Laying down on the Job. Myers reverting to his Buffalo +- days

[IBatch]

20 hours ago, Josepho said:

I don't know what "possession stats" you're referring to, but Pittsburgh was very good in certain categories, especially the 2016 cup. 5v5xGF% and 5v5SCF% are probably the best metrics to use, because they actually managed to include shot quality into their formula.

 

I went through literally every cup finalist (not just winner) since 07-08 (through naturalstattrick), and there definitely appears to be some correlation with the aforementioned stats and success.

 

Average Leaguewide ranking in 5v5xGF% for cup finalist: 8.1

Median Leaguewide ranking in 5v5xGF% for cup finalist: 6

Average Leaguewide ranking in 5v5SCF% for cup finalist: 7.5

Median Leaguewide ranking in 5v5SCF% for cup finalist: 6.5

 

I definitely understand that QOC/QOT/Deployment is a big factor and is definitely the biggest flaw of modern analytics at the moment. However, all the stats measuring that sort of thing don't indicate that Myers is taking on the "best of the best" -- that's Edler/Schmidt (the latter being an actual good Benning acquisition I will give full credit for). But he definitely isn't "shutting down all-stars", in the sense that he isn't really playing against all-stars or shutting them down.

 

Among players with a minimum 100 minutes, Myers ranks 12/21 in 5v5GA/60, Worst on the team in 5v5xGA/60, 9th worst on the team in 5v5SCA/60. If anything, Myers has been bailed out by decent goaltending, but is still surrendering a lot of high quality chances on the ice, without taking on the toughest minutes like Edler/Schmidt. Additionally, the fact that he takes a $&!# ton of penalties while playing is a genuine issue that gets shrugged off more than it should. And my entire point about Myers struggling defensively in tough minutes is that a 6mil player shouldn't be struggling this badly in their own zone in any sort of minutes. 

 

I've seen you and the other poster use the "appeal to authority" argument, but I can just as easily say that no other GM wanted to give Myers anything close to this contract and hence it is a bad signing. Cheveldayoff in particular, after already losing two other starting RD and being in an extremely unattractive market for free agents, had no problem letting someone else give Myers this bloated contract. 

 

Also, see my point below about how Burke actually values analytics -- he may have liked the signing at the time it happened, but he'd also acknowledge that his defensive metrics are poor.

 

Don't deny that he has some offensive touch, but at 6mil you need to be a better all-situations defender.

 

So you don't think anyone hired that isn't by an NHL team has a valid opinion? Why am I supposed to care about yours then? If JFresh gets signed by a team, is his opinion magically valid and will you change your mind about Myers?

 

Hell, Brian Burke himself admitted that he was going to trade for Ristolainen but only changed his mind when his analytics department told him not to. If that's not 100% basing your decision on analytics, I don't know what is.

 

You're once again resorting to insulting people who don't share the same opinion as you, and not actually refuting their points. Maybe you should grow the &^@# up and actually attempt to explain what specifically are wrong with modern analytics models. You still didn't bother explaining to me what Myers does that analytics doesn't measure, and how it specifically impacts the success of this team.

Overall possession of the puck - LA and CHI were tops or near tops in the league at this - during CHI last two cups and LAs two cups - and at the time it was becoming all the rage.   THN wrote about it all the time - including a history of Corsi and Fenwick which are the precusor's of advanced stats - before that it's taken a long time to add anything to a basic stat line .... plus minus has a finite date - save percentages for a goalie too, hockey historians have gone back to 1950 or so and added calculated sp to goalies that never used to have one for example.   Basically the league takes a long time before they will start adding certain stats... Jim Corsi of course started this movement...

 

THN wrote an article on how PIT was a middling possession team when they won their first cup - as in bang on average league wide - and predicted that this would make the possession stat movement uncomfortable .... they did it again which proved it was no fluke.   If you watched the series you'd understand why.   They were a counterattack team - allowing the other team to gain their zone,  coming back in a 5 man box (remember no Letang the first cup - gulp Shultz was their number one guy and they really had a very average D and that's being generous)  - sound familiar? - bomber lol - creating low scoring chances that could easily be picked off,  then quickly went back the other way using their strengths which of course was Crosby, Malkin, Kessel, Guenztal etc .... if you watched them they won games they probably didn't deserve to, but they had the guns to do it and were coached brilliantly with what they had.  Their D was at best just ok.  Their second cup they were a little better overall - but far from the leaderboard of the best or better overall possession teams too.    And curiously since that, THN hasnt brought up possession with both WSH, ST. louis and TB - or since they've pretty much abandoned the subject where as for a couple years it was in most of their issues in some ways ...guess they fired their analytics department lol.

