Rumor: Rumors & Proposals Thread: Movie Poster Edition VII- "Deadline"

[Jimmi McJenkins]

Did you misread this and think Makar is out indefinitely? Because Josh Manson being out indefinitely does not make an entire conference wide open

I did miss read, but avs are banged up, and there for the picking

 

[5 Mins 4 Ftg]

I don't think this is true. Some nerds actually understand the things they are studying. But even if they don't have a firm grasp of what you are describing that does not mean they can't be useful. The job of the analytic department can be to provide data in a form that is useful to the hockey people and help them to interpret the results.

The problem is actually the opposite of what you describe. It is people who don't understand what the numbers are actually saying and the limitations in that regard that tends to be the biggest issue.

The ones I’ve interacted with haven’t a clue about the game itself. Some admit they don’t even watch the games but look at the data after the game.

I’m not saying advanced stats can’t be useful but they are often presented as absolutes and present flawed conclusions often based on subjective data.

Case in point - WAR charts and player cards.

It’s also an ego trip for many of the Twitter nerds.

 

[Trafalgar Sadge Law]

It was stated ad nauseoum that PJ made McDavid better. Statements like this are commonplace in the nerd community as they try to legitimize their WAR charts and player cards.

Even today some of the nerds on Twitter were saying what a disaster trade for the Oilers.

I’m not saying advanced stats are all garbage and context is king but I am tired of getting preached to from the lofty perch from up on high the nerds claim as absolutes their unassailable mathematical high ground.

It is getting ridiculous with these player cards and WAR charts.

Puljujarvi was put on COVID protocol after the December 16, 2021 game. Up until that point, Connor McDavid had 16 goals and 31 assists in 28 games. During the bubble season, McDavid's PPG also went up after being moved with Puljujarvi in the second half of the season that saw him pop off and hit 100 in 53. Seems like McDavid was doing great with the version of Puljujarvi that was playing well and the analytics reflected that. This season McDavid's stats have been awful with Puljujarvi who is both by eye test and analytically playing far far worse. The numbers here are capturing the change in gameplay and results pretty damn well.

Now the context here is that the players Puljujarvi was replacing on McDavid's wing back in 2020-21 were Zack Kassian and Alex Chiasson. Analytics said that Puljujarvi was making McDavid better COMPARED TO Zack Kassian and Alex Chiasson. I think we can safely say that's the correct conclusion and paints exactly what happened. That's just an observable fact you can quantify with McDavid's production and also with the eye test that he did better with Puljujarvi than he did with Kass/Chiasson, just like how the same on/off stats of McDavid with Hyman this year indicate Hyman is making McDavid better COMPARED TO Puljujarvi and Yamamoto.

If anything the "stats nerds" you seem to hate so much are people who don't know how to use the numbers properly themselves (not claiming I'm an expert either btw). Using the people misusing the stats to generalize/dismiss analytics is not much better than people drawing dumb black and white conclusions with context-less numbers like "Nichushkin is better than Draisaitl".

EDIT: You're probs also climbing way too deep into the Twitter rabbit holes lmfao. The only analytics people I see that are this extreme are the Evolving-Wild crowd, most are pretty reasonable.

 

[guymez]

While I agree context should always be used with analytics and they're just a tool to help contextualize what you're seeing on ice, but when have analytics ever said that Caleb Jones is carrying Adam Larsson? If anything Larsson's analytics on the Oilers were favourable compared to his actual on ice play and results especially from 2017-20. We don't need to make up narratives that never happened just to make analytics seem worse than they are.

100%.

This ongoing need to demonize analytics is puzzling for sure..
Someone posts an interpretation of analytical data that lacks context and a number of people hold that up as an example that analytics are useless.

Its just a tool...like any tool its ultimately only as useful as the person utilizing it. When used improperly its not very helpful.

Its like someone using a shovel improperly and a group of people gathering around and proudly declaring that its proof that the shovel is useless.

Ridiculous.

 

[5 Mins 4 Ftg]

Puljujarvi was put on COVID protocol after the December 16, 2021 game. Up until that point, Connor McDavid had 16 goals and 31 assists in 28 games. During the bubble season, McDavid's PPG also went up after being moved with Puljujarvi in the second half of the season that saw him pop off and hit 100 in 53. Seems like McDavid was doing great with the version of Puljujarvi that was playing well and the analytics reflected that. This season McDavid's stats have been awful with Puljujarvi who is both by eye test and analytically playing far far worse. The numbers here are capturing the change in gameplay and results pretty damn well.

