Peaks project update: #15

[freethedevil]

It's noisy just like any other impact stats can be, you can take it or leave it but i wouldn't call it "misuse". As Eballa pointed out, if anything you miused RAPM by comparing them 5 years apart.

All impact stats are not equally noisy. This is why box score regression is used.

[Timmyyy]

Your on/off doesn't mean he lifted his team more. That actually is a big misuse of the stat (Wade having a way worse bench plays in it a lot and he gets credit for it which doesn't make sense).

It's noisy just like any other impact stats can be, you can take it or leave it but i wouldn't call it "misuse". As Eballa pointed out, if anything you miused RAPM by comparing them 5 years apart.

I already pointed out how I compared them by RAPM and repeat that this is not a misuse when you care at least a little how I worded it. I did compare them how they did within their years. When you take out Dirk and Wade in the respective years the RAPM scores of the competition on top looks pretty similar (PI slightly over +6 and NPI a little below). But Wade is actually around his competition while Dirk has a gigantic gap between him and the competition of +1-2 in every data set I have except one. Both had good competition (one prime Dirk, Duncan the other prime Lebron [his RAPM didn't magically fall because of the bad finals]). That is without comparing between years. What I then added is that unless you think the league average fell of a complete cliff worth of +1.5 or +2 it is at least worth noting/considering that Dirk had that much of a gap in RAPM scores between him and Wade in the years in question within the same data sets (meaning everything else equal) and this gap is again shown in all data sets (not just one random one). I didn't say we should take it and run away with it as the conclusion. That paired with the analysis relative to competition is giving a clear indicator RAPM wise. If you want to consider it is on you but RAPM shows us this.
You on the other hand literally said that on/off numbers show how much guys lift their team. THAT is a misuse. I applied context and added to be careful with the restrictions in use accross years and said that it can only be used as a weak indicator because of the big gap that it is. You on the other hand just handed out a number and said it equals lift. So are you serious with what you are saying here? What you did was as clear a misuse as it gets.

I do not care as much about the boxscore and you actually did that analysis already so why would I address it when I explicitly said I want to give a counterview to your boxscore numbers.

That's fine to ignore them as long as you are consistent. Let me ask this though, when it comes to a player like Russell Westbrook, at his peak he didn't fair particularly well by RAPM for a mega star. One source it's like 24th, and another i think has him maybe 17th or something for 2017. Can you say we just ignore his historical box production and floor raising, and call him not a top 10 player because that's where RAPM has him? I'm not being sarcastic, i just think there has to be some reasonable weighing of both.

And finally it might be a philosophical thing but being smart on defense time and time again shows up in good defensive RAPM and I don't think this is randomness. I take a smart big on D over the flashy perimeter defender most of the time and the fact that I have them close is a big credit I give to Wade.

Dirk is a good solid defender, no one should dispute that. At least not from like 03-11. If anything thinks though that Wade was all flash(no pun intended), and wasn't an intelligent defender i doubt they paid much attention to what he was doing on that end.

See the below video, is an evidnece of Wade being a defensive anchor for a big playoff game, i'm not saying this was an all the time thing but still he's not even in his prime, so it gives you a pretty good idea of what he could and did do.

I will address the rest together. If you would have truly tried to understand where I am coming from (as EBalla did for example, props to him), this whole rest wouldn't be needed.

I didn't use boxscores in this comparison because I wanted to give a counterargument to yours. You threw them out there without any other context, when Dirk has more value outside of these boxscores. To show that, I gave you and others a different approach. I didn't say that is the only thing I look at. I said it is a complement to what you did. Everybody can make his own picture out of all of that. I never said 'guys ignore the boxscore', so where is this argument coming from? I personally don't like the boxscore, true, but I never said I ignore it or anything.

I have Westbrook significantly lower in 17, yes, just outside of the top 5 if I remember correctly. Why not as low as +/-? Because I have two eyes I use to evaluate and compare it to what +/- is telling me. In this case I think Westbrook was worse than his boxscore suggests but better than RAPM or RPM suggests.
I am really annoyed that people always come with the same tiresome arguments when people use +/-. 'You only use RAPM', 'At least be consistent with it'.
I made a long post providing +/-, interpreted it and compared it to eye test things and in the end people come around acting as if I only judge according to +/- numbers. I gave a lot more qualitative analysis to my numbers than you did, yet you make it seem I am the one only judging by one number.

What is also tiresome is when guys are putting words in someones mouth to keep the argument going. I never said Wade was flash only, I never said Wade wasn't an intelligent defender.
What I said is, and please read only the words I write, that Dirk is a smart big, being a smart big like Dirk correlates quite well with having a high DRAPM. Wades defense is more obvious to see (that was the flash part). Wade is intelligent too, but he operates more away from the basket. The video you show is nice and it shows that Wade was big time for a perimeter defender. It doesn't show he is better than an intelligent big that positions himself well in the paint nearly all of the time cutting driving lanes, helping with rotations and so on. I watched both those guys a lot and of course Wade is better relative to position but what he does doesn't strike me as more impactful in a vacuum just more wooowy and that is partially shown in +/- numbers.

