement" would mean "How much a player contributes to their team's goal scoring rate (G/60) compared to a readily available replacement forward who would be deployed in the same situations (opposition, linemates, ice time)".

The idea of using a "replacement player" as a baseline is that, in theory, NHL caliber players (which includes 4th liners) should be expected to perform above a replacement player's level. In reality, however, because there's almost no difference between average 4th liners and replacement players, they're pretty much the same.

 

[Paddyjack]

Basically, "G/60 Impact Above Replacement" would mean "How much a player contributes to their team's goal scoring rate (G/60) compared to a readily available replacement forward who would be deployed in the same situations (opposition, linemates, ice time)".

The idea of using a "replacement player" as a baseline is that, in theory, NHL caliber players (which includes 4th liners) should be expected to perform above a replacement player's level. In reality, however, because there's almost no difference between average 4th liners and replacement players, they're pretty much the same.

You are likely correct, and I don't know much about analytics but I got the sense from his graphs that these players will contribute to a greater impact on the team this season compared to what was played in the same role last season. That he is correct or not to use this term for it I don't know. But that explains why Suzuki is low like that.

 

[LaP]

It can be helpful, because you can't possibly watch all players for 82 games so it helps painting an overall portrait of players, teams etc.

But, one has to validate their results with common knowledge, or at least address the obviously weird ones.

Suzuki being less inpactful than 3/4 of the roster is an obvious red flag. Whats going on there?

It's helpful at predicting one aspect of the game when it's based purely on math. Like PDO which is a logical objective advanced stats purely based on math. A unusually high PDO will often lead to regression specially for non elite players. But when it tries to predict the whole outcome using subjective terms it is where it falls apart.

I cringe every time i see an advance stats like "Shot from a scoring chance angle with a stick not longer than one player's arm on a Friday night of a full moon when the opposing player is no taller than his mother and with arms bigger than legs".

Some advanced stats are so subjective and not clearly mathematically defined that they are pretty much useless. Basically they are as good as the eye test i.e. as good as the guy watching the game can evaluate hockey. Like i'll always say i'd take a shot from Ovechkin from outside the faceoffs circle close to the board before a high danger scoring chance shot right in front of the net from Paul Byron where the goalie is ready to make the save.

 

[Gaylord Q Tinkledink]

I’m hoping no one is paying for this analysis. Sounds like a click bait site that is desperately seeking attention.

People are.

People always slap some shit from jfresh who's an idiot, or posts shit Dom says from the athletic

 

[Treb]

Forget about the predictions. What I want to know is who the hell is Mitchell Stephens and how did we get him and where the puck was I when we got him?

He signed with us on July 13th.

Played for Tampa an Detroit. Will most likely be in Laval this year.

 

[Miller Time]

Allen
Barron/Harris/Guhle
Gally/Armia/DVO/Anderson/Drouin/Dadonov/Hoffman/Monahan

If, and it's a big IF, we get solid goaltending, 2 of those 3 rookie D's play solid bottom 4 caliber level, and 3-4 of those 8 vets play @/above their career average productivity level... Habs are a playoff team this year.

It's unlikely, but unlike a yotes or kravken situation, the talent on our cap max roster is good enough to be a middle tier team this season.

3/4 of last year was absolutely worst case, and when MSL took over, we stabilized despite removing 3 contributing vets, inserting a bunch of young players, implenting a new system on the fly and having to play Monty a ton.

With a full camp, a healthy roster full of mid career vets (5 of them playing for one last big contract), and some promising young players ready to take forward steps in their careers, we very well could surprise a lot of pundits.