Now the context here is that the players Puljujarvi was replacing on McDavid's wing back in 2020-21 were Zack Kassian and Alex Chiasson. Analytics said that Puljujarvi was making McDavid better COMPARED TO Zack Kassian and Alex Chiasson. I think we can safely say that's the correct conclusion and paints exactly what happened. That's just an observable fact you can quantify with McDavid's production and also with the eye test that he did better with Puljujarvi than he did with Kass/Chiasson, just like how the same on/off stats of McDavid with Hyman this year indicate Hyman is making McDavid better COMPARED TO Puljujarvi and Yamamoto.

If anything the "stats nerds" you seem to hate so much are people who don't know how to use the numbers properly themselves (not claiming I'm an expert either btw). Using the people misusing the stats to generalize/dismiss analytics is not much better than people drawing dumb black and white conclusions with context-less numbers like "Nichushkin is better than Draisaitl".

I never read or heard once that Pj was better COMPARED to others on his right wing. It was “PJ makes 97 better” - over and over and over again.

We hear many a tall tale about how great certain analytic darlings are who never amount to a hill of beans. Often backed up with WAR charts and Player Cards.

I don’t think advanced stats are all useless. But I often find the conclusions as absolutes many bring to the table are near laughable.

Expected Goals and Saves seem to have become more important than real goals and saves - which NHL players and coaches laugh over - twitterlytics I believe one them called them. That’s where I first heard the name.

Anyway we probably agree more than disagree on analytics. I can get pretty abrasive in my disdain for how the game itself is often misunderstood in the quest for sanitized baseball like stats as absolutes on every given thing.

Cheers.

 

[guymez]

Puljujarvi was put on COVID protocol after the December 16, 2021 game. Up until that point, Connor McDavid had 16 goals and 31 assists in 28 games. During the bubble season, McDavid's PPG also went up after being moved with Puljujarvi in the second half of the season that saw him pop off and hit 100 in 53. Seems like McDavid was doing great with the version of Puljujarvi that was playing well and the analytics reflected that. This season McDavid's stats have been awful with Puljujarvi who is both by eye test and analytically playing far far worse. The numbers here are capturing the change in gameplay and results pretty damn well.

Now the context here is that the players Puljujarvi was replacing on McDavid's wing back in 2020-21 were Zack Kassian and Alex Chiasson. Analytics said that Puljujarvi was making McDavid better COMPARED TO Zack Kassian and Alex Chiasson. I think we can safely say that's the correct conclusion and paints exactly what happened. That's just an observable fact you can quantify with McDavid's production and also with the eye test that he did better with Puljujarvi than he did with Kass/Chiasson, just like how the same on/off stats of McDavid with Hyman this year indicate Hyman is making McDavid better COMPARED TO Puljujarvi and Yamamoto.

If anything the "stats nerds" you seem to hate so much are people who don't know how to use the numbers properly themselves (not claiming I'm an expert either btw). Using the people misusing the stats to generalize/dismiss analytics is not much better than people drawing dumb black and white conclusions with context-less numbers like "Nichushkin is better than Draisaitl".

EDIT: You're probs also climbing way too deep into the Twitter rabbit holes lmfao. The only analytics people I see that are this extreme are the Evolving-Wild crowd, most are pretty reasonable.

 

[Fourier]

The ones I’ve interacted with haven’t a clue about the game itself. Some admit they don’t even watch the games but look at the data after the game.

I’m not saying advanced stats can’t be useful but they are often presented as absolutes and present flawed conclusions often based on subjective data.

Case in point - WAR charts and player cards.

It’s also an ego trip for many of the Twitter nerds.

Most of those guys have limited bona fides on the nerd side. I know a lot of "nerds" who are perfectly capable of understanding what they speak of.

Truthfully, I don't pay much attention to all of these charts, nor do I really believe in WAR having much value in the hockey world. Hockey is just to much of a unit/team sport for some of the baseball style sabremetrics to translate. But some of the newer analytics guys who produce these things do actually have reasonable or even very good grasp of both the mathematics and of the game. The problem tends to come from their disciples who preach the stuff as if it was gospel without themselves having a decent understanding of what they say. A guy like Micah McCurdy for example knows his stuff on the math side and presents it in a way that I generally find quite illuminating.

