Speculation: Atlantic Division Predictions

[Ashdown2]

Leafs are always (in the Matthews era) top 2 in their division in those predictions. Often with the 3rd or 4th most points in the entire league.

i didnt mention the leafs . thats not where the issues of these metrics come from

 

[Fogelhund]

Is that predicting the Leafs win the Presidents Trophy?

Speaking from a stats perspective, it's predicting that the Leafs have the highest probability to win the President's Trophy. The standard deviations around those numbers, give a suggested range of outcomes.

 

[All Mod Cons]

i didnt mention the leafs . thats not where the issues of these metrics come from

No, I mentioned the Leafs. Every year these graphs have us destroying the NHL (sans Tampa, Colorado and Vegas) and every year we don't.

They're obviously missing something when calculating these metrics. What it is? Who knows. But they're definitely wide of the mark when it comes to our Leafs.

 

[Nineteen67]

Speaking from a stats perspective, it's predicting that the Leafs have the highest probability to win the President's Trophy. The standard deviations around those numbers, give a suggested range of outcomes.

And the Flyers finish 31st?

 

[Notsince67]

After being in a relatively easy division last year and coming out on top, the Leafs are back in the Atlantic.

How do you see the Division playing out?

My guess,

1. Tampa
2. Toronto
3. Florida
4. Boston
5. Montreal
6. Ottawa
7. Detroit
8. Buffalo

IMO, 2, 3 , 4 is a toss up as is 7, 8.

Pretty good guess but the #1 position is rarely a runaway from the second. 1 thru 3 could be a toss up for first.

 

[Fogelhund]

No, I mentioned the Leafs. Every year these graphs have us destroying the NHL (sans Tampa, Colorado and Vegas) and every year we don't.

They're obviously missing something when calculating these metrics. What it is? Who knows. But they're definitely wide of the mark when it comes to our Leafs.

20/21

https://hockeyviz.com/txt/preview2021

Team	Mean points	Standard Deviation
Leafs	66.8	7.0
Canadiens	63.1	7.0
Jets	62.4	7.1
Flames	61.4	7.1
Oilers	60.7	7.1
Canucks	58.0	7.1
Senators	55.5	7.1

[TBODY] [/TBODY]

Actual

	PTS
TOR	77
EDM	72
WPG	63
MTL	59
CGY	55
OTT	51
VAN	50

[TBODY] [/TBODY]

 

[SprDaVE]

Statistically, the Leafs added of one the better free agents at a fraction of a price --- Michael Bunting. They also have 2 goaltenders that had fantastic seasons. That skews the projections a bit. But realistically, Bunting won't have that huge of an impact.... probably.

 

[Ashdown2]

No, I mentioned the Leafs. Every year these graphs have us destroying the NHL (sans Tampa, Colorado and Vegas) and every year we don't.
They're obviously missing something when calculating these metrics. What it is? Who knows. But they're definitely wide of the mark when it comes to our Leafs.

your nitpicking if you are complaining where the leafs are predicted, wether they are 5th 6th or 7th overall isn't that egregious,
there are other predictions in that list are kind of insane

how anyone thinks the kraken will finish top 3 in the west is asinine.

 

[Fogelhund]

And the Flyers finish 31st?

Again a statistical probability, with a range of likely outcomes. It’s saying, of all of their possible outcomes, the median of the results gives them 31st place points. That’s not the same as saying, hey we think they’ll finish 31st. Hopefully that makes sense, it’s kind of math/stats nerd speak.

 

[notDatsyuk]

And the Flyers finish 31st?

One model shows Philadelphia with the second-worst record in the league, and the other shows them winning the Metro.

If 60 people predict the result of a die roll, the 10 who guessed right will brag, the others will be silent, and everyone will get the impression that most of the predictions were right.

 

[Nineteen67]

Again a statistical probability, with a range of likely outcomes. It’s saying, of all of their possible outcomes, the median of the results gives them 31st place points. That’s not the same as saying, hey we think they’ll finish 31st. Hopefully that makes sense, it’s kind of math/stats nerd speak.

I wonder how they come up with their range of likely outcomes. Maybe they know something but it doesn’t seem like The Flyers would be have ver many likely outcomes close to gathering 31st place points.

 

[Zybalto]

I wonder how they come up with their range of likely outcomes. Maybe they know something but it doesn’t seem like The Flyers would be have ver many likely outcomes close to gathering 31st place points.

I doubt Carter Hart has as bad a season as last year and even if he's in the midst of another sticker, he wont be given the rope to hang the team again.

 

[Dekes For Days]

I doubt Carter Hart has as bad a season as last year and even if he's in the midst of another sticker, he wont be given the rope to hang the team again.

Unfortunately, their other goalie is Jones.

 

[Zonk]

I doubt Carter Hart has as bad a season as last year and even if he's in the midst of another sticker, he wont be given the rope to hang the team again.

I am guessing that the reason there is such a wide range on how the Flyers will do depends on how the goal tending is evaluated by the different models. If a model puts a great deal of emphasis on how the goalies did last year, the Flyers will be horrible. I tend to agree that Carter Hart will not be as bad as he was last year, but if he sucks how much faith do you have in Martin Jones as his backup? I have no idea why Philly management thought that it was a good idea to sign Jones.