[No-more-rings]

Wade stuff

Isnt that the same series Danny Green went nuclear thanks to Wade’s poor off-ball D?

Don’t let the highlights fool you into thinking he was anything close to anchor. He constantly ball watched and roamed(with varying results)

That game i posted was the one where he got his knee drained so was feeling a lot better, prior to that he dealt with a lot of pain. As i said he didn’t do that every game, but that’s more representative of his prime ability than him getting roasted or whatever.

[Mavericksfan]

Wade stuff

Isnt that the same series Danny Green went nuclear thanks to Wade’s poor off-ball D?

Don’t let the highlights fool you into thinking he was anything close to anchor. He constantly ball watched and roamed(with varying results)

That game i posted was the one where he got his knee drained so was feeling a lot better, prior to that he dealt with a lot of pain. As i said he didn’t do that every game, but that’s more representative of his prime ability than him getting roasted or whatever.

Him getting roasted and those highlight reels pretty much represent his defense. He gambled a LOT. He had the athleticism to make a lot of eye popping plays and imo was overall positive because of that and his rebounding. But he was far from an anchor defensively or a shut down defender.

Also his injury issues became a problems in later playoff runs where his defense would regress further as he lost some explosiveness.

[E-Balla]

Fine.

Westbrook's team sucked. Westbrook suddenly posted amazing box stats when his team started sucking. And yet, two players with less impressive box stats were able to "carry" teams as much while taking them much higher that very season.

the main argument against Westbrook is that he couldn't perform on a better team/contender. To that I say look at the year prior.

In 2016 Westbrook played next to KD and had a great squad around him. He was arguably the best player on the team averaging 24/8/10 for a team that was +7 in the regular season, but beat a +10 team in the second round, and went to 7 with a +11 team in the WCF. Overall in the playoffs the 2016 Thunder are the best team outside of the 09 Nuggets to miss the Finals, the 15th best team ever, and they played at a +13.4 level. Westbrook was the clear best player in their postseason run averaging 26/7/11 on 52 TS% to KD's 28/7/3 on 54 TS% and having clearly better +/- numbers than KD.

If you read my post actually attempting to understand my reasoning you'd understand I already addressed this exact argument.

Box stats have little to no correlation with winning and westbrook got kickdropped despite his box stats in the first round. Since you care about postseason weaknesses, westbrook wasn't very good off the ball. This means his ability to carry or put up impressive box stats would plummet on a team with a team capable of winning.

I assume you meant dropkicked, but either way this is completely revisionist given what actually happened in 2017 and what Westbrook's whole postseason career post 2012 has been like.

He was horrible in game 1 and they got blown out.

Game 2 Westbrook had 51/10/13 with 4 steals and 4 turnovers on 50 TS% (116 ORTG). They outscored Houston by 11 in the 41 minutes with Russ on the floor. They were outscored by 15 in the other 7 minutes.

Game 3 Westbrook had 32/12/11 with 3 steals and 5 turnovers on 53 TS% (109 ORTG). They won and Houston only had the lead for 13 seconds.

Game 4 Westbrook had 35/14/14 with 3 steals and 5 turnovers on 52 TS% (111 ORTG). They outscored Houston by 14 in the 39 minutes with Russ on the floor. They were outscored by 18 in the other 9 minutes. In the best coaching decision ever Billy Donovan left Norris Cole out there in the start of the 4th quarter while a 7 point lead Westbrook got them turned into a 3 point deficit.

Game 5 Westbrook had 47/11/9 with 7 turnovers on 57 TS% (114 ORTG). They outscored Houston by 12 in the 42 minutes with Russ on the floor. They were outscored by 18 in the other 6 minutes.

To me watching those games it was obvious what was happening. Westbrook had a terrible supporting cast and coach, and he couldn't overcome such gross incompetence. I'm not blaming him and his 41.3/11.8/11.8 with 2.5 steals a game, and outscoring Houston by 10 ppg from games 2-5 for their loss. He clearly carried that team to a level where they should've won. They had a 108.6 ORTG with Westbrook on the floor in that series and a 113.1 ORTG in the last 4 games (Houston had a 108.4 ORTG in the regular season). He was dominant like few players we've ever seen before in losing efforts.

Adding to that he went to the CF in 2016 the year right before that as the best player on the team. He's clearly good enough to win with both a terrible squad, and a contender. Postseason weaknesses only matter if they clearly affect the team's ability to play at the highest level. I can't say Westbrook's inability to play off ball as a PG limited his teams to any meaningful extent. At least no more than anyone else on this list's weaknesses did the same. With Robinson (for example) he has no good series against tough defenses as a first option, it's a legitimate criticism to say his weaknesses crippled his team. Everything should have context, you can't just go "player x never won a ring" as if that's a rebuttal to me pointing out why player y didn't win.