 

[5 Mins 4 Ftg]

100%.

This ongoing need to demonize analytics is puzzling for sure..
Someone posts an interpretation of analytical data that lacks context and a number of people hold that up as an example that analytics are useless.

Its just a tool...like any tool its ultimately only as useful as the person utilizing it. When used improperly its not very helpful.

Its like someone using a shovel improperly and a group of people gathering around and proudly declaring that its proof that the shovel is useless.

Ridiculous.

Not demonizing anything. Advanced stats have a place - if you’re using the right stats.

Merely saying the ego nerds on Twitter like J Fresh, the Evolving Idiots and their WAR charts and player cards are ridiculous.

 

[alphahelix]

Did you misread this and think Makar is out indefinitely? Because Josh Manson being out indefinitely does not make an entire conference wide open

I would argue that Manson is still a critical player for them much as Ekholm will prove to be for us. They need that physicality/PK/stability element to hold onto their leads and close out games.

Obviously that guy‘s statement is predicated on the assumption that the Avalanche are the only team that are favoured to prevail among the crowd when healthy, which may well be accurate.

 

[5 Mins 4 Ftg]

Bill Gates implanted me with an XBox chip through the vaccine. I can’t help it.

 

[Macblender]

Well, it's never too early to start talking what Holland's summer plan needs to be.

It's going to have to be cap creative, but thankfully we have a lot of key players locked in. Fingers crossed the cap goes up by more than 1m.

Bouch and McLeod. Sadly he'll have to try and squeeze them on 2 year bridge deals, unless either are willing to sign a stupidly kind long-term deal.

Kostin, I love his play, but he's going to have to come in at another show-me price.

The key players to try and upgrade on are Soup and Ceci. Campbell I don't see as being realistic unless SJ is taking him in a mass overpay for Karlsson. All else failing, trying to find a top 6 RW would be great.

I hope we take better advantage of guys needing 1 year show-me deals or vets willing to sign league min to chase a cup. If we can upgrade some key positions, we will badly need these value deals.

Honestly I think at this point we should all just be hoping for an insanely good playoffs where large markets with actual viewership go very far so we are more likely to see the 85-86.5M cap than the 83.5M that is more likely right now

 

[guymez]

Bill Gates implanted me with an XBox chip through the vaccine. I can’t help it.

Ah yes...Bill Gates. Such an icon of truth and benevolence.
Just trust what he says....he's an expert after all.

 

[guymez]

Not demonizing anything. Advanced stats have a place - if you’re using the right stats.

Merely saying the ego nerds on Twitter like J Fresh, the Evolving Idiots and their WAR charts and player cards are ridiculous.

We absolutely agree...there are people that misrepresent the data and how it should be utilized.

Sincere question...if you know that already then why are you continuing to read their posts?

 

[belair]

Often subjective data combining subjective cherry picking.

But get a nerd to an actual live game and ask them to explain what the 2 teams forechecking or defensive systems being employed in real time are and the players employment and effectiveness are within said systems and it’s silence.

There's always the Kris Russell argument, too. There are players that teams employ--usually at a lower salary--who possess certain defensive skillsets. Statistically they're consistently bottom of the barrel in terms of things like shot and goal metrics, but the primary focus of their game is to limit the amount of damage that gets done when the weaker parts of a roster is on the ice. And in the cap era, it's inevitable for those parts of a roster to exist on most teams.

Are those players bad defensively?

 

[guymez]

There's always the Kris Russell argument, too. There are players that teams employ--usually at a lower salary--who possess certain defensive skillsets. Statistically they're consistently bottom of the barrel in terms of things like shot and goal metrics, but the primary focus of their game is to limit the amount of damage that gets done when the weaker parts of a roster is on the ice. And in the cap era, it's inevitable for those parts of a roster to exist on most teams.

Are those players bad defensively?

Yes?

Seriously.
It reminds me a little of Dallas Eakins suggesting (I am paraphrasing here) that a team that hits a lot is evidence of a team that doesnt have the puck very much.

 

[5 Mins 4 Ftg]

We absolutely agree...there are people that misrepresent the data and how it should be utilized.

Sincere question...if you know that already then why are you continuing to read their posts?

