A few points here, just so we’re all on the same page!
AdagioPace wrote:
Career best facilitation and FTr (basically aggressiveness + athleticism at 26 y o) are also due to luck/randomness? I dont get this sneaky attempt by Taylor to discredit TD, instead of viewing it as a quite predictable step in a career parable/context with slope in the right direction amd steepness. Yes it was a run not easy to replicate, as the 2004 season by KG was , as 2017 Curry was, et cetera.. He's trying to make Duncan look like Tracy McGrady ? (as if 00/2001 and 01/2002 seasons, both MVP worthy, didnt happen)
Good creation and free throw *rate* are not what Thinking Basketball means by shooting luck. Because, as you suggest, that’s not luck!
Shooting luck refers to general statistical variance on each of your different shots. Good shooting luck would mean you or your teammates make more shots than expected, based on the quality of shots you’re getting, while your opponents make fewer shots than expected.
When Ben was talking about shooting luck in the recent “Luck” podcast, he said a player having good shooting luck could be measured by:
-open and wide open 3 point shooting *percentage* of teammates going up compared to their full regular season average while a player’s on the court, and the open and wide open 3 point shooting percentage of opponents going down compared to their full regular season average when a player’s on the court.
In theory, a more complete measure of shooting luck would include other shots too (free throw shooting, open midrange and open paint shots, maybe even defended shots if you could quantify the quality of the defense), but Ben was going for something simple in the Luck podcast. I’m not sure exactly what shots Ben’s accounting for when analyzing Duncan’s luck -- did he update his Shooting Luck metric between the Luck podcast and evaluating Duncan's shooting luck? -- but this is the kind of thing he’s looking at.
How much does this matter? Well, this shooting luck inflates Duncan’s on-rating more than expected (it may also depress the off-rating too, but I’m forgetting whether he said anything there). Ben says Duncan has an outlier amount of luck, to the extent that it would make a non-negligible dent on peak Duncan’s on/off if you corrected for this. This would also lower his performance in downstream metrics, such as peak playoff RAPM and AuPM.
Correcting for shooting luck doesn’t suddenly make Duncan look like he doesn’t have a top 10 peak or career or anything. But it would mean that his 1st All Time (since 1997) ranking in peak Playoff AuPM is likely overrating him slightly.
Is there something a luck adjustment is missing? Maybe. But what would an argument like this look like? Did the Spurs passing out to the 3 point line get so much better that their shooters performed noticeably better, where the improved quality of the pass is what’s responsible for the shooting improvement beyond the fact that we’re only looking at open 3 point shots, moreso than it did for any other recent all-time player or all-time big men, but only when Duncan was on the court, only in the playoffs, and only during Duncan’s peak, in a way that Duncan deserves the credit for this? Did the exact opposite happen to Duncan’s opponents, again in a way that Duncan specifically deserves the credit for this? Was there something about Duncan’s pure presence that made his teammates more comfortable mentally when shooting their open 3s and made his opponents more uncomfortable when shooting their open 3s, but again only in the playoffs, and only during Duncan’s peak, and for Duncan moreso than any other all-time player?
These (or other similar arguments) are not totally impossible arguments to make. There absolutely could be other factors at play! But AdagioPace, I’d be hesitant to start accusing Thinking Baseball of bias against Duncan, while mischaracterizing his argument about shooting luck, without actually offering any sort of explanation for why an objective statistic on shooting luck should be missing something for Duncan moreso than any of the all-time players Duncan’s being compared to.
OhayoKD wrote:Ben is a more reasonable duncan hater, but is very much a hater
-> literally has duncan as a negative/nuetral portable player because defense isn't factored in
Quick clarification here. Yes, Ben’s portability score is an offensive portability score. Absolutely! But Ben does, and always has, given defensive portability credit to good defenders when calculating his overall value for a player.
