Speculation: Caps General Discussion (Coaching/FAs/Cap/Lines etc) - 2021-22 Season Part 7: Off-season Edition

[CapitalsCupReality]

I hate the offseason.

time to dust off the crazy!

Really hoping you stick around and post more. Really enjoying your contributions lately.

agreed….anyone willing to battle in the stats arena with Twabby is alright by me.


(But….so is Twabby…)

 

[HTFN]

Medical studies often involve thousands of participants such that it's possible to make reasonable conclusions not only based on the population as a whole, but in these stratified subsamples which can also contain hundreds of participants. We don't have that luxury in hockey. There are only in the hundreds of players that we can draw conclusions on in total, so stratifying is going to be subject to incredible amounts of noise.

Give me a stratification to work with that includes Fehervary, and I'll try to parse the WAR data. Do you want me to see how top 2 rounders perform after their rookie years? Do you want to see how Slovakian players do after their first year? Do you want to see how players who get more than X minutes in their rookie year perform in subsequent years? I'll try my best to parse through this data and post results. It'll likely have to use Evolving Hockey's WAR stat since I have access to their data, but it'll probably be close enough to Sprigings's WAR stat.

We shouldn't pretend that hockey is a solved problem. It wasn't that long ago where defensemen were still drafted because of size above most else, for instance. Fourth lines up until about 10 years ago consisted mainly of just dudes who could punch. Hockey is way different now than it was even 5-6 years ago.

Is it really unreasonable to think that maybe our ideas about player development timelines might be out of whack? Is it really that crazy to believe that teams give their young players who underperform, especially draftees, way more rope than they would otherwise deserve? Is it insane to think that those rookies who have shown well are not given the same opportunities as established veterans who are not as good?

I'm not saying that players don't develop. But in the case of defensemen, I suspect that almost all of that development is done before they even reach the NHL, on average.

I want to see the average of a ~10 player sample from a few different roles. I’d like to see if Norris caliber defensemen and 2nd pair defensemen see their WAR raise and lower at similar rates and ages, as well as players with a 10+ year career vs. 3-5 for starters. Might also be good to take some random draft years and compare similarly ranked prospects with various levels of NHL success and see how fully formed they were.

I’d also like to know how you expect WAR to do anything but lower year to year after what I asked you in my last post, because for the league-wide number to go up the league’s best players will have to outperform last year’s WAR, each and every year, just to buoy peers who are aging naturally. The chart has no appreciation for the league’s better players who can keep their play consistent for a number of years because no change in WAR won’t really mitigate anything. For the graph to do anything else they’d have to be achieving infinite growth

Thinking this solved anything is batshit. You talk about how WAR must be useful because the guy got hired but the quote most pertinent to this one is the executive that said (paraphrasing) the internet has most of the stats, but no grasp on where to apply them. I’m sure Colorado liked what he was bringing to the table, I’m also pretty sure he wouldn’t have that job if he suggested defensemen are all but fully formed out of the gates and definitely decline around 26 instead of entering their prime, as has been the general understanding for like 50 years.

“We don’t know everything, is it really so crazy to think that people have just been overcooking their pork and we could be having nice medium rare pork chops?”

 

[YippieKaey]

Stats notwithstanding i have to say that it's pretty common knowledge that defensemen get better with experience.

Re Fehervary we're talking a guy about to play his second full season and im very confident he'll get a lot better in the next 2-3 years. Probably not 1D but top 4 seems likelt.

 

[Jags]

Give me a stratification to work with that includes Fehervary, and I'll try to parse the WAR data. Do you want me to see how top 2 rounders perform after their rookie years? Do you want to see how Slovakian players do after their first year? Do you want to see how players who get more than X minutes in their rookie year perform in subsequent years? I'll try my best to parse through this data and post results.

The data is only useful when there are questions worth answering. The models that exist are generally helpful to suss out trends among players in general and, occasionally, a handful of telling things about an individual player. There is no perfect statistical model that allows you to predict anything, or even to fully assess a player's current or likely future performance.