I don’t — their crap bleeds into this forum and I usually just ignore it. Somehow got sucked back into the conversation. Anyway - all good!

 

[guymez]

I don’t — their crap bleeds into this forum and I usually just ignore it. Somehow got sucked back into the conversation. Anyway - all good!

I hear you. I just ignore that nonsense.

There are some quality analytics posters out there though.
It might be worth your time.

 

[Trafalgar Sadge Law]

I never read or heard once that Pj was better COMPARED to others on his right wing. It was “PJ makes 97 better” - over and over and over again.

We hear many a tall tale about how great certain analytic darlings are who never amount to a hill of beans. Often backed up with WAR charts and Player Cards.

I don’t think advanced stats are all useless. But I often find the conclusions as absolutes many bring to the table are near laughable.

Expected Goals and Saves seem to have become more important than real goals and saves - which NHL players and coaches laugh over - twitterlytics I believe one them called them. That’s where I first heard the name.

Anyway we probably agree more than disagree on analytics. I can get pretty abrasive in my disdain for how the game itself is often misunderstood in the quest for sanitized baseball like stats as absolutes on every given thing.

Cheers.

Fair enough, I agree with the general notion that analytics have their place in the game but need proper utilization and are not the end all be all. There's a reason why so many well run franchises like Tampa, Colorado, Vegas, Carolina, and Toronto (yes Toronto, specifically in regular season ) heavily use analytics, but they also make sure to back it up with in person scouting/film review to obtain the context behind the numbers. The only question is whether you believe analytics should be used to paint 10%, 30%, 50%, 70% etc of the picture and I suspect I have a higher number than you, but I think it's fair to say it's not 0% and certainly not 100%.

Also I still think you're diving into the hallowed depths of Twitter that you're better off avoiding. The Evolving-Wild crowd in particular I personally find insufferable even as a "stats nerd" myself. There are plenty of completely reasonable analysts/communities out there that don't spew bullshit like Nichushkin>Draisaitl.

On a somewhat related note, even something as well refined as baseball sabremetrics is constantly evolving. One big new things in their industry is swing path tracking which studies collision/launch angles, swing speed, "shape" of swing etc relative to pitch velocity/spin/movement/location etc which only started the past 2 or 3 years. Information is power in the current world regardless of industry and being the first to figure out something gives you a significant advantage over your competitors. Who knows when the next new hockey stat makes xGF% and HDCF% irrelevant just like how xGF% and HDCF% have made Corsi largely obsolete within the past half decade. I'd just rather the Oilers be the team to figure it out first and not someone like the Avs.

 

[Fourier]

I am going to try to show why in certain cases even a bad stat can tell you things that the eye test might miss.

One of the most obviously abused stats is the simple +/-. I think most people realize that as a raw stat it is ridiculous to use it without context to compare players on different teams for example. But that gets converted into "+/- is always useless".

What I am going to say will be controversial but I'll say it anyway. Nurse makes a lot of highly visible mistakes. And some of these result in goals that stick in people's minds. But for the last three years he has consistently been one of if not the top +/- guy on the Oilers. In fact he is 9th overall in +/- amongst NHL defensemen despite playing more minutes against the oppositions best players than almost any other defender and without a high-end partner to absorb the load. So what can one deduce from this "useless stat". I claim that even with its many flaws +/- still shows that despite his very visible mistakes when Nurse is on the ice things tend to go very well for the Oilers. The sample size is large enough that even this very basic stat can provide compelling evidence suggest that the common narrative that Nurse is a defensive liability is just false.

Now to be clear, I am not claiming that +/- can prove that any player is good defensively. What I am saying is that a it is unlikely that a player in Nurse's role could achieve such numbers if he was as bad defensively as some suggest.

 

[Fourier]

Yes?

Seriously.
It reminds me a little of Dallas Eakins suggesting (I am paraphrasing here) that a team that hits a lot is evidence of a team that doesnt have the puck very much.

He is right though. Hits do tend to correlate with lack of possession. The problem is that this tends to be interpreted that throwing hits is bad. The two statements are not equivalent. And that is one of the most common errors we see on these boards. Linking two statements incorrectly.

 

[guymez]

He is right though. Hits do tend to correlate with lack of possession. The problem is that this tends to be interpreted that throwing hits is bad. The two statements are not equivalent. And that is one of the most common errors we see on these boards. Linking two statements incorrectly.

