Toggle sidebar

Jets Advanced Stats thread

  • Xenforo Cloud will be upgrading us to version 2.3.5 on March 3rd at 12 AM GMT. This version has increased stability and fixes several bugs. We expect downtime for the duration of the update. The admin team will continue to work on existing issues, templates and upgrade all necessary available addons to minimize impact of this new version. Click Here for Updates
surixon said:
Yeah Maurice has a bit of a blind spot on Wheeler. I get the loyalty thing and I get him wanting to get Blake going but I think it's more the coach being in denial at this point.

I think if you cut Blake's minutes down to 17 and play him against softer comp and he is more able to maintain his energy levels throughout a game and likely perform better.

Dubois is fully capable of winning tough matchups.

Ehlers/Copp/Schiefle could also handle tough matchups imo.

I'd try to build a soft scoring line around Stastny and Wheeler and grinding fourth line around Lowry.

I think that moves us towards the Tampa model and I think puts everyone in a spot to succeed.
Click to expand...


Agree with this, it would be my preferred lineup too. I'd love to maybe see Perfetti with Stastny and Wheels by season's end.
 
Guffman said:
Re: xG

I said, "Jets always outscore xGoals (xGoals doesn’t actually factor shooter quality)". What was the point of your little exercise?

Re: Corsi

Corsi and #'s explain for the layman

I guess if we take a look at a more recent example (2019-2020), let's see how the top Corsi teams did:

Corsi Rank: 1 NHL Rank: 8
Corsi Rank: 2 NHL Rank: 24
Corsi Rank: 3 NHL Rank: 11
Corsi Rank: 4 NHL Rank: 28

shrug
Click to expand...

Point is this: The fact that Jets outscore xGoals just means for Jets there's a upward shift in the f(xG)=G curve, not that the curve is nonsensical. The curve is still there. Better xG still means better, and worse still means worse. It still matters in determining the weaknesses and areas to improve the team. The Jets are still typically at their best when they out Corsi and out xG opponents, whether looking game to game or season to season.


To your second part, you are still confusing in and out of sample correlation, and the point of Corsi:
Behind the Numbers: What Makes a Stat Good

Corsi is/was important for the 4 following reasons over history:
1) It represents a portion of the game that matters: shot volume.
There are 3 ways to improve offense
- get more chances than your opponent (shot volume, best measured by Corsi)
- make those chances the best you can (shot quality, best measured by xGoal per Corsi/Fenwick)
- make the best of those chances (finishing, best measured by goal per xGoal)
There are 3 ways to improve defense
- allow fewer chances for your opponent (shot volume, best measured by Corsi)
- make those chances the worst you can (shot quality, best measured by xGoal per Corsi/Fenwick)
- stop those chances (goaltending, best measured by goal per xGoal)
How good a team or player is does not end in just one of these six, but all six parts still matter. While no one part is everything, no one part is nothing.
2) It represents the portion of the game that is most consistent. A team that out shoots opponents in the first half of the year is more likely to still out shoot opponents the second half, then a team that stops more chances more than average. This is why Corsi is so predictive despite the other factors pushing teams.
Screen-Shot-2019-05-15-at-9-02-58-PM.png

3) The market at one point (it is better now but still not perfect) paid WAY more in $/win in finishing talent than shot volume talent. Recognizing and taking advantage market inefficiences is how one maximizes their chance at winning. Otherwise winning mostly comes down to $$$, timing, and luck.
 
Last edited:
  • Like
Reactions: John Agar and SLAYER
Guffman said:
To illustrate this, you should take the Corsi after half a season from... say 2018-2019, and tell us what the projected standings should be based on this. That would be a neat trick if it actually worked, wouldn't it? :)
Click to expand...

Okay, but I won't just do 2018-19 but instead look at all seasons:
cqwg8jlwcaig8na.png


Looking half way (or close to) at 40 games, we see red is higher than yellow. This means that Corsi in the first tells you more about how a team will perform in the back half more than goals in the first half does.

EDIT: Note the above xG is a different model than the ones used today that are more explanatory and less predictive.
 
Last edited:
  • Like
Reactions: Mathil8, John Agar, SLAYER and 2 others
surixon said:
And now the Jets are back above 50% and in thr middle of the pack in terms of XGF%.