Why would a team wanting to win take westbrook over anybody else left?

Because he's clearly good enough to win if you focus on his results on the court and not whether or not he plays the style of basketball you like.

[liamliam1234]

The quote was directed at a user who was in quote chain on the original post. Frankly, though, you assuming it's direction was towards you is quite telling...

Yeah, it is telling how you treat any argument which does not capitulate to your own personal assessment. I did not say it was a non-sequitur there meant for me, but it does fit your attitude all over this thread.

After all, how could anyone imagine bias ever being an issue?

Did I say i was unbiased? No? Strawman has straw. The issue here is you're bringing up bias out of thin air when i made no assertion regarding my bias, or lack there of.

Is disingenuous arguing second nature to you? Why on earth would you be the one to need to initiate a conversation on bias. "I did not self-identify this potential issue, so do not bring it up!"

It is not your job to vet other people for bias.

Good to see you sincerely believe my sarcastic response. Yes, oh so logical to just accept biased arguments at face value because, oh, they decided they were unbiased. Brilliant system. No wonder you get along so well with WarriorGM.

It is people's own buisness how they formulate the points they make.

This is incomprehensibly un-self-aware. Is that why you police the formulation of those who make arguments which do not correspond to your "impact metrics + portability = quality" formula?

If those result in shaky claims, then point out how thye're shaky.

Like the past three threads.

What someone's biases are, is, simply put, none of your business.

Absolute bollocks.

When someone engages you in a discussion, you should have the decency to respond to what they say, not... whatever this is:

You did not engage me in a discussion, you started a fight because I did not arrive at the same conclusion as you using an entirely different assessment. "Decency". Yeah, really decent to drip condescension all over your responses to a debate you initiated. Really decent to ignore the other person saying "I am not assessing this the way you are." You are a bastion of decency.

Who hurt you?

The guy who has wasted my time over three threads because he cannot believe that someone does not just use some nebulous combination of "impact + scalability" to assess players, and then has the gall to act as if he is in the right.

As for the red herrings, i specifically explained why i considered each of the things i called a red herring a redherring, so. i'm not sure what this is for:

Dismissing everything as a "red herring"

I did not dismiss everything. I dismissed specific things as red herrings and which each dismissal gave you specific case by case reasoning for my dismissals. However, since you seem to have missed that, I'll recap:

Rejecting the premise of your "dismissals" is not missing them. But you were never good at accepting that someone would reject any of your assumed premises.

You claimed i am being inconsistent, however you have failed to showcase the contradictions. My argument for curry was based on three different things. So bringing up examples of me having a lower view of players based on one or two of those things is not showcasing a contradiction. If it is not mutually exclusive, it is not, by definition, contradictory. So saying this...

You make it too easy by deliberately blinding yourself to the myriad of inconsistencies you leave in your wake

...tells me you didn't read.

Oh, sorry, I should have been more sensitive to your arbitrary evaluation that says "impact metrics matter until they go against the guy I like, in which case I can subjectively declare him more portable and thus advantaged." I mocked this mentality in like the fourth post in this discussion, yet here we are again.

Here is an example to illuminate my point:

So you seriously think James Harden and Chris Paul were two of the three best players that year.

I do seriously think they could have been two of the three most impactful per minuite players in the regular season. However, as I've already explained to you, I also consider
portability
I don't think either are very portable. You can say "portability is overblown" but that isn't relevant to the consistency of my criteria.

It is considering portability is not in itself a consistent metric, even though you want to pretend otherwise.

Furthermore, I don't simply consider efficiency, i also consider volume. this is why I asked do you know what the "pm" stands for. Look at wins added that season. Paul(and curry who had injuries) plummet and the jameses come to the top. So if a player is being efficient on the basis of not doing much, I wouldn't bring them into a comparison of superstars.

Oh, so VORP.

In the playoffs, however, if players are being used similarly, then i'll value their efficiency accordingly. Green is a 3 and d wing who was a fringe starter during the playoffs. Curry is his team's offensive engine whose minuites are prioritized, just like kobe. So what wouldn't make sense for danny green, would be perfectly sensible for curry. There is no contradiction.

Ah, yes, PIPM is great, unless it suggests something weird, in which case you need to look at cumulative minutes. And if that is still weird, like with Danny Green or Brook Lopez, then we assess it by who is the "engine". What a great system. Totally not manufactured at all.

My opinion on 2001 kobe does not necessitate me changing my opinion of curry, because it was never solely based on how good curry's team was. Much like my opinion on draymond green does not neccesarily mean i must change my opinion on curry because it was never solely based on impact #'s. Ditto for westbrook. I've established why i consider these red herrings and consequently given you something to address and scrutinize.