You drop all these stats on us along with some conclusions, we comment on those things critically, and your response is, "Well what do you want me to do??? What data should I get? What else do you need to see?" And the answer is nothing because we didn't ask for any of it in the first place. You gather it all up to help sell your ideas and arguments. Sometimes it works, sometimes it doesn't. Sometimes when we disagree you're like, "BUT THE NUMBERS! Look at the numbers! The charts! The heat maps!" And it just ain't that simple.

Sometimes the numbers tell a story all by themselves. Most of the fancy stats require A LOT of context and perspective to be used meaningfully, otherwise you can land on some pretty outlandish conclusions, like hockey players not improving in the NHL.

Is it really unreasonable to think that maybe our ideas about player development timelines might be out of whack?

Not at all. But it's hubris to think that you're the guy that's gonna un-whack them. And it IS unreasonable to think that human behavior and performance can always be distilled down to a number. That's rarely the case. You can't look at a bunch of fancy stats and just decide that, no, in fact NHL defensemen don't improve at the NHL level. That's nutty.

Is it really that crazy to believe that teams give their young players who underperform, especially draftees, way more rope than they would otherwise deserve?

Not at all. Given the free agency rules in the NHL, a young, cost-controlled player is an extremely valuable commodity. Teams want to make damn sure there's nothing to salvage before punting on a kid they invested a pick on because when young players pay off it's years of performance at a fraction of the typical cost. In a league with a meaningful salary cap, that makes a HUGE difference.

Is it insane to think that those rookies who have shown well are not given the same opportunities as established veterans who are not as good?

I'm not sure I understand the question. If I'm right about the gist, that's something that's entirely the fault of coaches and management. Most teams in the league fully evaluate their teams and play their best combinations of players. Coaches and managers that deviate from that general rule should be beaten with a sock full of nickels.

 

[RedRocking]

Is it really unreasonable to think that maybe our ideas about player development timelines might be out of whack? Is it really that crazy to believe that teams give their young players who underperform, especially draftees, way more rope than they would otherwise deserve? Is it insane to think that those rookies who have shown well are not given the same opportunities as established veterans who are not as good?

All roads lead back to CMM, and his perceived lack of opportunities, don’t they? Is there an axe to grind against the rookie that got tons of TOI?

I hope for forwards that player development isn’t a boring thing that never happens, because CMM sure needs to develop his finishing skills so he starts to have a chance of actually solving NHL goalies.

I hope with Fehervary he hit some kind of rookie wall and/or that COVID adversely affected him. I remember in the first half of the season all the rave reviews here, and whispers of “is Fever our best D-men?”

And yea, by and large, he regressed in the second half of the season. And no, he shouldn’t be off the table if there is a good move to be made, but we really need him to keep developing and solidify himself as a Top 4. Let’s hope so.

I think analytics are always helpful as part of the picture, and I appreciate views which hold them in such high regard. But I personally trust the eye test a bit more, and the experience of a guy like Lavi (over anything I have to say really, ha). I just can’t vibe with the idea that a guy like Fever can’t get markedly better.

 

[Kalopsia]

Medical studies often involve thousands of participants such that it's possible to make reasonable conclusions not only based on the population as a whole, but in these stratified subsamples which can also contain hundreds of participants. We don't have that luxury in hockey. There are only in the hundreds of players that we can draw conclusions on in total, so stratifying is going to be subject to incredible amounts of noise.

Give me a stratification to work with that includes Fehervary, and I'll try to parse the WAR data. Do you want me to see how top 2 rounders perform after their rookie years? Do you want to see how Slovakian players do after their first year? Do you want to see how players who get more than X minutes in their rookie year perform in subsequent years? I'll try my best to parse through this data and post results. It'll likely have to use Evolving Hockey's WAR stat since I have access to their data, but it'll probably be close enough to Sprigings's WAR stat.

We shouldn't pretend that hockey is a solved problem. It wasn't that long ago where defensemen were still drafted because of size above most else, for instance. Fourth lines up until about 10 years ago consisted mainly of just dudes who could punch. Hockey is way different now than it was even 5-6 years ago.

Is it really unreasonable to think that maybe our ideas about player development timelines might be out of whack? Is it really that crazy to believe that teams give their young players who underperform, especially draftees, way more rope than they would otherwise deserve? Is it insane to think that those rookies who have shown well are not given the same opportunities as established veterans who are not as good?