For example, if you have two players who have a net value of +6.0, and both have a neutral offensive portability score of 0, but one gets far more of their net value from defense, then the better defender will end up with the higher CORP, because of the implicit positive defensive portability of the better defender.
So this is a non-issue for Ben’s overall assessment of Duncan. You could disagree with the assessment that Duncan’s offensive portability is natural to a slight negative. You could argue the assessment of Duncan’s defense is wrong. But it’s not true that Ben doesn’t consider defense portable, and that he doesn’t give credit to good defenders for that portability.
OhayoKD wrote:-> brought up playoff metrics that favor duncan over garnett but dismissed it as noise because in other three-year stretches duncan scores than extremely small samples from kg
Of course, we also have a large regular season sample to compare the two, and the broader context of playoff runs outside their 3 year peak, to help boost our analysis of a smaller playoff sample for KG vs Duncan in 2003.
OhayoKD wrote:-> curves duncan down on the basis of being with popavich and a co-star who plays the same position and is a very similar player, notably does not do this with steph/kerr/draymond or jordan/pippen/jackson even though the latter two's success are significantly more tied to said co-stars/coaches
As always, the ability to collect evidence =/ being able to apply that evidence consistently/coherently and with "plodding" bigs he'll just finds whatever he can to justify more "skilled" guys even when the evidence doesn't line-up
AdagioPace wrote:when he did the top 20-25 (I don't remember) some years ago his inconsistency was already glaring but his biases got a bit overwhelming lately. I feel like the stereotypization of players has gotten worse.
On the 3rd point you mention: yeah the problem with Taylor is what I would call "selective granular analysis". You won't see that amount of inquiry/scrutiny in dispelling anomalies for his favourite hyper-portable players, because he makes a lot of assumptions and grants them often the benefit of the doubt.
Players play in a given context. They have teammates, coaching staffs, training staffs, etc. In any situation, be it good or bad, if you’d like to evaluate the goodness of a player, it’s important to try to isolate them from their context. A good coach can make a player look better, a bad coach can make a player look worse, and it’s up to the analyst to decide how much of the player’s good or bad performance comes from the player and how much comes from their context, such as their coach. This is something we all have to do if you want to evaluate goodness, and it requires
subjective analysis.
You can say you think the “latter two's success are significantly more tied to said co-stars/coaches”. But it’s not a certain, definite, inarguable point that Jordan or Curry benefited more from their context than Duncan did from his. I’m sure you could offer evidence in favor of your point — I imagine it would include some sort of multi-year WOWY for example, based off a coach or a player being added to the lineup. But I imagine Ben also has evidence in favor of his assessment. He values film analysis quite a bit, and I imagine he could provide film analysis for how the coaching and costars alongside Duncan benefited Duncan as much or more than was the case for Jordan or Curry. (at least if your characterization of him thinking Duncan's situation benefited him more than Jordan's or Curry's did is true)
So I’d be careful accusing someone else of bias (“he'll just finds whatever he can to justify more "skilled" guys even when the evidence doesn't line-up” / " You won't see that amount of inquiry/scrutiny in dispelling anomalies for his favourite hyper-portable players, because he makes a lot of assumptions and grants them often the benefit of the doubt.") as if your evidence or your point of view is the objective one, as if your evidence or point of view is not also subject to bias, without actually providing any explanation for why your evidence is superior to the opposing evidence and where the opposing evidence specifically went wrong.
AdagioPace wrote:I respect Taylor's contributions, of course, in terms of footage analysis, constant flow of youtube entertaining material, but I'm starting to look at this theoretical work with more caution. Although I don't understand metrics-weighting in depth, I can see the results of his thought process materializing itself in rankings,conclusions he draws and constant hagiographic transpositions of his ideas in the PC board. Ipse Dixit on trust alone.
After all this Ben-bashing, accusing him of bias against Duncan in favor of KG, I think it's worth mentioning... Ben has 2003 Duncan > 2003 KG