Then you are just arguing in bad faith. You adjust your little formula as needed, weighing portability and impact and team success and whatever as you see fit, and then you attack others for not matching that invented, nonexistent formula. Being "consistent" with nonsense is not consistency. You do not get to claim consistency simply by claiming to consider extra variables. Anyone can do that. And it would be one thing if you let people play with their own variables, but no, instead you decided that it was unthinkable to have Curry below Wade or Kobe because they fall short of the two variables you values most.

You choosing not to do so doesn't warrant this:

(and should they criticise that means of assessment, ooh, I guess we can just call it a "silly tangent"!).

I have given you things you can address or criticize. You haven't done so. That is not my problem. Launching criticisms with no tie to what I've said doesn't change that.

I have, twenty posts ago. And it is your problem, considering you started this. I do not care about your little recipe. I think it is inane and do not value it in the slightest. That is not my problem, that is yours.

You're the one whose felt emboldened to make accusations of character. I'd say you're the kettle calling the pot black, but the pot itself is an imaginary one you've conjured out of thin air to justify being the kettle.

It is not imaginary when this is consistent and prolonged behaviour. Yeah, I am going to make an accusation of character. I naively accepted your query in good faith, and here we are days later talking in circles because you still cannot accept that anyone would not buy into how you rate players.

Curry does more than Kobe does until he encounters adversity.

I'm gonna guess you forgot about him
-> coming back from 3-1 down against westbrook and durant off an injury

Team accomplishment.

-> anchoring an offense with only klay as a legit option that posted a 110 rating against the second best modern playoff defense.
-> leading the warriors to two straight wins against the rockets with klay as his only legit offensive option.

Oh, look, more disingenuous framing. Yes, whine about having Klay as a second option, as if none of us can remember when he was enough of a second option to win 140 regular season games over two years. More of a second option than Wade or Kobe had during their peak years.

And considering how much you love to whine about injuries, should I point out Kawhi and Siakam, two of their best defenders, were hobbling around the court? Bet that was great for team defence.

-> scoring 45 and anchoring an offense that managed a 107 rating against the second best postseason defense of the modern era while klay went down.

Oh, he had a good game? That settles it. Kobe and Wade never had a good game against a top defence. Oh, that is right, they had multiple on a semi-frequent basis.

What amazing "adversity" have kobe and wade overcome that curry couldn't? Why would it outweigh whatever situations curry could come that kobe couldn't?

At their peaks? How about not playing on teams anywhere near as stacked at the Warriors.

A couple of years ago Elgee did this neat little calculation, with PIPM I believe, on relative team quality (http://www.backpicks.com/2017/07/06/supporting-casts-are-more-important-than-stars/). Wade never led a positive supporting cast. Kobe led a strongly positive supporting cast once. I wonder, if they were available for Curry, what do you think those numbers would indicate? Especially given what we know of Draymond's and Durant's impact scores.

Yes, Danny Green, the low usage 3-point sniper, is just so similar to Wade and Kobe.

You claimed citing pipm for green is analogous to citing pipm for curry as curry does less than kobe does. I claimed the opposite. I'm not really sure where the incredulity is coming from. Regardless you said pipm painted green as a superstar, but it didn't.

It is cute that you are really hanging your hat on "top fifteen to twenty is not superstar". Is the cognitive dissonance about your precious metric that bad?

Not that I think it is that simple, given that the Wolves have three... what is that, three top twelve players?

They don't? They don't even have 2 in the top 12 and don't even have 3 in the top 20. And 14th placed butler drops out of the 20 once i look at volume. The timberwolves also won nearly 50 games and posted a solid srs. Not that it's "inconsistent" for a player on a bad team to have a high pipm. If a player plays efficiently, luck adjustment mitigates for their efficient play not translating into impact for things out of their control. That a player like anthony davis grades out as high despite his team sucking isn't inherently a flaw in the model. It's a result of him being efficient in spite of his team.

I messed this up by associating Covington with them. Speaking of Covington, there is another top ten PIPM guy, by impact or volume. Ah, but I forgot, that does not count because we need to do the imaginary "actually a top player" adjustment.

There is something seriously off with this metric you keep hailing as the definitive means of adjustment, but you keep tritely pretending they do not exist.

There's something seriously off about countering a large sample size suggesting the metric does a good job predicting how players do with a few extreme examples:

Especially since these examples can be addressed by looking at volume as well as effiency.

Except when volume leaves them be, like with Covington and Brook Lopez and Danny Green.

What exactly is the "sample size" suggesting it does a good job? Sorry, does a good job notably beyond being a glorified BPM.

It's also puzzling you keep accusing me of solely relying on the metric when you've also dismissed the other criteria i've listed for my evaluation. It's doubly puzzling you're picking on this metric when i've cited a variety of different metrics which support my opinion.

