Owly wrote:Doctor MJ wrote:Owly wrote:This may be oversimiplified but trying to boil this down to its essance:
If I'm correct you're saying injuries don't matter if the comp is to a player eliminated around the same time or especially earlier?
If so, I think I would strongly disagree.
Playoffs are a thorny, uneven playing field.
What to do with injuries is tough and hard to balance fairly.
But for someone campaigning on Chris Paul's late series struggles, where we know he's had some injuries ...
Across all players considered functional absence from the last two rounds (missed a bit of g5, all g6 round 2 [one could argue the "critical" end of it - I wouldn't], 12m cameo in the finals) their absence should nuke your chances.
If someone's absent for round 1 but you're mostly healthy RS and can expect to have locked in a strong seed, maybe you can expect to have some chance to get through it. (And I ... think ... if this happens they happen not to do so this ends up being a bigger ding on the injured guy than an injury that locks you out of tougher opponents).
In terms of the season, Durant had he locks you into "do what you can" in conference finals and finals. He happened to be on a team that could survive that [but that, along with the general long-term impact picture, kind of just pointed out who the heavier lifter was] but the title probability add he gives you is close to 0. If you're good enough to win the sort of teams faced in the final two rounds ... your odds are already really good and there may be diminishing returns on what you bring and then you can't bring it for what should be the tougher challenges anyhow.
Particularly jarring is "Just keep in mind that if Jokic had simply sat out the games after Durant got hurt, it literally wouldn't have mattered to the results." So if the team loses performance doesn't matter? Is that the implication. I don't know what else I could be expected to take from this. I don't know how much of a slippery slope it requires to get from here to "Literally only players on the champs matter". Because ... "if [any or every non-champ player] had simply have sat out the season, it literally wouldn't have mattered to the results" ... I guess it depends how much one thinks individual games for teams now rendered non-contenders (if the player lost is at this level) matter as results in this context.
As I said there are different ways to do this and it's hard to do fairly but the implication that contributions in a loss don't matter and "if X's team don't advance, Y's injury doesn't matter" ... you do see how Y's injury curtails his teams chances in about all scenarios. And okay scenarios is playing with different worlds and maybe he doesn't get injured but [a] Kevin Durant gets hurt quite a lot for the back half of his career, perhaps especially the last 5 or so years, [b] how is this any reflection on Jokic's value or championship probability added or whatever you're seeking to measure?
I don't know I just ... struggle to make sense of this approach as ... I understand it(??)
Responding to various points:
a) In the context of any given year, I find it hard to penalize Player A relative to Player B due to Player A getting injured after Player B stopped accomplishing anything.
b) This doesn't mean that we can draw broader conclusions about about endurance, durability, etc, when a player gets injured, and in the context of a Career project, that stuff can certainly act as a penalty.
c) Pertaining to my posted evaluation of Chris Paul, note that I was focusing on season-by-season analysis, and as far as I now, wasn't doing anything pertaining to Paul that was inconsistent with (a). If you see a specific bug in my logic, please do let me know. As I've said, I've certainly made such mistakes before.
d) Re: "functional absence...should nuke your chances". Not sure precisely what you're saying here. Please elaborate, and feel free to give specific examples, particularly if you think I might be inconsistent with regard to them.
e) re: "So if the team loses performance doesn't matter?". Well, let's first clarify that I said it wouldn't matter to the result. The results were losses, and presumably would have been losses had Jokic not played too.
f) I'd emphasize though that I'm not literally saying that/how Jokic played in those 2 games provides no information and can't be used in our evaluation of him. Rather, I'm saying that if those two games are what elevate Jokic ahead of Durant for someone, I think that's pretty strange. Having Jokic ahead of Durant before that point doesn't seem bizarre to me at all, but the idea that it was those 2 games that stick in people's minds as the deciding factor just doesn't seem realistic to how anyone does analysis.
Yet, for anyone whose first thought about Jokic vs Durant in '18-19 is Durant's injury, I'd argue that that's effectively what they're doing whether they realize it or not.
g) Now, I do think it understandable if one's evaluation of these players in general factors in durability/endurance explicitly, particularly in a project like this one. Just keep in mind that in the context of a particular season, if an injury-prone guy doesn't get injured, I think pretty much everyone would look at you crazy for knocking his season on the basis of the injuries he might've had be didn't.
h) Re: "how is this any reflection on Jokic?". My evaluation of Durant has no direct bearing on Jokic. This came up in the context of Top 5 seasons which you're perfectly fine to not care one bit about. Does it really matter that a guy was 6th instead of 5th in one year? Arguably no...but I'd note that when I gave my tallies for guys - and Chris Paul in particular - it sure seemed like it mattered to some people, and the response I got from them was not apathy but objection.