What comes to mind for me is that puck retrieval is a very important part of the game.
So if a team is built in a way that allows then to uses body checks as a means of retrieving the puck then that isnt necessarily suggesting that puck possession is an issue for that team.
In isolation it's not really providing any useful information about puck possession.

 

[Fourier]

What comes to mind for me is that puck retrieval is a very important part of the game.
So if a team is built in a way that allows then to uses body checks as a means of retrieving the puck then that isnt necessarily suggesting that puck possession is an issue for that team.
In isolation it's not really providing any useful information about puck possession.

That's not the point though. The correlation follows because you don't throw hits if you already have the puck. And as such the more hits a team has the more likely it is that they don't have the puck. What you are talking about is how hits can help you regain the puck. As I mentioned these are different but related things.

 

[Trafalgar Sadge Law]

The eye test can be and often is rooted in biases due to people looking for things they want to see and ignoring things that don't fit their preconceived notions. One major example of this was this past playoffs where a significant portion of the "eye test" crowd on these boards was acting like Darnell Nurse was out here committing war crimes whereas Duncan Keith was this stable veteran carrying us to the conference finals. Nurse has been a pseudo-whipping boy on these boards since his rookie season and people will pick out every mistake he makes despite him having the most ridiculously overworked deployment I've ever seen from a hockey player and trying to ignore the player's strengths. Keith meanwhile was heavily protected narrative wise by people trying to justify an egregious overpay both from a trade value and cap allocation perspective and trying to justify or ignore the player's weaknesses.

The narrative that "Keith is the reason we made the top 4" and "Nurse was the reason the team couldn't make the finals" is still something many posters here believe to this day. The reality is Darnell Nurse played over 21 minutes a night with a torn hip flexor matching up against monsters like Kopitar/Gaudreau/Tkachuk/MacKinnon/Rantanen every shift, put up 5 even strength points, and was +5. Duncan Keith played less than 20 minutes a night primarily against middle 6/bottom 6 guys like Iafallo/Lizotte/Backlund/Compher/Newhook, put up 3 even strength points, and went -4.

The following clips are pretty much the exact same play that led to pretty much the exact same result. Yet I barely heard a peep about Keith's deflection, meanwhile you had pages and pages of flame directed towards Nurse for the exact same self goal and the PGT was full of posters blaming the loss on him and Bouchard (another player who gets a disproportionate amount of flame).



Then you have the series losing play (not that we were beating Colorado anyway just like we were pretty donezo vs Winnipeg). What Duncan Keith here did was beyond egregious. Far worse than what Ethan Bear did on the last play vs Winnipeg. Yet 2 years later and I still sometimes read posts here about Bear's bad change vs Winnipeg. Keith losing his man here had a far more direct consequence on an even more heartbreaking OT goal than Bear but this board acts as if it never happened. Make number 2's skin complexion a bit darker and change the name to "Ethan" or "Darnell" and you'd never hear the end of it.

 

[Harry Curry]

Yes, I am suggesting the guy who blew $44B because he's dramatically insecure is not smart. He was just born rich.

Musk has a degree in economics and and a degree in physics from an Ivy League school. Probably smart.

Musk's family was rich, but by no means generationally wealthy.

The born rich thing is odd. Sure the father probably had a single digit millions net worth (probably 1 - 3 million) at the time of Elon's birth. Somebody pointed out a $28,000 loan Musk got from his father. Turning that into a net worth of $189 B as of today is phenomenal. Net worth was $300 B at one point. Musk will get it back there.

Saying Musk "blew $44 B" misses the point. Musk could never have made as much money as he has without being willing to take huge risks.

If Musk was worth $233B today would his life be any different than it is with a net worth of $189B? He gets to troll Twitter any time he wants now. Which makes his life better. Sure it cost him some money, but he could afford it.

 

[TopShelfGloveSide]

I appreciate most of the analytic posters here. I look to them a lot especially when it comes to new players. I always find it interesting to see if my opinion matches the numbers. I believe people see what they want to see. Numbers don’t do that.

However there are a couple not to be named posters who get almost a god complex. Their word is gospel and if you disagree at all you are automatically stupid.

Put those few guys on ignore and you will appreciate the “nerds” a lot more.