 

All the teams were buying into a team with 55-58%overall or so possession as a way to win after LA and CHI kept winning...then PIT came along and - guess what it all became about speed and skill...who cares about a few percentage points either way if it doesn't matter etc.   Every time a team wins, especially back to back - or more then once,  old ideas are curbed - new ones come up.   Since then more advanced stats have popped up - including zone entries (something TG was doing before it was vogue, was manually recording it with his staff because nobody was doing it yet).. Again most are meaningless without a lens added... where a player starts (D or Ozone), and most importantly the actual quality of competition.   Last year JVs advanced stats looked very good - until you take away that block of ten games - and added quality of competition..then not so good at all.   JV for sure isn't the best player on the team - but advanced stats suggest he might be - even McEwen at one point last year  - his PP60hmm  guess he should be on the first line and first unit PP then right?  Point is if you don't watch the player, and just use stats, and don't consider all the stats too (just cherry pick certain ones) .... well your WAR model had Steen from St. louis as the best player in the league one year - seriously.   Hilarious stuff. 

 

Just to be clear i'm not saying Myers is a top pairing defenseman.  And i would even agree that he'd be ideally suited on the third pairing on a contender - because that would mean we'd have one heck of a right side right?  But realistically he's shown enough to prove he's a capable top four guy, that's how he's been used, and it's working.   Given his D pairings he's done very well for us this year - and basic stat wise done very well league wide too.   More SH minutes then any of the top 50 matters - as does the fact he's getting limited 2 pairing minutes compared to guys feasting on the first unit - shows he's got it even without the same chances... and as another poster mentioned - QHs, Benn and OJ have been his pairings ... of those OJ looked good - Myers was barking at him all the while too...

Edited March 15, 2021 by IBatch

[JM_]

15 hours ago, kanucks25 said:

Because there's a thing called a "sample size" instead of basing your opinion on 60 minutes.

OK but thats just one game, tweety guy is saying Myers is complete trash, and its not justified at all. This kind of thing represents the worst of the stats guys, they decide the narrative and that feeds the stats trolls that like to proceed to tell everyone how stupid they are if they like what Myers brings to the team. Which is a lot, if you actually watched the games. Anyone that thinks Myers is a bottom 1% d man didn't watch the bubble playoff e.g. 

 

He happens to be 29th in points for all d men. But that must be a mistake somehow, doesn't fit Buddys tweet. 60th in toi... no that can't be right, he's to crappy for that kind of ice time. 

 

The problem with these models is they don't provide a full picture of a player or what they provide, particularly on the defensive side of the game.If tweety guy was correct, we should just be able to throw Chatifeld or Sbisa into Myers spot and expect the same results. Do you think thats reasonable? 

 

14 hours ago, ilduce39 said:

He’s at the bottom of the barrel with Hughes, Provorov, Rasmus Andersson, Josh Morrisey and Seth Myers.  
 

Pretty obvious the “model” penalizes guys who play big minutes.  It’s one thing to play around with trying to boil down players to an overall score and it’s another to slurp it up with a spoon.  You could have a pretty awesome defence going off the “worst war” charts.

yeah it has a snarky/pissy quality to it thats off putting, and even worse they've created a false narrative around a player. 

[Gurn]

How some people use stats.

 

Tyler Myers is 6 foot 8 inches tall.                                                                        True

https://www.hockeydb.com/ihdb/stats/pdisplay.php?pid=94206                  A link? Excellent!

 

That means he sucks at clearing the puck...                                                      WTF?   

 

 

[stawns]

He's been the teams best damn this season but, yeah, let's teach him a lesson

 

The stupidity is staggering

[kanucks25]

9 hours ago, Jimmy McGill said:

OK but thats just one game, tweety guy is saying Myers is complete trash, and its not justified at all. This kind of thing represents the worst of the stats guys, they decide the narrative and that feeds the stats trolls that like to proceed to tell everyone how stupid they are if they like what Myers brings to the team. 

Nobody is "stupid" for liking Myers, I can easily see the case for him on the other side.

 

If the tweety guy and some random dude across the continent that doesn't even watch our games says Myers sucks because of X stat then yes, I don't blame you for scoffing at that.

 

But, you see, people on this forum watch the games. Some watch them and pay attention to the stats. So you can't really lump all the "stat trolls" together and completely disregard the argument.

 

9 hours ago, Jimmy McGill said:

Which is a lot, if you actually watched the games. Anyone that thinks Myers is a bottom 1% d man didn't watch the bubble playoff e.g. 

Again, not really much of a sample size.

 

I remember when people were trying to convince me Gudbranson was a beast because of an allegedly good 7-game stint he had in the playoffs with Brian Campbell as his partner.

 

9 hours ago, Jimmy McGill said:

He happens to be 29th in points for all d men.

He has been good offensively. Those are good numbers despite not getting 1st PP time.