And I have criticised them all, because impact metrics are not objective reflections of reality. And "portability" is not even an objective concept, let alone something that determines player quality.

Regardless, since i've multiple things i consider for player evaluation, you can only show me being inconsistent with examples where my opinion would contradict all my criterion of evaluation. But you have not, and instead are fixating on the trees rather than forest.

Because I do not value your "forest". A forest full of infected trees is hardly worth keeping. If you want to tend it on your own, fine, but I do not care to.

Cheap dodge. It is still just drawing from basic boxscore numbers

No, it isn't:
https://fansided.com/2018/01/11/nylon-calculus-introducing-player-impact-plus-minus/

There are three components of Player Impact Plus-Minus: a box-score prior, luck-adjusted on-off data, and luck-adjusted net rating.

It draws from luck adjusted data, a 15 year sample of rapm AND more specific knobs of box score. Simply noting that there is a box score prior does not mean it's bpm.

My apologies, it is 50% BPM mixed with 25% net rating (where Danny Green gets a lot of his value) and 25% ortng/drtng (guessing this is where Brook Lopez gets most of his value). And we make it sound special because it factors out three-point noise.

The results?
PIPM is more stable across smaller sample sizes than rapm(which needs 250 days to stabilize). IOW, it can be used for a season. So if I wanted to look at curry's value in 2019, pipm would be more reliable than eballa spamming rapm. This is why box score prior's are used.

I am not saying they have no possible utility, but there a is massive gap between "this can indicate some interesting things" and "this is what I use to evaluate players". Apologies, this and "portability".

Westbrook isn't portable because he is extremely ball dominant. Again, you can consider my criteria "myopic", if the inconsistency does not address all the criteria, it is not an inconsistency.

The disingenuousness of this was expressed prior. Who exactly do you think is going to be convinced because, "Oh, he had that extra subjective variable to cite when needed."

You set the arguments, and everyone needs to respond, oh, but only in the way you see fit; disputing the starting point of the argument itself is a non-starter.

Actually, I let you set the stage for discussion after you said you weren't convinced by my criteria. I then met you on the stage you set

The gall of this. No. No, no, no. I do not know what is worse, trying to what in any context without receipts could qualify as gaslighting... or just assuming I would not bother to check the receipts.

I made my post with Wade and Bryant, and you came objecting because I put them over Curry even though Curry had "portability" and "impact metrics", and for the past three threads you have continued on like that. That is you setting the stage. You did not say, "Explain why impact and portability is bad." You said Curry is better because of impact and portability. At no point was this ever about me wanting clarification. I said outright, repeatedly, and pretty immediately, that I disagreed with your fundamental premise and was not assessing players that way. I did not ask for you to walk me through why your method is better, because I do not take your method seriously. At no point did I say, "Make the case for Curry." I in fact explicitly said I do not have a broad problem with the case for 2017 Curry, or even for 2016 Curry when people admitted they only really cared about the regular season. You asked why I thought Wade and Kobe were better at their best, and I answered, and you never let it go. This has never been about my criteria. This has always been about your criteria. Why would I care about your criteria. And how could you possibly think that a discussion revolving entirely around your set criteria is actually something I am orchestrating. Are you seriously that full of yourself?

I've also entertained the questions you've asked me despite them making the discussion a one-sided scrutinization of my views. I've let you set the table and you're still upset.

If it cheers you up, you can have this:

Screw off. This has all been about your ego, and now you want to play if off as if you were doing a favour. The discussion is about your views because you think your views are the only ones worth considering, and that anyone who uses a different assessment – most recently, me or E-balla – needs to be promptly informed that "ackshually, that does not correspond to portability and impact metrics."

You can strong disagree, but kobe was well below average for his position in blow by rate and gambled more than curry with little results. Conidering the most important part of perimiter defense is preventing penetration, being in the bottom third, to the bottom decile in walling off the perimiter isn't gonna cut it. Kobe also was regular placed on lesser threats and hid on the post, much like harden. Curry on the other hand is above average in lowering his opposing #"s efg%, been a steals leader with low fouling rates and lesser defensive errors. He's been effective at forcing bigs into help, and he doesn't get parked in the post. The end game here is that curry's consistently made his team's defense better throughout his prime, while kobe's defensive value fluctuated from negative, to slight positive to nuetral.

The case for Wade here is certainly better, but Kobe's defensive reputation among media, analysts, and other players is not wholly invented just because he was a popular guy on the Lakers.

And of course it is a lot easier for Curry to basically play into a limited role behind Draymond than it is for Kobe to replicate the same on a team like the 2005-07 Lakers.

Hahahaha, as opposed to the three point and midrange advantage, right.

Huh?
No, I mean as in the three points you get when you hit a three pointer and the two points you get when you score a midranger. Quite embarrasing that someone whose been spending multiple threads whining about my evaluation being simplistic can't grasp the most simple example of nuanced distinction.