Funny how single games can still really swing those results one way or the other.
Click to expand...

The last 3 Jets games have been really unusual with 2.5+ xGoal differential on each one of them. In the entirety of last season there was only 1 such Jets game that had such a wide margin. Curious to see where the dust settles.

In the past 3 years the Jets have overperformed their GF% over xGF% by ~4.5 points (it has been 2, 5 and 6 points in the 18-19, 19-20, 20-21 seasons respectively). If the baseline of xGF% goes from 45-46% to 49-50% this season, that would be huge as then a ~4 points over-performance in GF% would put them as a 53-54 GF% team which would easily put them in the top 3rd of the league.

Ofcourse this is assuming they continue to over-perform their xG% which I would say is likely but not guaranteed.
 
  • Like
Reactions: TommyKillian, hn777, AlaskaJet and 3 others
garret9 said:
Point is this: The fact that Jets outscore xGoals just means for Jets there's a upward shift in the f(xG)=G curve, not that the curve is nonsensical. The curve is still there. Better xG still means better, and worse still means worse. It still matters in determining the weaknesses and areas to improve the team. The Jets are still typically at their best when they out Corsi and out xG opponents, whether looking game to game or season to season.
Click to expand...

Yes, the more you shoot at favourable locations, the more goals you should expect to score. As Schiefele would say in his interviews, "Obviously...".

My original point is that the number spewed out for xG should not necessarily be viewed as the expected goals for a particular team because shooter quality matters.

If a team has xGF of 2.5 and xGA of 2.8 over a number of games, that doesn't necessarily mean the team is hopeless if they don't change the style of game. They might still be producing more goals than the opposition due to differentials in shooter and goaltending strength.

Let's take a look at this ranking summary:

9. Winnipeg Jets (6-3-2)

Previous rank: 19
Precarious: The Jets are collecting wins but are doing so in an unsustainable manner. Winnipeg ranks 24th in five-on-five expected goals at 47.07%, but the team is third with a 9.42 shooting percentage. That can't last forever. If the concerning trends persist while Connor Hellebuyck struggles uncharacteristically, regression could hit hard.

What part is unsustainable? Teams can still have a sub 50% xG differential and still outscore the opposition in the long-term. I think people misuse xG and Corsi to suggest that if you are not above 50%, that means a predictive doom and gloom regardless of your current positioning. You might qualify that and say that these are just a part of a myriad other stats someone should factor in... and this is true... but it is simply not done by those that just throw out these particular metrics.

Ultimately, all "Advanced Stats" are trying to do is provide components to the most important stat that supports success... The Guffman (i.e., goal differential): Guffman discovers new Advanced Stat
 
Guffman said:
Yes, the more you shoot at favourable locations, the more goals you should expect to score. As Schiefele would say in his interviews, "Obviously...".

My original point is that the number spewed out for xG should not necessarily be viewed as the expected goals for a particular team because shooter quality matters.

If a team has xGF of 2.5 and xGA of 2.8 over a number of games, that doesn't necessarily mean the team is hopeless if they don't change the style of game. They might still be producing more goals than the opposition due to differentials in shooter and goaltending strength.

Let's take a look at this ranking summary:

9. Winnipeg Jets (6-3-2)

Previous rank: 19
Precarious: The Jets are collecting wins but are doing so in an unsustainable manner. Winnipeg ranks 24th in five-on-five expected goals at 47.07%, but the team is third with a 9.42 shooting percentage. That can't last forever. If the concerning trends persist while Connor Hellebuyck struggles uncharacteristically, regression could hit hard.

What part is unsustainable? Teams can still have a sub 50% xG differential and still outscore the opposition in the long-term. I think people misuse xG and Corsi to suggest that if you are not above 50%, that means a predictive doom and gloom regardless of your current positioning. You might qualify that and say that these are just a part of a myriad other stats someone should factor in... and this is true... but it is simply not done by those that just throw out these particular metrics.

Ultimately, all "Advanced Stats" are trying to do is provide components to the most important stat that supports success... The Guffman (i.e., goal differential): Guffman discovers new Advanced Stat
Click to expand...