 

9 hours ago, Jimmy McGill said:

60th in toi... no that can't be right, he's to crappy for that kind of ice time. 

Y'all really need to stop with this ice-time thing. Bad players on every team every year get ice-time. Ice-time doesn't automatically =/= good.

 

9 hours ago, Jimmy McGill said:

The problem with these models is they don't provide a full picture of a player or what they provide, particularly on the defensive side of the game.If tweety guy was correct, we should just be able to throw Chatifeld or Sbisa into Myers spot and expect the same results. Do you think thats reasonable? 

I'd say the problem more is people who think the models are the bible. And then we always end up with this strawman pissing contest that starts with "omg stat nerd how about u watch teh gameszzZ?????" Not even the guys who make the models think they are the bible (at least not the intelligent ones), especially when it comes to predictor models. If you listen to a podcast where one of these guys talks about their model, they'll usually explain what they are trying to measure/show/predict but also what the pitfalls are.

 

Models that describe what's already happened are just a combination of stats AKA facts. I don't know what this guy uses but just for a simple example, one model might be a combination of GF% and FF% weighted against QoT and QoC or whatever. That's not evil voodoo magic, that's literally just a bunch of real stats put together.

 

People are afraid of what they don't understand. You could make a model that uses just G, A, and +/- and that would be much more palatable to the average fan but as most of us ought to know, some of the "traditional" stats that have been considered gospel in the past (like +/- to judge a player's two-way ability) are completely outdated. 

 

Realize that we're only like 10-12 years into analytics becoming an openly public thing, and only a few years since it's been mainstream. The new age stats aren't perfect but people are trying to figure out how to analyze a player statistically in better ways. Currently you can consider them supplemental tool.

 

As for Myers in particular it comes down to this: you and I both watch the games. We can agree to disagree on what our eye-test shows and we can agree to disagree on what stats matter. But there are real stats out there that depict Myers as not being as good as you think he may be, so I don't think this is a conversation that should be so easily dismissed as much as some of the bum-hurt people here might like it (scroll up).

 

Edited March 16, 2021 by kanucks25

[JM_]

32 minutes ago, kanucks25 said:

Nobody is "stupid" for liking Myers, I can easily see the case for him on the other side.

 

If the tweety guy and some random dude across the continent that doesn't even watch our games says Myers sucks because of X stat then yes, I don't blame you for scoffing at that.

I think thats inherent in dude's tweet. 

 

32 minutes ago, kanucks25 said:

 

But, you see, people on this forum watch the games. Some watch them and pay attention to the stats. So you can't really lump all the "stat trolls" together and completely disregard the argument.

thats fair. 

 

32 minutes ago, kanucks25 said:

 

Again, not really much of a sample size.

 

I remember when people were trying to convince me Gudbranson was a beast because of an allegedly good 7-game stint he had in the playoffs with Brian Campbell as his partner.

 

He has been good offensively. Those are good numbers despite not getting 1st PP time.

but very important, and far more intense minutes than the regular season. Playoff time is like dog years. 

 

As for Guddy, he did have a pretty OK year with Campbell. I guess it came down to thinking Guddy had more room to improve... 

 

32 minutes ago, kanucks25 said:

 

Y'all really need to stop with this ice-time thing. Bad players on every team every year get ice-time. Ice-time doesn't automatically =/= good.

no, but he's not taking the easiest matchups during that time so you do have to look at the whole picture.

 

32 minutes ago, kanucks25 said:

I'd say the problem more is people who think the models are the bible. And then we always end up with this strawman pissing contests that starts with "omg stat nerd how about u watch teh gameszzZ?????" Not even the guys who make the models think they are the bible (at least not the intelligent ones), especially when it comes to predictor models. If you listen to a podcast where one of these guys talks about their model, they'll usually explain what they are trying to measure/show/predict but also what the pitfalls are.

 

Models that describe what's already happened are just a combination of stats AKA facts. I don't know what this guy uses but just for a simple example, one model might be a combination of GF% and FF% weighted against QoT and QoC or whatever. That's not evil voodoo magic, that's literally just a bunch of real stats put together.

 

People are afraid of what they don't understand. You could make a model that uses just G, A, and +/- and that would be much more palatable to the average fan but as most of us ought to know, some of the "traditional" stats that have been considered gospel in the past (like +/- to judge a player's two-way ability) are completely outdated. 

 

Realize that we're only like 10-12 years into analytics becoming an openly public thing, and only a few years since it's been mainstream. The new age stats aren't perfect but people are trying to figure out how to analyze a player statistically in better ways. Currently you can consider them supplemental tool.

 

As for Myers in particular it comes down to this: you and I both watch the games. We can agree to disagree on what our eye-test shows and we can agree to disagree on what stats matter. But there are real stats out there that depict Myers as not being as good as you think he may be, so I don't think this is a conversation that should be so easily dismissed as much as some of the bum-hurt people here might like it (scroll up).