Distinction to the point of extenuating eight different types of three-pointers is not "nuance".

How does struggling against a box one translate to struggling against van vleet?

By all the times he struggled against Van Vleet outside of the box one.

No wait hey, here's a better question, when did kobe make lebron go 0-8 against him in a finals game?

Oh right, never.

Narratives are so easy to spin.

This is truly the most shallow possible defensive analysis, assuming it even qualifies as anything more than nonsense.

So different from spacing.

-> Spacing: Defenses going up higher to guard you
-> Drawing coverages: Teams sending an extra man to double or triple you.

Amazing how you've gone from whining about how stats don't capture nuance to getting sarcastic when someone breaks down the game for you.

You are not breaking down the game. In Curry's case any coverage drawing is absolutely the result of his spacing, considering Kobe arguably drew even more coverage on average (more a consequence of his team than anything) despite lacking the same shooting ability.

add wing defence, man-on-man defence, and lockdown defence

Kobe wasn't good at any of those, Why would I?

Plenty of players and analysts would disagree.

Basically every 3-and-D player can be said to fit into more systems than the top tier stars.

If that 3 and d player was the
-> one of the best playmakers ever
-> one of the best scorers ever
-> best of ball player in the league
-> one of the best on ball players in the league
then yes, they would. Alas, only the second applies to kobe whose closer to a 3 and d wing than curry is.

Since when have 3-and-D players struggled for fit because they were not "one of the best scorers" or "ball-players". Another super genuine framing.

This is a dumb and highly unobjective means of assessing overall player quality.

This is the criteria you brought up for basketball evaluation.

It literally was not. I said Wade and Kobe were more reliable in the toughest environments, i.e. up against physical playoff defence and not surrounded by one of the greatest rosters ever assembled. The i.e. is my clarification, since apparently you missed all the other times I referenced those being my major issues with Curry.

If you want, as this criteria has stopped working for you, you're welcome to give us a new one to compare them on.

For three threads we have been using your criteria, at your behest. Again, screw off.

The argument for kobe is.....

This is the problem I have seen with a few of the "veteran" posters on this board. You have this one-track mind that refuses to deal with any criticism which does not fit in with your own belief of what qualifies as "legitimate".

Oh, right, you don't have one.

I gave it twenty posts ago, but you just ignored it. And I repeated myself, and you ignored it. Because you did not start this conversation to promote discussion or to have your mind changed; you started this conversation because you wanted people to fall in line with your method of assessment.

Too bad.

But for those people who love regular season impact metrics but do not fetishise some nebulous concept of "portability", this may be too late, but have fun with this (http://www.backpicks.com/2017/10/02/the-plus-minus-goat-list-1994-2016/). I especially like the part where both 2016 and 2017 Curry trails 2016 and 2017 Draymond, 2010 Wade, and 1995 Robinson.

[No-more-rings]

I already pointed out how I compared them by RAPM and repeat that this is not a misuse when you care at least a little how I worded it. I did compare them how they did within their years. When you take out Dirk and Wade in the respective years the RAPM scores of the competition on top looks pretty similar (PI slightly over +6 and NPI a little below). But Wade is actually around his competition while Dirk has a gigantic gap between him and the competition of +1-2 in every data set I have except one. Both had good competition (one prime Dirk, Duncan the other prime Lebron [his RAPM didn't magically fall because of the bad finals]). That is without comparing between years. What I then added is that unless you think the league average fell of a complete cliff worth of +1.5 or +2 it is at least worth noting/considering that Dirk had that much of a gap in RAPM scores between him and Wade in the years in question within the same data sets (meaning everything else equal) and this gap is again shown in all data sets (not just one random one).

But just calling the competition "good for both years" is slightly lazy. As Eballa pointed out, we seen a pretty decent drop in Wade and Lebron's RAPM from 2010 to 2011, they didn't get way worse they just couldn't produce the same impact having to share the ball. Dirk probably wouldn't have led the league let alone by a good margin if that duo never teamed up.

Wade beat Dirk in RAPM in 06, 09 and 2010. 2011 Dirk finally beat him but propping up the one year he did in all of Wade's 4 best seasons feels..idk like you're ignoring anything that goes against it. It's a one year vs one year thing, but i don't see the sense in ignoring the other years they were in their prime as if Dirk suddenly acquired magical powers. That wasn't even Dirk's best regular season, it was either 06 or 07.

You on the other hand literally said that on/off numbers show how much guys lift their team. THAT is a misuse. I applied context and added to be careful with the restrictions in use accross years and said that it can only be used as a weak indicator because of the big gap that it is. You on the other hand just handed out a number and said it equals lift. So are you serious with what you are saying here? What you did was as clear a misuse as it gets.

This is kind of the pot calling the kettle black here. Raw on/off vs RAPM vs the ortgs you're leading all are still just estimates of your impact.