I'm not saying that players don't develop. But in the case of defensemen, I suspect that almost all of that development is done before they even reach the NHL, on average.

The really big studies exist, but they're insanely expensive. It's much more common to have studies with sample sizes around 100 with extremely strict inclusion criteria in order to get the most bang for your buck.

Of course we should always be open to challenging our preconceived notions, but the bigger the challenge the stronger the evidence needs to be to support it. I did an internship in grad school where I was instructed to look at how the mRNA expression of a family of genes affected the ability of different cancer drugs to kill cancer cells. I got a couple of really counterintuitive results, where the expression of genes made drugs that were supposed to target those genes less effective when they should've made them more effective. My supervisor ran a different type of correlation analysis to corroborate my findings, then looked for other datasets to repeat the analysis on, then commissioned a lab to see if they could re-create the results. It's been three years since my internship and they're only just writing the manuscript now. Obviously I'm not expecting that for a discussion on hockey stats, but I think you should at least be more open to the possibility that there are weaknesses in that analysis and the metric that need to be addressed before the finding's taken seriously.

Ideally, I'd want to limit the aging curve analysis to defensemen who were in the top 120 in minutes in a season prior to their 23rd birthday. Guys who were given a heavy workload at a young age. That probably has too small a sample size though, so maybe just the top 120 cutoff? In all likelihood, there's probably no subset of the dataset that's simultaneously homogenous enough for the analysis to matter and large enough to have adequate statistical power.

 

[twabby]

Ideally, I'd want to limit the aging curve analysis to defensemen who were in the top 120 in minutes in a season prior to their 23rd birthday. Guys who were given a heavy workload at a young age. That probably has too small a sample size though, so maybe just the top 120 cutoff? In all likelihood, there's probably no subset of the dataset that's simultaneously homogenous enough for the analysis to matter and large enough to have adequate statistical power.

I’ll take a crack at this tomorrow when I am back at my desktop. I bet there’s at least some useful data to look at, even if it might not be as predictive as looking at the entire data set.

 

[ClevelandCapsfan]

A trade target not contemplated?

 

[SecretaryofDefense5]

A trade target not contemplated?

Gimme

 

[um]

A trade target not contemplated?

That's dumb as heck. Montreal would be nuts to move him unless its an overpayment.

 

[um]

Here I was thinking Fehervary was one of the bright spots this season, but apparently, we just wasted his peak season!

The opinions of people who don't watch the games are always fun.

 

[HTFN]

Here I was thinking Fehervary was one of the bright spots this season, but apparently, we just wasted his peak season!

The opinions of people who don't watch the games are always fun.

seriously. first half: good, second half: not super good. in regular real-world terms that's the same rookie wall nearly everyone hits, maybe with a little COVID thrown in for good measure, and a real good season to build on.

average of those, apparently Tyler Sloan and not really worth the time.

I’ll take a crack at this tomorrow when I am back at my desktop. I bet there’s at least some useful data to look at, even if it might not be as predictive as looking at the entire data set.

you gonna do my numbers too or just continue to ignore every single question I ask you

 

[SDBondra]

Here I was thinking Fehervary was one of the bright spots this season, but apparently, we just wasted his peak season!

The opinions of people who don't watch the games are always fun.

I stepped away from this thread for two weeks and come back to seeing him compared to TFS. He’s literally the only prospect we have successfully developed in the past five years. Fehervary is the last of our problems.

 

[Holtbyisms]

*Stirs the pot...* Bob 4m retained for Vitek and Haglin.

 

[trick9]

*Stirs the pot...* Bob 4m retained for Vitek and Haglin.

I doubt they will retain that much for that long.

Still it's an easy no. They will need to move few first-rounders or take back another contract like Kevin Hayes if they want to move him.

 

[twabby]

Ideally, I'd want to limit the aging curve analysis to defensemen who were in the top 120 in minutes in a season prior to their 23rd birthday. Guys who were given a heavy workload at a young age. That probably has too small a sample size though, so maybe just the top 120 cutoff? In all likelihood, there's probably no subset of the dataset that's simultaneously homogenous enough for the analysis to matter and large enough to have adequate statistical power.

Sorry for the delay.

