sp6r=underrated wrote:Texas Chuck wrote:AEnigma wrote:He said five-year sample. I agree there is no real point in filtering out 2019, but I also would not filter out “low leverage” possessions out of some idea that it is not an issue for Jokic to be attacked when there is a notable deficit.
I do not know at what point a large chunk of this forum became more interested in plus/minus variances than in watching the games, but it makes conversation difficult. I have never felt it requires a particularly advanced eye test to watch Jokic in the postseason and see how he has been a problem, and while there is a possibility I need to scale back my standards in general, “do postseason offences specifically target Jokic” should not be the reason why.
If everything in the postseason can be dismissed as sample noise, why even bother with them. Jokic’s postseason DRAPM is +1? Hm, we need at least ten more postseasons before we can confidently say whether he is an issue. Then retroactively we may be able to confirm that, oh yes, teams were in fact exploiting him. Good for future us.
You are dead on with the overreliance on +/-. But then you turn around and take the reality that teams are going to try and attack him defensively in the playoffs and imply he's just a regular season player. That's not accurate evaluation either.
It's almost like a player can be contributing so much overall that even if they get attacked(exploited if you prefer) defensively at times they can still be very much a net positive. Even if their +/- in individual games or series isn't outstanding because the overall team gets outplayed.
This forum champions KG as this +/- darling, but that's RS. In the playoffs in Minnesota(the vast majority of his prime), he's not a plus/minus darling but we are able to look at his level of play and still realize he was a great player in the playoffs. Oh his scoring didn't hold up as well as it did in the RS but we know he contributes far beyond that.
Same with Jokic and some PNR attacks. This doesn't make him just a RS guy because of on/off stuff.
The real on/off contradiction is the love of on/off and Steve Nash, who looks quite for on/off but nothing like his backers allege.
So with all of this I can't help but see myself in the group being discussed, and I can't help but chafe at allegations of inconsistency in my analysis being implied but not specifically specified. Maybe that's me making things too much about me and my personal posting history is irrelevant, but with all of this I'm just trying to evaluate things as best I can with incomplete data - which doesn't mean I can't be inconsistent, but I'm surely trying not to be.
Starting with Nash here, I'd like to think I've been pretty clear that from an overall perspective, I don't think there are indicators saying that Nash was ever the best player in the league. That's not quite the same thing as saying he shouldn't have won one (or both) MVPs, because not everyone is at their best all season every season. For the record, I would call Nash my '04-05 RS MVP due to Duncan's injury late in the season. For '05-06 I never had Nash as my MVP choice (I chose LeBron at the time), but it was a crazy year where all the candidates were flawed and I do think Nash had a case.
Now offensively, I always did see a ton of indicators that Nash was the most impactful guy and so if that's what's being debated, well, I suppose I'm your man.
Regarding +/- in RS & PS, some thoughts:
- I'm generally a big believer in the analytical value of +/- with sufficient sample, and I do think that one regular season is generally enough to give us meaningful information. That information doesn't necessarily tell us how capable a guy is at basketball generally, but it does give us indicators of average impact over the sample.
- With that said something I've tried to emphasize frequently is that there is a potential disconnect between +/- and true impact because the goal of the game is to outscore the opponent, not outscore the opponent as much as possible. In general as we've gotten NBA data this hasn't appeared to be much of an issue, but I've considered the possibility that it would be, and that's why I started making a stat like OnWin and it's variants.
The player with the most OnWins (games with a positive +/-) obvious isn't necessarily the most valuable player, but if you have a guy who seems a bit disappointing by +/- but does better by OnWins, it's possible that the latter is a better judge of the player's capacity for impact. Kobe was the one on my mind when I made this, and I do think Kobe looks a bit stronger by OnWin than by +/- (though the difference is not as stark as I thought it might be). In today's game, Luka has been the focus there, and again, I think he looks a bit better by OnWin, but not drastically so.
- Regardless of my pet stats, what I've always said about Luka is that it's possible his resilience in the playoffs makes him an entirely different animal by impact measures, and the on/off data from his early playoff seasons has looked promising there. But as things stand, with Luka having a young career without championship runs, this is one of those "time will tell" things for me still.
- What about the opposite? When a player goes from great on/off to meh on/off in the playoffs, When is that meaningful? and What precisely does it mean?
I think to start with here is the recognition that generally superstars play big enough minutes in the playoffs that it's hard for their team to lose the series while they themselves have a positive +/-. Those who are known for OnWins when their team loses in the playoffs are worth noting of course, but they're generally guys who play a bit less than those who play the absolute most, and so I think we need to recognize that when guys lose playoff series, the entire scoreboard family of stats is probably not going to look that great for them. Maybe On/Off will spare them because the team falls off an absolute cliff without them, but the sample of those Off minutes is almost definitionally too small to have that much confidence in.
For this reason, I'm actually considerably more cautious about the possibility of using playoff +/- to say a guy struggles in the playoff than I am about the possibility of using it to show he thrives, and I really don't see any logical contradiction about about that.
When looking to identify guys who struggle in the playoffs, I'm generally looking more at the guy's production data, along with the team as a whole regularly disappointing relative to regular season performance. Of course, those are still things that have myriad confounding factors in them - they are not free from noise - but I think they tend to tell us more than the more granular +/- in these situations.