You say i applied no context(which is fair)when that was exactly what you did with bringing up how they performed vs their playoff competition.

I will address the rest together. If you would have truly tried to understand where I am coming from (as EBalla did for example, props to him), this whole rest wouldn't be needed.

You threw them out there without any other context, when Dirk has more value outside of these boxscores. To show that, I gave you and others a different approach. I didn't say that is the only thing I look at. I said it is a complement to what you did. Everybody can make his own picture out of all of that. I never said 'guys ignore the boxscore', so where is this argument coming from? I personally don't like the boxscore, true, but I never said I ignore it or anything.

You must be ignoring it if the gap is that big though?

It does not show definitively that Dirk had more value outside the box score though, it's estimates the impact he had in his role.

I have Westbrook significantly lower in 17, yes, just outside of the top 5 if I remember correctly. Why not as low as +/-? Because I have two eyes I use to evaluate and compare it to what +/- is telling me. In this case I think Westbrook was worse than his boxscore suggests but better than RAPM or RPM suggests.

Which this is more what i'm trying to get out of this whole thing, do your eyes tell you Dirk was better than Wade? If yes, then of course I'm not going to convince you otherwise. My eyes would disagree, and unless you're a Mavs fan/Wade hater i don't get how you'd think Dirk was better by eye test.

What I said is, and please read only the words I write, that Dirk is a smart big, being a smart big like Dirk correlates quite well with having a high DRAPM. Wades defense is more obvious to see (that was the flash part). Wade is intelligent too, but he operates more away from the basket. The video you show is nice and it shows that Wade was big time for a perimeter defender. It doesn't show he is better than an intelligent big that positions himself well in the paint nearly all of the time cutting driving lanes, helping with rotations and so on. I watched both those guys a lot and of course Wade is better relative to position but what he does doesn't strike me as more impactful in a vacuum just more wooowy and that is partially shown in +/- numbers.

I still take issue with this part, because it's all assumption. And it keeps going back to the default stance "Dirk was a solid big so he surely has more impact than a good guard". Well, i don't think it's as simple as that. Dirk to my knowledge had by far and away his best DRAPM next to Tyson Chandler. Did that have nothing to do with it , and did Dirk somehow become a defensive force at 33 years old?

-Dirk does not block or contest a lot of shots at the rim

-Dirk was a good defensive rebounder in the playoffs, but not in 2011the year argued

-Dirk was a solid post defender, but he's not shutting anyone down in there

-Dirk is relatively slow moving his feet, and no he's not doing that great at cutting off drivers

[No-more-rings]

Isnt that the same series Danny Green went nuclear thanks to Wade’s poor off-ball D?

Don’t let the highlights fool you into thinking he was anything close to anchor. He constantly ball watched and roamed(with varying results)

That game i posted was the one where he got his knee drained so was feeling a lot better, prior to that he dealt with a lot of pain. As i said he didn’t do that every game, but that’s more representative of his prime ability than him getting roasted or whatever.

Him getting roasted and those highlight reels pretty much represent his defense. He gambled a LOT. He had the athleticism to make a lot of eye popping plays and imo was overall positive because of that and his rebounding. But he was far from an anchor defensively or a shut down defender.

Also his injury issues became a problems in later playoff runs where his defense would regress further as he lost some explosiveness.

Prime Wade did not get roasted when healthy.

Your last sentence doesn't matter, since prime Wade was typically healthy in the playoffs.

I should dig up Elgee's breakdown of Wade's defense, where his tracking shows low defensive error. Then again why bother? You probably have your mind made up anyway.

[Timmyyy]

But just calling the competition "good for both years" is slightly lazy. As Eballa pointed out, we seen a pretty decent drop in Wade and Lebron's RAPM from 2010 to 2011, they didn't get way worse they just couldn't produce the same impact having to share the ball. Dirk probably wouldn't have led the league let alone by a good margin if that duo never teamed up.

Wade beat Dirk in RAPM in 06, 09 and 2010. 2011 Dirk finally beat him but propping up the one year he did in all of Wade's 4 best seasons feels..idk like you're ignoring anything that goes against it. It's a one year vs one year thing, but i don't see the sense in ignoring the other years they were in their prime as if Dirk suddenly acquired magical powers. That wasn't even Dirk's best regular season, it was either 06 or 07.

This is about peaks. Dirk was better than in his other prime years mainly because the team surrounded him better, but he still was better. Why would I ignore that.

This is kind of the pot calling the kettle black here. Raw on/off vs RAPM vs the ortgs you're leading all are still just estimates of your impact.

You say i applied no context(which is fair)when that was exactly what you did with bringing up how they performed vs their playoff competition.

Actually I wrote text to nearly every number I came up with and applied a lot of context (Wade having worse teammates, Dirk having Terry).

You must be ignoring it if the gap is that big though?