I went ahead and did the following: I took every player-season from every defenseman from 2007-08 until 2021-22. For each season, I took the top 120 TOI player-seasons. This gave 1800 player-seasons.

Of these 1800 player-seasons, I further culled them by taking the first instance of a player under the age of 23. For instance, Aaron Ekblad has been in the top 120 TOI seasons since he entered the league, so I only considered his age 18 season (2014-15) to start.

This gave me about 109 players under the age of 23 who both got "top 4" minutes but also have more records in the database. For instance, Martin Fehervary has the former (an age 22 season with top 4 minutes) but does not yet have a subsequent season to compare to. So he's not a part of this dataset, which makes sense because we are trying to forecast his future based on last season. I took the WAR/60 of the first instance of a player under 23 getting top 4 minutes.

I then took the cumulative WAR/60 of these players following their initial top 4 season as a comparison. For instance, I took Aaron Ekblad's 2014-15 WAR/60 and plotted it against his 2015-16 thru 2021-22 cumulative WAR/60. I did this for each of the 109 players. Here is the scatterplot:

A few notes:

1. The correlation coefficient of this dataset is about r = .44, and r^2 = 0.19. This is actually a somewhat stronger correlation than I was expecting. The trendline is also shown in blue.

2. Martin Fehervary's WAR/60 in his first Top 4 year (e.g. this past year) was -0.016.

3. In the plot above, the red line represents the median top 4 player's subsequent WAR/60 (about 0.04).

Overall, I think this kind of further supports what I've been saying. Most players under the age of 23 who played a lot of minutes and had initial results similar to Fehervary never really progressed to becoming above average top 4 players. That's not to say they flame out of the league (though many did), but it's more to say most of them become expendable.

I think Washington needs to consider that Fehervary may not progress into much more than a third pairing talent for his career, especially if he doesn't improve next year. Next year will be key, and I think by the time they reach the TDL they should know much more about his projected future.

 

[Holtbyisms]

I doubt they will retain that much for that long.

Still it's an easy no. They will need to move few first-rounders or take back another contract like Kevin Hayes if they want to move him.

I dunno...sounds horrible at first but Bob is a significant upgrade over Sammy(and might be able to mentor him). Assuming Vitek gets 1.5ish+Haglins cap hit we only gain 1.75 million in salary. His window is probably over the same time as Ovi's so it takes us right there. He's also BFFs with Orlov so he'd probably waive his NTC to come here. Florida only has to eat Haglins cap hit for a year, viteks a cheap and talented backup who can further push Knight. I think it works.

 

[g00n]

Re: twabby's stats which I'm only half paying attention to...

Or, young guys rushed into top 4 roles might typically be filling in for cap-crunched teams, or crappy teams who aren't spending on top defensemen, so the sampling bias might come from a situation where you're naturally putting guys into slots they were never suited for.

Look at all the rebuilding years for the Caps. A lot of draft picks and prospects played on lines above what they should have, which made the draft record look better for McPhee if you went strictly by NHL games played from a given draft class.

What if you were to repeat this with only playoff teams? Any difference?

 

[twabby]

seriously. first half: good, second half: not super good. in regular real-world terms that's the same rookie wall nearly everyone hits, maybe with a little COVID thrown in for good measure, and a real good season to build on.

average of those, apparently Tyler Sloan and not really worth the time.


you gonna do my numbers too or just continue to ignore every single question I ask you

The numbers take time. I'll get to them when I get a chance. I do have other things on my plate (the NFT market is kind of in a tailspin right now, so I need to figure out how to pump 'n dump a bunch of PsychoKitties).

To answer your other question:

I’d also like to know how you expect WAR to do anything but lower year to year after what I asked you in my last post, because for the league-wide number to go up the league’s best players will have to outperform last year’s WAR, each and every year, just to buoy peers who are aging naturally. The chart has no appreciation for the league’s better players who can keep their play consistent for a number of years because no change in WAR won’t really mitigate anything. For the graph to do anything else they’d have to be achieving infinite growth

Thinking this solved anything is batshit. You talk about how WAR must be useful because the guy got hired but the quote most pertinent to this one is the executive that said (paraphrasing) the internet has most of the stats, but no grasp on where to apply them. I’m sure Colorado liked what he was bringing to the table, I’m also pretty sure he wouldn’t have that job if he suggested defensemen are all but fully formed out of the gates and definitely decline around 26 instead of entering their prime, as has been the general understanding for like 50 years.