 

thats true. This particular tweet thing is a pretty easy example of when a model goes sideways, but eye test is also highly subjective and then we run into discussions around the value of intangibles. 

 

[kanucks25]

4 minutes ago, Jimmy McGill said:

thats true. This particular tweet thing is a pretty easy example of when a model goes sideways, but eye test is also highly subjective and then we run into discussions around the value of intangibles.

See my signature below for my favorite player of the modern era. I'm no stranger to intangibles 

 

I'm sure he's a great guy and seems to be a leader but weren't Tanev and Markstrom as well? They seemed rather essential to this team from that POV.

 

The team will move on in that respect simply because they must, the same way they had to get over the guys who left this past off-season.

 

IMO we've spent way too many assets prioritizing intangibles instead of good hockey.

[JM_]

29 minutes ago, kanucks25 said:

See my signature below for my favorite player of the modern era. I'm no stranger to intangibles 

 

I'm sure he's a great guy and seems to be a leader but weren't Tanev and Markstrom as well? They seemed rather essential to this team from that POV.

 

The team will move on in that respect simply because they must, the same way they had to get over the guys who left this past off-season.

 

IMO we've spent way too many assets prioritizing intangibles instead of good hockey.

see for me, I'd throw the numbers out the window and take Captain Serious anyway. He produces when it matters. Hope he's OK btw.

 

Same reason I'd take Crosby over McDavid for a playoff run. 

[Lil B From The Pack]

19 hours ago, Me_ said:

Just illustrating, the incalculable components of the game; heart, will, and doing what you need to do go get it done.

 

Myers is a good example of that which cannot be fully calculated with analytics.

 

I'm not gonna post my thoughts on analytics vs "heart of the game" type of stuff. You have your opinion and I respect that.

 

However I feel that wishing acts of violence towards people who have different points of view on the topic (and may not agree with your views), is extremely aggressive.

 

We all enjoy the sport in different ways and I think that should be respected.

[stawns]

The single biggest problem with the modern fan base is they think every player needs to be flawless or they're trash.  They're humans, not AI

[IBatch]

57 minutes ago, stawns said:

The single biggest problem with the modern fan base is they think every player needs to be flawless or they're trash.  They're humans, not AI

Your probably on to something.   Before video review and the digital - internet era - i remember fans being fans, connecting to a certain player, liking how he played by mostly eye test, and buying their jersey/sweater and wearing it proudly at games.   Ryan Walter was a C for a while playing for the habs, and finished off his career centering our fourth line and doing a good job.   It wouldn't be weird to see anyone walking around wearing that proudly.   Now if you saw a Beagle sweater you'd probably think - maybe he's related or something.   Fandom has changed, and not really for the better.    Sedin, Burrows, Kesler, Bieksa and a few Luongo's during the last core was all you really saw when walking around between periods (from that era - still a ton of Linden and Bure ones)   Silly really.   If i was tall enough i'd buy a Myers jersey - but instead maybe i'm thinking  Zack MzEwens.    It stayed like that through most of the WCE era too.   Harold Druken  - i saw one of those lol.  

Edited March 17, 2021 by IBatch

[kanucks25]

6 hours ago, stawns said:

The single biggest problem with the modern fan base is they think every player needs to be flawless or they're trash.  They're humans, not AI

The single biggest problem with this forum are the hyperbolic strawman arguments that make it impossible to engage in any sort of proper discussion.

[JM_]

8 minutes ago, kanucks25 said:

The single biggest problem with this forum are the hyperbolic strawman arguments that make it impossible to engage in any sort of proper discussion.

[kanucks25]

56 minutes ago, Jimmy McGill said:

Not really, Alanis. 

 

If you look at my posts, I usually try to explain my stance in-depth unless I'm replying to such a garbage post that it doesn't warrant it (ironically, like the one I quoted there).

 

The problem with some here is that their first reaction to someone having a different opinion is "I'm bum-hurt about this and here is my vitriol" instead of "okay, here's my counterpoint, what do you think?"

Edited March 17, 2021 by kanucks25

[JM_]

1 hour ago, kanucks25 said:

Not really, Alanis. 

 

If you look at my posts, I usually try to explain my stance in-depth unless I'm replying to such a garbage post that it doesn't warrant it (ironically, like the one I quoted there).

 

The problem with some here is that their first reaction to someone having a different opinion is "I'm bum-hurt about this and here is my vitriol" instead of "okay, here's my counterpoint, what do you think?"

I know, we're getting along lately so I thought it was time for some ribbing. 

 

[kanucks25]

1 hour ago, Jimmy McGill said:

I know, we're getting along lately so I thought it was time for some ribbing. 

 

o ya? well take this: