The Athletic has ranked their top 30 NHL teams... Leafs and Flames at the top

[Sidney the Kidney]

I mean, last season was the most bifurcated season of the modern era between playoff and non-playoff teams - it was "weird" in that sense. All the public models were off more than usual, yet Dom's model was still the most accurate and still beat the sportsbook odds. You had MSM people saying Winnipeg and Ottawa were making the playoffs, lmao. If anyone had a perfect standings prediction, I'd like to see it.

Dom's model also had the Stanley Cup Winner predicted from his top-2 ranked teams 5 of the 6 last seasons.

Who didn't have the teams who won the Cup the past 6 seasons as one of their top 2 or top 3 teams? Other than St. Louis, those are the teams I would have had in my top 2 or top 3 heading into those respective seasons, wouldn't you? Being correct about something someone who doesn't use a statistical model likely predicts doesn't mean the criticisms of his model aren't justified.

Just like going into this season. If my top 5 (off the top of my head) are Tampa, Colorado, Carolina, Toronto and Edmonton, for example and one of those teams goes on to win the Cup, that doesn't mean I'm some genius with a top notch method of determining my rankings. So a model being "correct" about teams who were consensus top 2/top 3 teams going into those seasons (other than the Blues - which his model also got wrong) isn't exactly proof of its strength.

 

[AvsFan29]

Isn't this based on regular season? If so, Toronto at 1 isn't a bad take

The Leafs lost the least out of the top teams except for maybe Carolina and are essentially the same team as last year. Based on that, winning the Presidents isn't far off for them

Look at the quality of competition in the east that Toronto will play the majority of their games against.

Look at the absolute garbage that Colorado and Calgary will play the majority of their games against.

I'd wager Colorado and Calgary put up more points than Toronto.

 

[Yepthatsme]

Link for anyone curious:

2022-2023 Season Preview

Visualizations and predictions for understanding hockey.

hockeyviz.com

This is how that model predicts the Pacific.
Just going through Calgary, some gems I’ve found after sifting through it quickly is that Edmonton’s goaltending is projected to be better than Calgary’s, that Tyler Toffoli is projected to generate A LOT more offense on a power play than Huberdeau, and that Darryl Sutter’s effect on the team is that we actually allow more shots than we would with average coaching. Also that Zadorov is better defensively than Tanev.

Not saying Don’s model is good, but this one is terrible.

 

[tfwnogf]

2. Calgary Flames
27. Columbus Blue Jackets

Johnny Bros we f***ed up

 

[bossram]

Ah, a fellow bettor who understands.

Dom before he got his writing gig was actually a fairly well known hockey sharp in betting circles. He used to make a living betting on hockey. Not a high roller by any means but a very sharp mind.

He's consistently been the most accurate projector among the big media outlets since he started doing these things.

- He was one of the few analysts that predicted the 17-18 Sens missing the playoffs despite their ECF Run.
- He had Montreal as a bottom 10 team going into last year despite their finals run.
- For the Oiler fans here: Last year he had you guys as the 6th best team in the league in a year where almost everyone was low on the team including many of you.
- He had the Leafs as a bubble team in 16-17 despite them finishing last in the league the year prior.

I give this thread about 20 pages before the mods close it.

I 100% agree with you. I'm very sad that Dom will no longer be publishing his daily game predictions because I followed that religiously for the past few seasons and made some decent money betting hockey.

Of the public models and predictors, he is essentially year-in and year-out the most accurate, relative to the sportsbooks odds. If anyone wants to claim his model is terrible, I'd like to see the receipts on their predictions.

 

[leafsfan5]

Look at the quality of competition in the east that Toronto will play the majority of their games against.

Look at the absolute garbage that Colorado and Calgary will play the majority of their games against.

I'd wager Colorado and Calgary put up more points than Toronto.