“We don’t know everything, is it really so crazy to think that people have just been overcooking their pork and we could be having nice medium rare pork chops?”

I think you might not be understanding the methodology in the article. It's measuring average WAR differentials by age. For instance, it is taking Player X's age 22 year WAR, and comparing it to the same Player X's age 23 year WAR. The difference between those WARs is one data point. This process is then repeated for every player with an age 22 and 23 year old season. All of these differences are then averaged.

In essence, the process is roughly "What is the average WAR difference between a player's Age N season and their Age N+1 season?"

Why would this have to decrease year to year? If 25 year olds performed better than their 24 year old selves on average, the their cumulative WAR would increase on average.

And indeed forwards do on average see their WAR increase from age 18 to about 23-25. Just because this same trend is not noticed with defensemen does not mean it's flawed. Maybe it just means defensemen don't follow the same aging curve that we though they did. It is being axiomatically stated that defensemen take longer to cook, when the data seems to show the opposite.

Your point about the league's better players keeping their play consistent doesn't really go against anything the data is showing. The chart is showing how WAR changes over time, not the raw values of said WAR values. All it's saying is that on average defensemen see their WAR lower from the time they enter the league. Great players can still be great at older ages, mainly because they had such a high initial starting point.

It's also a bit odd to say that Sprigings would not have his job if he thought defensemen peaked at an early age, considering they:

1. Have the best defenseman in the league and he is 23 years old (Makar)
2. Traded for a 19 year old defenseman and has been playing big minutes for them since he was 20 years old (Girard)
3. Traded for an undervalued 26 year old who was cast off from the Islanders despite having a fantastic impact in earlier in his career (Toews)
4. Have another 20 year old in the lineup playing significant minutes (Byram)

 

[twabby]

There's no writeup for this one that I'm aware of, but here's another aging curve for defensemen based on a different WAR stat:



Kind of shows the same thing: a slight increase in WAR rates until age 22, and then a pretty gentle decline on average.

 

[twabby]

How to build a contender – Part 2: the Impact of Aging

<p align="center" style="font-size: 10px"><img src="//thenationnetwork...

canucksarmy.com

And here's another one, that I believe is based on WAR On Ice's (rip) old WAR data. One of the founders of WAR On Ice was recently hired by the Sabres last year after being with the Penguins since 2015.

Again, the data seems consistent: very little overall progression in a defenseman's early 20s, kind of plateauing from 22-25ish, and then a gentle decline on average.

 

[Jags]

Most players under the age of 23 who played a lot of minutes and had initial results similar to Fehervary never really progressed to becoming above average top 4 players. That's not to say they flame out of the league (though many did), but it's more to say most of them become expendable.

So here's a good microcosm of your methodology...

1. You put a lot of work into generating a pretty sophisticated set of statistics.
   
   
2. You determined that few players that fit Fehervary's profile progressed to become more than an average top-4 defenseman.
   
   
3. You then use this logic to define an average top-4 defenseman as expendable.

That is a ridiculous conclusion, twabs. A player that is an average top-4 defenseman in the NHL and is also young and cost-controlled for years to come is anything but expendable. In fact, you'd really like to have 3 or 4 of them on your team at all times.

So again, I really appreciate the time you take to provide us with some of the data that you do. A good portion of it is pretty interesting and contextually relevant. But your interpretation of that data is often bonkers.

A guy that fits the profile you just created is a pretty valuable player available to us at a bargain price. That's really good news.

I think Washington needs to consider that Fehervary may not progress into much more than a third pairing talent for his career, especially if he doesn't improve next year.

You just proved (to yourself at least) that he's an average top-4 defenseman. I'm good with that. There's room for a guy like that on our bottom 2 pairs. Yay! So he might not be a top pair guy. So what? Slot him where he belongs and focus on the next guy.

To be clear, I'm not sold on the statistical model because of the things it doesn't and likely can't weigh that could easily skew the results dramatically. But if you're satisfied with the results and feel they rank him as a rank-and-file middle-pair defenseman, I'm not sure I understand why you wouldn't be happy with that.