Briefly and if I've got a response
a) I guess I just don't get the chronological framing. We're after the fact. See (d)
d) (and maybe others). Durant was functionally absent from late in round 2. I put functionally to cover that technically he appeared in a finals game. "Nuke your chances" ... almost every team with a player of present level consideration on would if that guy was absent have no chance of winning a conference finals and a finals series. KD happened to be on one that could. But in 99.9% of cases if you've got a superstar on a superstar contract and that guy goes down and can't play those rounds your team is done. Their chances of winning are infinitesimally small. And this is probably the core of my point. If one is on a injuries happen as they happened model Durant's "championship probability added" is near zero. You're probably dead and if you aren't there's a fair chance Durant was fairly redundant anyhow.
e) See (a)
f) See (a). We've seen the season. It's not Jokic passes in two games (assuming one had him behind before). It's we don't have evidence that Jokic nukes his value by rendering himself unavailable.
g) Struggling to parse here. With the knowledge we have Durant seems injury prone at this point. He also gets injured. In "exactly as it was" models he wasn't available for the later rounds. In fuzzier likelihood models we seem to see he's injury prone and also did happen to get injured (safe to say at this end or some middle balance ... there's evidence injury affects his value this season). The "if he weren't injured ..." I'm not sure of the relevance ....
Fwiw if a Walton team got eliminated in the first round (or didn't make the playoffs) and didn't have a serious injury ... it's difficult but I could see a "we can't be sure on his continued health" penalty. It would require getting pretty deep into the weeds to apply that fairly and consistently though.
h) Yeah, arbitrary line rankings don't matter to me (at a glance it did seem very low on Paul but it's not my fight). But yeah it's just this Jokic/time framing was invoked. See (a) I guess.
Re: Why does chronology matter after the fact? Well, I suppose I'd say that it's a question of why you're changing your assessment after the fact.
So for example, if Golden State's performance after Durant's injury made you retro-actively credit him with less value-add during the time he played, to me that's perfectly reasonable. But that's not about knocking Durant for being injured, it's about the injury giving you access to new information that changed your assessment.
But if you're just talking about having a set view of 2 players over the vast majority of their respective seasons, and then one gets injured while the other loses out, shouldn't that vast-majority-of-the-season be the thing that counts for the vast majority of your evaluation of their season?
There's a philosophical concern here where I notice a tendency that we all have to try to cut the basketball out of our basketball analysis. "Player A got hurt and missed time so Player B gets the nod." Sometimes the injury is such a big deal that it really should be the dominant factor, but other times, I think we blow it out of proportion.
Let me give a hypothetical cousin to the Durant-Jokic debate:
After the regular season, Player A has the advantage over Player B.
The both players fall prey to the same freak injury and are out for the next 3 months.
Player A's team is in the playoffs and must play without him.
Player B's team missed the playoffs.
Question: Does it make sense to end up switching to rank Player B over Player A because Player A's injury hurt his team?
I can see the argument, but to me this seems clearly wrong because Player B would be getting the advantage specifically because of his team's lesser regular season performance.
Now, I think you might say: It's wrong because Player B got the same injury as Player A, which I think IS more reasonable in theory but makes us ask the question: Can we really expect to know that Player B couldn't play if his team isn't playing? I wouldn't expect any kind of certainty there, so in practice, guys on eliminated teams will tend to get the benefit of the doubt here.
We can debate that further, but heading back to the Durant-Jokic situation and how that differs. I see it as 2 things:
1. Jokic played 2 more games after Durant's injury.
2. Durant getting injured potentially says something about his general lack of reliability.
And as I've said, I don't think (1) is that big of a deal the way I would if, say, he played considerably more.
(2) clearly has a lot more room for philosophical divergence.
Re: "nukes his value by rendering himself unavailable". So, the implication of this statement to me is that a player's season achievement goes way down if he gets injured in the playoffs, and that's just not how I see it. I see it as he played how he played what he played and contributed the value he did when he was there. His missed time gives others the opportunity to surpass him, but they don't surpass him simply because they didn't get hurt.
Let me end up making sure it's clear that I'm talking about something that's analogous to MVP. So in an MVP race, while getting injured tends to hurt your MVP candidacy, it doesn't make it so that every player who isn't injured immediately jumps above you in the race.
If it did, then for example, Walton wouldn't have won MVP when he did.
) 