I wouldn’t necessarily disagree, I personally have Colorado winning the presidents with Toronto #2 in the league due to the conference discrepancy you outlined

I just don’t believe it’s a bad take to have Toronto winning the presidents. It’s an elite team who put up 115 points last year and is essentially the same club

 

[bossram]

Who didn't have the teams who won the Cup the past 6 seasons as one of their top 2 or top 3 teams? Other than St. Louis, those are the teams I would have had in my top 2 or top 3 heading into those respective seasons, wouldn't you? Being correct about something someone who doesn't use a statistical model likely predicts doesn't mean the criticisms of his model aren't justified.

Just like going into this season. If my top 5 (off the top of my head) are Tampa, Colorado, Carolina, Toronto and Edmonton, for example and one of those teams goes on to win the Cup, that doesn't mean I'm some genius with a top notch method of determining my rankings. So a model being "correct" about teams who were consensus top 2/top 3 teams going into those seasons (other than the Blues - which his model also got wrong) isn't exactly proof of its strength.

Picking the Stanley Cup winner is just for demonstration.

His model is the most accurate because its log loss is better than almost every other model, every season. And even outperforms the sportsbook odds.

If someone has a better game prediction model, I'd very much like to see it. And again, I haven't seen anyone make any legitimate criticisms of his model other than "lol, this team is too high/low".

 

[CantHaveTkachev]

I mean, he literally had Colorado #1 last year....

5 of 6 seasons, the Stanley Cup champion was either #1 or #2 in his preseason model.

he also predicated the Flames to finish 4th in the division..

 

[bossram]

This is how that model predicts the Pacific.
Just going through Calgary, some gems I’ve found after sifting through it quickly is that Edmonton’s goaltending is projected to be better than Calgary’s, that Tyler Toffoli is projected to generate A LOT more offense on a power play than Huberdeau, and that Darryl Sutter’s effect on the team is that we actually allow more shots than we would with average coaching. Also that Zadorov is better defensively than Tanev.

Not saying Don’s model is good, but this one is terrible.

Micah's model routinely grades out as one of the worst for predictions every season.

I like his website for the information, but wouldn't use his predictions for betting.

 

[dmac7719]

and therein lies the problem is Dom's "model" and why he's rarely right

He played to the level of a 1C in the NHL last season, so I don't see the problem with Dom's model calling him a 1C last season

Kadri is not a true 1C. He had a career year and flourished in the 2C role. C'mon man.

If you don't think Kadri didn't play to the level/production of a 1C last season, than I don't know what to say. He may have been the 2C on the Avs last season but he finished 9th in scoring for centers. A team can have 2 players on there team that have the production of a 1C.

 

[The Nuge]

Ah, a fellow bettor who understands.

Dom before he got his writing gig was actually a fairly well known hockey sharp in betting circles. He used to make a living betting on hockey. Not a high roller by any means but a very sharp mind.

He's consistently been the most accurate projector among the big media outlets since he started doing these things.

- He was one of the few analysts that predicted the 17-18 Sens missing the playoffs despite their ECF Run.
- He had Montreal as a bottom 10 team going into last year despite their finals run.
- For the Oiler fans here: Last year he had you guys as the 6th best team in the league in a year where almost everyone was low on the team including many of you.
- He had the Leafs as a bubble team in 16-17 despite them finishing last in the league the year prior.

I give this thread about 20 pages before the mods close it.

Nobody in their right mind had Montreal outside of the bottom 10. Most had them bottom 3. That’s not anything to brag about.

And if he had us 6th, and his model thinks that adding Kane, Campbell, Holloway, Kulak etc makes us worse, that’s pretty telling about the flaws in his system

 

[thaman8765678]

Care to explain how the better team lost 4 games in a row?

You are really embarassing yourself this off/preseason with some gems of idiocy.

It is simple. Goaltending is a very important factor and without it you have nothing.

If goaltending was equal on both sides in the series, Flames sweep.

 

[Golden_Jet]

Ah, a fellow bettor who understands.

Dom before he got his writing gig was actually a fairly well known hockey sharp in betting circles. He used to make a living betting on hockey. Not a high roller by any means but a very sharp mind.

He's consistently been the most accurate projector among the big media outlets since he started doing these things.

- He was one of the few analysts that predicted the 17-18 Sens missing the playoffs despite their ECF Run.
- He had Montreal as a bottom 10 team going into last year despite their finals run.
- For the Oiler fans here: Last year he had you guys as the 6th best team in the league in a year where almost everyone was low on the team including many of you.
- He had the Leafs as a bubble team in 16-17 despite them finishing last in the league the year prior.

I give this thread about 20 pages before the mods close it.

Yep had Seattle at 97 points or something as well lol.

 

[Markster]

Leafs #1 according to the Athletic. Time to hang another banner?

 

[Voight]

Seattle is too high IMO.

They essentially have the same roster + Beniers/Wright. Great prospects, but still only prospects. I guess they did add Burakovsky & Bjorkstrand but they aren't needle movers.

 

[Voight]

Someone should tell him that models should be based on reality. He's making a model and hoping that reality will follow it - that's not how it works.

Your post reminds me of this quote, which usually applies to advanced stats.

Theoretical physics can also prove that an elephant can hang off a cliff with its tail tied to a daisy! But use your eyes, your common sense.

 

[93gilmour93]

We’re number 1 we’re number 1 we’re number 1

 

[Face Of Bear]

They got #1 right, not sure about the rest.

 

[bossram]

Seattle is too high IMO.

They essentially have the same roster + Beniers/Wright. Great prospects, but still only prospects. I guess they did add Burakovsky & Bjorkstrand but they aren't needle movers.

This doesn't really make any sense. They're adding three top-six forwards (Burakovsky, Bjorkstrand, Beniers) but that's the "same roster".

I mean, they're bound to improve simply by not getting ECHL goaltending alone.

 

[Tufted Titmouse]

Comments about last year are a bit strange considering how much change went on this season.

 

[Voight]

Cam and Strick did a podcast with Coach Bedner and Cogliano, and if I remember correctly, they both said St. Louis was Colorado's hardest opponent in the playoffs.

The Athletic is a joke. Dom's model predicts roughly a 50% chance the Blues make the playoffs. They are a top 3-4 team in the West.

I get Binnington has struggled but I don't get why people sleep on the Blues. Minus Piets (who was obviously a big loss), they have the same core from 3 years ago when they won, with guys like Thomas & Kyrou getting even better.

 

[Rcknrollkillnmachine]

Amazing what a handshake line will do .

 

[LeafGrief]

Surprised we're at #1 with the goaltending we have. Our skaters are as good as anyone's, however.

Unfortunately for the Leafs, playoff rounds aren't won on paper and aren't won based on how many good players you have. Having us at #1 is a hot take, but the Althetic writers also aren't holding 800 Alex Kerfoot trade threads against us.

Carolina and Tampa should be quite a bit higher.

 

[Video Nasty]

Look, this list is mostly pedestrian, outside of a few picks where they don’t seem to realize that the core is only aging and/or has injuries or is injury prone (Bruins and Penguins), but I don’t see why some are so up in arms about Toronto and Calgary leading the way with Colorado filling in the #3 spot.

As much fun as we all have dogging on Toronto and the nuisances inside their fanbase, their first round struggles should have nothing to do with the perception of that team in the regular season. There’s little reason to expect them to be any worse than the team that finished 4th in points, tied for 3rd in Wins, 3rd in regulation Wins, 2nd in goals and so on. If anything, they have every motivation in the world to gun for the #1 seed to draw a great matchup.

I’m going to be fair to Calgary and assume that they all weren’t mentally destroyed by the Oilers defeat. They did a great job replacing their top 2 guys and they seem to have made a lateral move at worst. It’s not crazy to me that they put up a similar record.

Colorado are the champions and I doubt their motivation is going be driven by the desire for another President’s trophy. I can see them slipping a handful of points off of purely not having their foot on the pedal for all 82 regular season games.

 

[theoriginalBCF]

I think it's pretty accurate. I would move the Flames down myself, I think they're so over rated, and the loss of Tkachuk and JG are going to be felt hard, factoring in the team had career years, like they did 4 seasons ago, 2018/19, then missed for 2 while they over hauled the Defense.

Other than that, I'd say it's pretty decent.