One of the founders of WAR On Ice was recently hired by the Sabres last year after being with the Penguins since 2015.

This kind of thing gets mentioned by stats fans from time to time. I don't find it compelling at all. Yes, these folks have shown a talent, curiosity, and passion for this type of analysis. That's why they get these jobs. It doesn't make the models they've developed up to that point more believable to me. It just indicates that they're people that a team would want to hire to help them fully evaluate their talent.

In other words, I don't always hire coders because they've created something I think is incredible. Their work just demonstrates to me their ability to create what I ask them to.

Again, the data seems consistent: very little overall progression in a defenseman's early 20s, kind of plateauing from 22-25ish, and then a gentle decline on average.

The "decline" can start as late as 25 for a kid that may have come into the league several years earlier. So not at all the "arriving at the NHL level fully cooked" part of your argument that everyone objected to. You took a deeper look and found that their "prime" can still easily cover as much as half of a lengthy NHL career, and it's the half that comes cheapest.

If we had a crystal ball that told us definitively that Fehervary is an average, middle-pair minute-chewer that might "gently decline" to a 3rd pair guy at some point, I'd be okay with that. Your idea that players in that category are a dime a dozen and can be picked up cheaply in free agency at any time is total fiction. I'll keep the one we've got, thank you very much. And if the day comes that he wants money that's out of line with his ability, then he can go be that UFA acquisition for someone else.

 

[twabby]

So here's a good microcosm of your methodology...

1. You put a lot of work into generating a pretty sophisticated set of statistics.
   
   
2. You determined that few players that fit Fehervary's profile progressed to become more than an average top-4 defenseman.
   
   
3. You then use this logic to define an average top-4 defenseman as expendable.

That is a ridiculous conclusion, twabs. A player that is an average top-4 defenseman in the NHL and is also young and cost-controlled for years to come is anything but expendable. In fact, you'd really like to have 3 or 4 of them on your team at all times.

So again, I really appreciate the time you take to provide us with some of the data that you do. A good portion of it is pretty interesting and contextually relevant. But your interpretation of that data is often bonkers.

A guy that fits the profile you just created is a pretty valuable player available to us at a bargain price. That's really good news.

I never said an average top 4 defenseman was expendable. I merely included the cutoff to show how far Fehervary was from being an average top 4 defenseman, and to show that the deck is stacked against him reaching that relatively modest level of success given his performance so far.

You just proved (to yourself at least) that he's an average top-4 defenseman. I'm good with that. There's room for a guy like that on our bottom 2 pairs. Yay! So he might not be a top pair guy. So what? Slot him where he belongs and focus on the next guy.

To be clear, I'm not sold on the statistical model because of the things it doesn't and likely can't weigh that could easily skew the results dramatically. But if you're satisfied with the results and feel they rank him as a rank-and-file middle-pair defenseman, I'm not sure I understand why you wouldn't be happy with that.

That’s not at all what was shown. He played like a sub-replacement level defenseman last year. As in not even a top 6 guy.

The question is what does this sub-replacement level performance this year mean for his future performance? There is plenty of uncertainty of course, but it appears that a pretty big majority of players who profile similarly to Fehervary go on to top out at a third pairing level or worse.

The "decline" can start as late as 25 for a kid that may have come into the league several years earlier. So not at all the "arriving at the NHL level fully cooked" part of your argument that everyone objected to. You took a deeper look and found that their "prime" can still easily cover as much as half of a lengthy NHL career, and it's the half that comes cheapest.

If we had a crystal ball that told us definitively that Fehervary is an average, middle-pair minute-chewer that might "gently decline" to a 3rd pair guy at some point, I'd be okay with that. Your idea that players in that category are a dime a dozen and can be picked up cheaply in free agency at any time is total fiction. I'll keep the one we've got, thank you very much. And if the day comes that he wants money that's out of line with his ability, then he can go be that UFA acquisition for someone else.

Again, I’m not saying Fehervary is starting his decline now. I’m saying that, on average, players of his profile don’t improve much. Given his poor year last year, not improving much from that performance would mean he’d top out as a third pairing defenseman at best. That, to me, would make him expendable. He’s not the type of guy I’d want to rely on in my top 4 next season with no contingency plan.

