Rumor: Rumors & Proposals Thread | With Klingberg in the Mix Who Are Our 7D After the Deadline?

[guymez]

Even the better xG models are rudimentary and alienate the majority of the real world context for any individual data point. If any of the models accounted in any way for unpredictable behavior, i.e. the nachos on the ice distracting the goalie to allow a floater from the blue line, you would have something more valuable to this discussion.

The entire point has been that regardless of how well things go for this Oilers team full of chaos D-men, they are still prone to the occassional peewee level 5 alarm firedrill.

McDavid can rack up HDSC all night long (and often does) with no goals to show for it. The majority of the HDSC we allow are easily handled by the goaltender. Then we have a handful of chaos plays that are literally worse than a breakaway where guys are behaving in completely unpredictable fashion and allowing attackers carte blanche to do whatever they want.

Appreciate the post...very well articulated.

Regarding the highlighted....opposing coaches have to know this and so if they can generate a style of play (a system) that focuses on shutting down McDavid and Draisaitl they can take adavantage of those peewee level breakdowns and give themselves a very strong chance of winning the game.
This is especially true because the Oilers secondary scoring is so subpar.

So while I do think its important to upgrade the goalie tandem that change is not going to solve these problems and IMO these problems need to be solved if this team is going to win the SC.
I dont care who is in net. A top 5 goalie may be able to mask these deficiencies over the short term but it wont be a long term solution.
It also wont be a solution in the later rounds when the quality of teams (and quality of coaching) is distilled down to 4 very high quality teams.

 

[Tobias Kahun]

So where does this account for contested and uncontested shots?

More nuance than that, not as black and white

 

[stewy04]

There was also a time where Talbot let in the first shot of the game.... all the time.

I think that was a few of our goalies! ...ughh

 

[Fourier]

Even the better xG models are rudimentary and alienate the majority of the real world context for any individual data point. If any of the models accounted in any way for unpredictable behavior, i.e. the nachos on the ice distracting the goalie to allow a floater from the blue line, you would have something more valuable to this discussion.

The entire point has been that regardless of how well things go for this Oilers team full of chaos D-men, they are still prone to the occassional peewee level 5 alarm firedrill.

McDavid can rack up HDSC all night long (and often does) with no goals to show for it. The majority of the HDSC we allow are easily handled by the goaltender. Then we have a handful of chaos plays that are literally worse than a breakaway where guys are behaving in completely unpredictable fashion and allowing attackers carte blanche to do whatever they want.

Proprietary xG models include data from fairly sophisticated puck tracking. As it turns out adding in these additional layers does not significantly change the outcomes player to player.

If you want to look at a single game I agree that chaos could explain anomalies. But this data represents tens of thousands thousands of repeated events. I would be the last person to claim these models were perfect. In contrast your claims are based on a few observations that are by their nature subject to extreme observational bias. And here I am not suggesting that you individually bias. We all suffer from the same issue. It is indeed possible that reality is different from what the data says. But in this case, I think that would be very surprising. It would require the Oilers to be highly exceptional in terms of their inability to perform their basic duties. Were that the case, I suspect that this would show up in far greater ways.

 

[Devilmademedoit]

How did I do?

View attachment 972385

View attachment 972384

If Kane is back he takes that top 6 spot, otherwise see if Kuzmenko works in his spot.

Why the hell would you want Kuzmenko? HE has been run off two team now!

 

[alphahelix]

Proprietary xG models include data from fairly sophisticated puck tracking. As it turns out adding in these additional layers does not significantly change the outcomes player to player.

If you want to look at a single game I agree that chaos could explain anomalies. But this data represents tens of thousands thousands of repeated events. I would be the last person to claim these models were perfect. In contrast your claims are based on a few observations that are by their nature subject to extreme observational bias. And here I am not suggesting that you individually bias. We all suffer from the same issue. It is indeed possible that reality is different from what the data says. But in this case, I think that would be very surprising. It would require the Oilers to be highly exceptional in terms of their inability to perform their basic duties. Were that the case, I suspect that this would show up in far greater ways.

Why dont you pull up a chart for Troy Stecher and see where on ice GA occur for him as an individual player? Then pull up the average defender, and we'll see if he allows a higher % of GA at the slot against elites per 60.

We can run that chart for a couple of "eye test" culprits, and we can design a few more tests to identify potential oddities with the data that reflect what the eye test is demonstrating to many observers here.

Or we can agree to disagree, I guess--- but the charts you've pulled so far do not dispel any of the things the eye test people are saying.