I think the failing of people who criticize advanced statistical analysis of hockey (or any sport really) is that they start from the point that proponents of advanced stats think that they are the end-all, be-all when it comes to hockey analysis. They are PREDICTIVE and ANALYTICAL models. Of course, sometimes teams will not perform exactly as a model suggests they might. I don't think people who make these models are suggesting that they will.

The stats help to explain the why X is happening and how it might continue. That's all. I'd also say it's a bold move to come into a thread discussing the Jets advanced statistics to state that they're "hogwash", to use another Scheifele term.
 
Guffman said:
Yes, the more you shoot at favourable locations, the more goals you should expect to score. As Schiefele would say in his interviews, "Obviously...".

My original point is that the number spewed out for xG should not necessarily be viewed as the expected goals for a particular team because shooter quality matters.

If a team has xGF of 2.5 and xGA of 2.8 over a number of games, that doesn't necessarily mean the team is hopeless if they don't change the style of game. They might still be producing more goals than the opposition due to differentials in shooter and goaltending strength.

Let's take a look at this ranking summary:

9. Winnipeg Jets (6-3-2)

Previous rank: 19
Precarious: The Jets are collecting wins but are doing so in an unsustainable manner. Winnipeg ranks 24th in five-on-five expected goals at 47.07%, but the team is third with a 9.42 shooting percentage. That can't last forever. If the concerning trends persist while Connor Hellebuyck struggles uncharacteristically, regression could hit hard.

What part is unsustainable? Teams can still have a sub 50% xG differential and still outscore the opposition in the long-term. I think people misuse xG and Corsi to suggest that if you are not above 50%, that means a predictive doom and gloom regardless of your current positioning. You might qualify that and say that these are just a part of a myriad other stats someone should factor in... and this is true... but it is simply not done by those that just throw out these particular metrics.

Ultimately, all "Advanced Stats" are trying to do is provide components to the most important stat that supports success... The Guffman (i.e., goal differential): Guffman discovers new Advanced Stat
Click to expand...
Ah yes predicting the past, that's what everyone is trying to do.
 
All xgf/xga is create a baseline(for individuals or teams) based on shot data in order to assign values to shot attempts. Difference between actual and expected can be shooter talent (or lack of) or good goaltending (or poor). A team oriented xgf probably exists but hard to compare team to team or across different seasons. Xgf/xga vs actual gf /ga probably nets similar results or conclusions
 
Last edited:
  • Like
Reactions: Jack7222
Interesting comment by Maurice, he said that the Jets are 2nd best in the league when it comes to the opponent having puck possession in the Jets zone.



It is probably based on the NHL puck tracking or Sportslogiq data. That sounds true enough, other than one of the Ducks games and parts of the 3rd period of the Wild game, I don't remember the Jets being hemmed in their own zone for shifts on end. Something that use to happen all the time last 2-3 seasons. That is not to say they don't give up quality chances (the Islanders/Stars didn't have the puck much in those games but still generated a few grade A+ chances in those games) but seems like a step in the right direction compared to last year when the Jets would go an entire forward rotation without getting out of their own end.
 
  • Like
Reactions: hn777, Ducky10, AlaskaJet and 1 other person
JetsFan815 said:
Interesting comment by Maurice, he said that the Jets are 2nd best in the league when it comes to the opponent having puck possession in the Jets zone.



It is probably based on the NHL puck tracking or Sportslogiq data. That sounds true enough, other than one of the Ducks games and parts of the 3rd period of the Wild game, I don't remember the Jets being hemmed in their own zone for shifts on end. Something that use to happen all the time last 2-3 seasons. That is not to say they don't give up quality chances (the Islanders/Stars didn't have the puck much in those games but still generated a few grade A+ chances in those games) but seems like a step in the right direction compared to last year when the Jets would go an entire forward rotation without getting out of their own end.
Click to expand...


Also seems like our zone exit microstats are very good as well. That is the foundation we need to spearhead our offensive game.
 
  • Like
Reactions: hn777
JetsFan815 said:
Interesting comment by Maurice, he said that the Jets are 2nd best in the league when it comes to the opponent having puck possession in the Jets zone.



It is probably based on the NHL puck tracking or Sportslogiq data. That sounds true enough, other than one of the Ducks games and parts of the 3rd period of the Wild game, I don't remember the Jets being hemmed in their own zone for shifts on end. Something that use to happen all the time last 2-3 seasons. That is not to say they don't give up quality chances (the Islanders/Stars didn't have the puck much in those games but still generated a few grade A+ chances in those games) but seems like a step in the right direction compared to last year when the Jets would go an entire forward rotation without getting out of their own end.
Click to expand...


look at the 1st game of the year. the Jets out-corsi'd the Ducks 41-21. but i found a lot of the shot attempts low danger, minimal traffic, goalie movement or rebound opportunties.

so 41 to 21 in corsi
1.53 to 1.07 in xGF

per attempt; the Ducks actually scored better in quality, which kind of matched with how viewed that particular game.

islanders game
Jets won corsi 58-50
lost xgf 1.59-3.09.

islnaders were far superior in quality per attempt than the Jets. they're probably content with the Jets taking low-danger shot attempts w/ low probability of going in. especially with a goalie like Sorokin.
 
  • Like
Reactions: hn777
tbcwpg said:
I think the failing of people who criticize advanced statistical analysis of hockey (or any sport really) is that they start from the point that proponents of advanced stats think that they are the end-all, be-all when it comes to hockey analysis. They are PREDICTIVE and ANALYTICAL models. Of course, sometimes teams will not perform exactly as a model suggests they might. I don't think people who make these models are suggesting that they will.

The stats help to explain the why X is happening and how it might continue. That's all. I'd also say it's a bold move to come into a thread discussing the Jets advanced statistics to state that they're "hogwash", to use another Scheifele term.
Click to expand...

All models are wrong, some are useful.

Shot quality and quantity are components of winning, as is shot quality. xG is important as it is important to know that the Jets failings and weaknesses are in a particular area. Part of what keeps them from being good the past few seasons is because they fail at the shot quantity and quality despite being above average at goaltending and finishing.
 
Romang67 said:
It's simply more fun hockey to watch than the past couple of years (arguably dating back to spring 2019).

If I don't ever have to see CSW again, it will be too soon.
Click to expand...

Agreed, our play has been much more enjoyable this year. We need to get Blake and Mark back playing fast and direct hockey and putting Conner with them right now likely goes against that. I am happy Nik is on their line now and should help push the pace for the other two.
 
  • Like
Reactions: MardyBum and Romang67
surixon said:
Agreed, our play has been much more enjoyable this year. We need to get Blake and Mark back playing fast and direct hockey and putting Conner with them right now likely goes against that. I am happy Nik is on their line now and should help push the pace for the other two.
Click to expand...

Yup, and Copp/Lowry instantly went back to just out hustling lines.
 
  • Like
Reactions: John Agar, surixon and Romang67
I think my favourite thing from the Sharks game was how they kept the pedal down, forechecked hard and covered back through the neutral zone. Good gap too. Last two games have been very 2017-18 reminiscent, been a few others as well.
 
  • Like
Reactions: GNP, Mud Turtle, John Agar and 2 others
Ducky10 said:
I think my favourite thing from the Sharks game was how they kept the pedal down, forechecked hard and covered back through the neutral zone. Good gap too. Last two games have been very 2017-18 reminiscent, been a few others as well.
Click to expand...

Yup they are certainly trending that way. There are elements of 14-15 in our game as well. This to me is what Jet hockey is, what we saw the prior two years was the anthisis of it.

Going to games is fun again as the team is good and entertaining to watch.
 
  • Like
Reactions: JetsFan815, John Agar, Ducky10 and 1 other person
Now that they have played a few games against good teams

Team Season Totals - Natural Stat Trick

7th league wide in Corsi and 11th in xGF%, comfortably over 50% in both these metrics. I like our chances if we can stay over 50% all season given the goaltending and shooting talent. Much better than being in the low-mid 40s in these metrics like the last 2-3 seasons.
 
  • Like
Reactions: Jack7222, Mathil8, Ducky10 and 1 other person
That PK :laugh:


Offense - good with chances and attempts, but finish is lacking.
Defense - not many attempts, but given up high quality per attempt. good goaltending.
 
  • Like
Reactions: JetsFan815
You must log in or register to reply here.

Latest posts

Ad

Upcoming events

Ad

Top Bottom