And he’s certainly not a guy I’d consider close to untradeable. My suspicion (and this is just a suspicion) is that a player like him could be seen as valuable by many teams in the league. We all see the way he’s talked up by Washington, so it wouldn’t shock me if a big minute, young, physical, smooth skating defenseman could be a valuable trade piece. It’s probably annoying to read at this point, but I think his perceived value is probably much higher than his actual value.

 

[HTFN]

The numbers take time. I'll get to them when I get a chance. I do have other things on my plate (the NFT market is kind of in a tailspin right now, so I need to figure out how to pump 'n dump a bunch of PsychoKitties).

To answer your other question:



I think you might not be understanding the methodology in the article. It's measuring average WAR differentials by age. For instance, it is taking Player X's age 22 year WAR, and comparing it to the same Player X's age 23 year WAR. The difference between those WARs is one data point. This process is then repeated for every player with an age 22 and 23 year old season. All of these differences are then averaged.

In essence, the process is roughly "What is the average WAR difference between a player's Age N season and their Age N+1 season?"

Why would this have to decrease year to year? If 25 year olds performed better than their 24 year old selves on average, the their cumulative WAR would increase on average.

And indeed forwards do on average see their WAR increase from age 18 to about 23-25. Just because this same trend is not noticed with defensemen does not mean it's flawed. Maybe it just means defensemen don't follow the same aging curve that we though they did. It is being axiomatically stated that defensemen take longer to cook, when the data seems to show the opposite.

Your point about the league's better players keeping their play consistent doesn't really go against anything the data is showing. The chart is showing how WAR changes over time, not the raw values of said WAR values. All it's saying is that on average defensemen see their WAR lower from the time they enter the league. Great players can still be great at older ages, mainly because they had such a high initial starting point.

It's also a bit odd to say that Sprigings would not have his job if he thought defensemen peaked at an early age, considering they:

1. Have the best defenseman in the league and he is 23 years old (Makar)
2. Traded for a 19 year old defenseman and has been playing big minutes for them since he was 20 years old (Girard)
3. Traded for an undervalued 26 year old who was cast off from the Islanders despite having a fantastic impact in earlier in his career (Toews)
4. Have another 20 year old in the lineup playing significant minutes (Byram)

I think I get it just fine.

The problem, again, is that this leaves a lot of good players “declining” as far as you see it with absolutely zero context to their peers. We’ve accepted for a very long time that a forward’s prime is somewhere between 23 and 27, earlier side for goal scorers and a little later for IQ players so it’s not unexpected to see WAR rise in the early 20’s because you’re either contributing and growing or not playing. Easy to understand, backed up by how there’s less data for younger players.

We also know that the average career length of an NHL player is about 4.5 years, because the turnover on bottom 6 players is always going to beat the ~25% who stay for 10+ years. So… the same borderline guys who are dragging the career average down are going to be the same guys who drag the WAR graph down, go figure, right around 4-5 years after they make it. How do you not already see the problem here? Those guys aren’t washing out with year to year WAR increases, are they? And there are more of them than there are All-Stars, so what does that do to an average?

The 25% with long careers aren’t going to be gaining WAR at a rate that will ever counterbalance the 75% that come and go. They themselves can experience zero loss and that’s going to flatten the line a little but it won’t raise it, and that’s stupid because that player can still be more productive than 90% of his peers. They could be putting up twice the WAR of any other player and it really won’t make an impact unless it’s changing, and even if they lose a tenth here or there it shouldn’t really matter if they’re still lapping the rest of the league.

So… yeah. It doesn’t matter on an individual predictive level at all unless you already have reason to believe a player fits one of the two categories, at which point the chart is only really useful if you think he’s a plug. The whole thing is built around the lowest common denominator, you might as well just be here saying the Caps should trade anyone in the 6th year of their career.

 

[Alexander the Gr8]

I can’t think of a single defenseman in Caps history whose best season was his rookie season.

Fehervary has a lot of upside, he’s not afraid to throw the body. I don’t think he’ll ever be like Brooks Orpik in that department, but he will bring a lot of qualities to the table at both ends of the rink.