Im pretty sure if you asked the coaches off the record if they think this D corps is reliable and steady against elites in tight playoff games, theyd ask for help.

 

[brentashton]

Bouchard played 27 mins and 4 seconds in that game. How much more do you expect him to play?

He didn’t have to play anymore minutes, nor was I suggesting he do so. Read my post.

I said that perhaps he could have stepped in and played some PK minutes in Eckholm’s temp absence. The offset 5 on 5 could be shared by the lower depth players perhaps. GYB was suggesting he could have stepped up for some of the tougher minutes in a teammate’s absence. I agree.

 

[The Nuge]

I think that was a few of our goalies! ...ughh

It’s every goalie. On average there’s like a 9% chance of any shot going in, so every 4 games, there’s theoretically bound to be a goal within the first few shots. HFOil just has an obsession with it as if it’s an anomaly instead of the norm

 

[McDoused]

Why the hell would you want Kuzmenko? HE has been run off two team now!

Hes free. I used all the assets on a goalie and defender but still had cap left.

 

[McTonyBrar]

This is from the Athletic. Who wrote this?

 

[Burnt Biscuits]

How did I do?

View attachment 972385

View attachment 972384

If Kane is back he takes that top 6 spot, otherwise see if Kuzmenko works in his spot.

The Byram trade I don't think BUF would bite, they rumors are they are looking more for a hockey trade that helps them more near term than more picks.

Skinner trade is fine, I don't think TOR would want Skinner unless they find a new home for Nick Robertson.

Kuzmenko I'm not interested in at all especially with that cap-hit and PHI just acquired him for Michkov. I also don't think Knob would like Kuzmenko's game.

Vejmelka sounds great, but that's a cheaper acquisition cost then I would guess.

Lastly I don't think Frederic is anywhere close to being worth Akey, we wouldn't be the first team to overpay for toughness, but that's a trade I would not make.

 

[Fourier]

Why dont you pull up a chart for Troy Stecher and see where on ice GA occur for him as an individual player? Then pull up the average defender, and we'll see if he allows a higher % of GA at the slot against elites per 60.

We can run that chart for a couple of "eye test" culprits, and we can design a few more tests to identify potential oddities with the data that reflect what the eye test is demonstrating to many observers here.

Or we can agree to disagree, I guess--- but the charts you've pulled so far do not dispel any of the things the eye test people are saying.

Im pretty sure if you asked the coaches off the record if they think this D corps is reliable and steady against elites in tight playoff games, theyd ask for help.

My comments were team focused. I actually agree with you on Stecher by the way. He is not great in front of the net. I'd be happy to provide those charts but they are behind a paywall.

As to the eye test they are somewhat ironically analogous to the plane diagram you posted though for the exact opposite reason. Both exclude data that is highly relevant. What the eye test too often focuses on is what went wrong without including data on what went right. The latter is often far more illuminating but is seldom noted.

Again, my post is not to suggest that anyone's "eye test" has no value, specifically yours. I think it is an important part of the equation. It just concerns me when the eye test and the data differ markedly.

 

[McTonyBrar]

Bob Stauffer said that, if he was GM, he’d bring in one more top six forward for the Oilers.

Friedman believes Oilers are looking at a lot of things because they want to win and points out that their 1st round pick is very valuable

 

[russ99]

I’d love to see a metric for undefended chances against. This is the culprit with our defense and how Skinner is getting cooked, not xG, IMO

 

[Devilmademedoit]

Hes free. I used all the assets on a goalie and defender but still had cap left.

You clearly have not seen him play. He is toast. Thought the idea was to win?

 

[McDoused]

My comments were team focused. I actually agree with you on Stecher by the way. He is not great in front of the net. I'd be happy to provide those charts but they are behind a paywall.

As to the eye test they are somewhat ironically analogous to the plane diagram you posted though for the exact opposite reason. Both exclude data that is highly relevant. What the eye test too focuses focuses on is what went wrong without including data on what went right. The latter is often far more illuminating but is seldom noted.

Again, my post is not to suggest that anyone's "eye test" has no value, specifically yours. I think it is an important part of the equation. It just concerns me when the eye test and the data differ markedly.

Agree with this. Analytics and data are a tool we can use, especially when it doesnt match the "eye test". I find the most common problem with people's eye tests is two fold.

One, we tend to have a recency bias. Part of this ks because our brains cant accurately recall what happened months ago. The only things we might remember are the extreme positives (goals and assists) and extreme negatives (blunders that resulted in a goal).

Two, this is a results orientated business. We always hear coaches talk about the process, playing the right way and doing the little things. These all result in better expected numbers even if the results aren't there. Heck, look at Vancouver last year or Washington this year. At some point those teams PDO were bound to see regression. I'm not convinced most people's eye test and memory are trained to see expected results over actual results.

Data driven analysis allows us to remove inherit biases and calculate (model) expected results. It's never going to be perfect but the more work and data we put into the models the better the model becomes.

 

[alphahelix]

I’d love to see a metric for undefended chances against. This is the culprit with our defense and how Skinner is getting cooked, not xG, IMO

this would be awesome to see, not sure how exactly you'd identify it-- nobody within sticklength of the shooter?

 

[alphahelix]

All I care about is the numbers against the top ten teams (at the time they played the teams)

I can see numbers all over to start the year because it takes time for teams to learn systems etc.

I have to look it up, but more so the last month or month and a half

That's what matters. You will play those teams in the playoffs. I never hear any media talk about this

All they talk about is how his numbers are decent for him....

That means nothing once we play the top dogs

I know Connor Haley on Gregor show yesterday thinks he's top 15 lol. Connor must be drunk on the show

I feel exactly the same way except I am more cautious about whether these GA and Sv% against the actual teams that matter is more on the fault of the goaltending or on the defence. I feel that it is ‘obviously’ the combination of both together, and you can save on budget goaltending if you have excellent defenders - but you cant save on both and compete with the best for a cup. I’d love to see some of the numbers under the hood that might attempt to peel the issue apart.

 

[sepHF]

*Wrong thread

 

[Burnt Biscuits]

He didn’t have to play anymore minutes, nor was I suggesting he do so. Read my post.

I said that perhaps he could have stepped in and played some PK minutes in Eckholm’s temp absence. The offset 5 on 5 could be shared by the lower depth players perhaps. GYB was suggesting he could have stepped up for some of the tougher minutes in a teammate’s absence. I agree.

The deployment and coaches mindset I think is rather obvious, we were already down 1-0 prior to that PPG being scored, he obviously was hoping they would kill the penalty off and then roll out our loaded up best offensive unit to tie it up, which is always McDavid + Drai + Bouchard; we've been doing that same strategy going back to Woodcroft

You said Bouchard not being out there was an "indictment of Bouchard's lack of growth in itself"

When Nuge missed a game to illness, did you comment about McDavid not filling in for his PK minutes and say it is an "indictment of McDavid's lack of growth in itself".

Yeah, I think not.

Come on Bouchard is held to an elevated level of scrutiny, that is not entirely reasonable. The bottom line is Bouchard played 27+ mins and he was the best player on either team and the coaches used him in a way to take advantage of the strengths (in a game where we were chasing all night and badly needed his offense).

 

[brentashton]

The deployment and coaches mindset I think is rather obvious, we were already down 1-0 prior to that PPG being scored, he obviously was hoping they would kill the penalty off and then roll out our loaded up best offensive unit to tie it up, which is always McDavid + Drai + Bouchard; we've been doing that same strategy going back to Woodcroft

You said Bouchard not being out there was an "indictment of Bouchard's lack of growth in itself"

When Nuge missed a game to illness, did you comment about McDavid not filling in for his PK minutes and say it is an "indictment of McDavid's lack of growth in itself".

Yeah, I think not.

Come on Bouchard is held to an elevated level of scrutiny, that is not entirely reasonable. The bottom line is Bouchard played 27+ mins and he was the best player on either team and the coaches used him in a way to take advantage of the strengths (in a game where we were chasing all night and badly needed his offense).

lol. You’re twisting things into pretzels. I’m not even talking about Nuge or McDavid. Not sure how that entered the discussion.

Enjoy the game Burnt.

 

[bellagiobob]

Rantanen might have to settle for league minimum next year with how he’s playing in Carolina.

 

[North]

Rantanen might have to settle for league minimum next year with how he’s playing in Carolina.

I don’t think the team as a whole is looking all that hot.

 

[Jarvi]

Has anybody heard a wild rumor about what really went on behind the scenes with Petey and Jt Miller? I am not going to post it here, but heard it from some insiders

Can you send me a dm? I'm very curious

 

[Tarus]

Rantanen might have to settle for league minimum next year with how he’s playing in Carolina.

The whole team is sick with the flu apparently, might be a factor.