 

[Zybalto]

Unfortunately, their other goalie is Jones.

I am guessing that the reason there is such a wide range on how the Flyers will do depends on how the goal tending is evaluated by the different models. If a model puts a great deal of emphasis on how the goalies did last year, the Flyers will be horrible. I tend to agree that Carter Hart will not be as bad as he was last year, but if he sucks how much faith do you have in Martin Jones as his backup? I have no idea why Philly management thought that it was a good idea to sign Jones.

The big difference between the two is that Jones put up an .896 save% on an awful defensive team in the Sharks (7th most shots against and 4th highest expected goals again in the league) and Hart put up an .876 on a team that had above-average defensive metrics. (15th best for shots and 12th best for expected goals). Philly management probably betting on Jones being better than his numbers suggested and will do some heavy lifting while Hart works getting back to what he was. I don't see them in last place myself.

 

[zeke]

I'd imagine the hardest thing to do for the flyers is predict goaltending.

Also how they choose to deploy Risto vs Ellis.

 

[zeke]

I'd imagine the hardest thing to do for the flyers is predict goaltending.

Also how they choose to deploy Risto vs Ellis.

 

[Zybalto]

I actually wonder more why Carolina would have tossed away the young goalie who finished 3rd in Calder voting to throw that big contract at Andersen who had a worse save% than Jones on a good defensive team.

 

[Zybalto]

Speaking from a stats perspective, it's predicting that the Leafs have the highest probability to win the President's Trophy. The standard deviations around those numbers, give a suggested range of outcomes.

Good find.

I like this one:

Team	Playoff Chance
Leafs	95%
Lightning	83%
Panthers	70%
Bruins	60%
Canadiens	42%
Senators	36%
Wings	21%
Sabres	10%

[TBODY] [/TBODY]

 

[Fogelhund]

I wonder how they come up with their range of likely outcomes. Maybe they know something but it doesn’t seem like The Flyers would be have ver many likely outcomes close to gathering 31st place points.

This was their comment on the Flyers.

The Flyers re-worked their defence corps over the summer without any obvious net improvement and even a hefty "bounce-back" for Carter Hart, their presumptive starter, still leaves him well below average given how dire he was last season.

The values I use for goaltender and shooter ability are not as easily interpretable as the previous measurements, but they can still be understood with odds ratios; for instance, the odds of a shot taken against a Philadelphia or Buffalo goaltender becoming a goal are 6% higher than a similar shot taken against a league-average goaltender

---------------------------

In summary, it appears that their statistical model, based on last years numbers, where Hart was bad, and their D wasn't good either... are the downfall. Again people are looking at this as an absolute, instead of seeing that it is a range of possibilities. If Carter and the D are just as bad as last year, that's the median likely result. Change the parameters, and of course the results change. Historic statistical models, aren't predictors of changes in play, they aren't based on gut feelings. Variances in actual vs. expected are easy to identify in hindsight... well, Carter actually was decent this year... and if you believe personally, for whatever reason that players, teams, coaches, goalies, will regress, or progress, than you can personally handicap that.

 

[zeke]

I actually wonder more why Carolina would have tossed away the young goalie who finished 3rd in Calder voting to throw that big contract at Andersen who had a worse save% than Jones on a good defensive team.

Carolina had an extremely weird/awful offseason all around.

 

[Zybalto]

I can say one thing for Hart, he was average at saving pucks flipped into him..other than that....

Out of the 51 goalies with 1000+ minutes last year:

Low Danger save%:

27. Hart

Medium Danger save%:

51. Hart

High Danger save%:

51. Hart

Save % on the PK:

51. Hart


Woof.

 

[Lightsol]

I think 1-4 can be a toss up.. Florida had more regular season points than Tampa last year. It's going to be a dog fight all year, which is good playoff prep.

Buffalo depends on Eichel to some degree... does he get traded, what's the return, or is he on LTIR? I don't think Buffalo can avoid the #8 spot, and might be generationally bad.

Detroit has some good young players, and if Matt Murray's form isn't great in Ottawa, Detroit could challenge them.

1. Florida
2. Tampa
3. Toronto
4. Boston
5. Montreal
6. Detroit
7. Ottawa.






8. Buffalo

Yeah, I'd probably go:

1. Florida
2. Tampa
3. Toronto

4. Boston

5. Montreal

6. Detroit
7. Ottawa.






8. Buffalo

 

[Zybalto]

At least we arnt getting screwed in this way again. Habs getting the breaks and poor Ottawa but at least they are a young team.

 

[Dekes For Days]

The big difference between the two is that Jones put up an .896 save% on an awful defensive team in the Sharks (7th most shots against and 4th highest expected goals again in the league) and Hart put up an .876 on a team that had above-average defensive metrics. (15th best for shots and 12th best for expected goals). Philly management probably betting on Jones being better than his numbers suggested and will do some heavy lifting while Hart works getting back to what he was. I don't see them in last place myself.

I don't necessarily see them in (second) last place either, but I'm not sure betting on a goalie who has been below 0.900 over his last 157 games in the NHL to do "heavy lifting" was a great idea, and I could see how a model that heavily weights recent goalie results could be down on Philly.