I see the +/- stuff and think that it paints a different picture and in the end come to the conclusion that it is close.

Which this is more what i'm trying to get out of this whole thing, do your eyes tell you Dirk was better than Wade? If yes, then of course I'm not going to convince you otherwise. My eyes would disagree, and unless you're a Mavs fan/Wade hater i don't get how you'd think Dirk was better by eye test.

I answered this already. If you only look at the obvious things in basketball like PPG, athleticism, ok then there is no case, go ahead call guys out for bias. I analyse with my eyes what I see and come to a different conclusion because I watch basketball a different way....without being biased.

I still take issue with this part, because it's all assumption. And it keeps going back to the default stance "Dirk was a solid big so he surely has more impact than a good guard". Well, i don't think it's as simple as that. Dirk to my knowledge had by far and away his best DRAPM next to Tyson Chandler. Did that have nothing to do with it , and did Dirk somehow become a defensive force at 33 years old?

-Dirk does not block or contest a lot of shots at the rim

-Dirk was a good defensive rebounder in the playoffs, but not in 2011the year argued

-Dirk was a solid post defender, but he's not shutting anyone down in there

-Dirk is relatively slow moving his feet, and no he's not doing that great at cutting off drivers

You didn't bring a single indicator on the table that proofs me wrong, I at least have DRAPM to back my claim up. That is just a difference in philosophy.

I will check out now, since we aren't really getting any further and just go into details I never intended to discuss.

[Mavericksfan]

That game i posted was the one where he got his knee drained so was feeling a lot better, prior to that he dealt with a lot of pain. As i said he didn’t do that every game, but that’s more representative of his prime ability than him getting roasted or whatever.

Him getting roasted and those highlight reels pretty much represent his defense. He gambled a LOT. He had the athleticism to make a lot of eye popping plays and imo was overall positive because of that and his rebounding. But he was far from an anchor defensively or a shut down defender.

Also his injury issues became a problems in later playoff runs where his defense would regress further as he lost some explosiveness.

Prime Wade did not get roasted when healthy.

Your last sentence doesn't matter, since prime Wade was typically healthy in the playoffs.

I should dig up Elgee's breakdown of Wade's defense, where his tracking shows low defensive error. Then again why bother? You probably have your mind made up anyway.

www.backpicks.com/2018/02/15/backpicks-goat-22-dwyane-wade/

It’s right there. I’m also no longer sure which version of Wade you’re talking about. You posted a video of 2013 Wade who DID get roasted and had injury issues. That’s what I was responding to. Are we talking about Pre-Lebron Wade only? Only his absolute peak?

Wade was a positive defender and I said as much. I just disagree with the idea that he was an anchor at the position. Elgee has Wade as elite for good defensive possessions but he put Wade in about the 66th percentile for containing dribble penetration.

His shot blocking, and ability to play the passing lane is what provided his main value defensively imo.

I think part of this is semantics. When you call Wade an anchor what guards do you compare his defensive impact to?

[No-more-rings]

Still Wade for me, on top of this quote i think i’ve more than made enough points regarding him even if others still disagree. Number 3 ill add in:

3. 95 David Robinson- If KG is in, Drob probably shouldn’t be that far behind. Very arguably the best defender left, probably wins a ring if he doesn’t face someone better than himself. Kobe didn’t have that problem, he faced good teams but not an all timer like Hakeem close to his peak. Not super confident, but getting my votes in before it’s too late. May consider switching next round.

1. 09' Wade- Carried a putrid supporting cast to 43 wins, with Wade on court they ran a 110.5 ORTG which would've been tied for 5th best in the league that year, which is actually something considering the Heat ran a slow defense-first type of pace. Without Wade, they were playing with the worst offense in the league by far.(99.4 ORTG). The Heat were -3.2 DRTG better with Wade on the court also. Heavy lifting on both ends of the court. Playoff performance could've been better(though 29/5/5 on 57 ts% and 26.3 PER is nothing to sneeze at), but Wade dealt with back spasms in the series and had to play nearly 39 mpg on high usage over the season just so his team could be competitive. I believe if Wade had a legit supporting cast in 09 and 10 these seasons would be viewed more highly than they are.

2. 06' Wade- In 06 with Wade on court the Heat ran a 112.3 ORTG which would've been the best in the league. With Wade sitting, they were again the worst offense in the league. Wade led the league in RAPM, NPI RAPM, and had an historical championship playoff run for a guard. I don't think anyone left quite had a title run like Wade's considering both numbers and competition.

So during Wade's playoff run he played the 7th ranked defense, 3rd ranked defense, 5th ranked defense and 11th best in that order. Aside from the bulls in the first round, absolutely destroyed the other 3 teams. In the ECF against the 5th ranked defense, 64 win, fresh off a finals trip Pistons, Wade averaged 26.7/5.2/5.5 on 68.4 ts%, and helped hold Rip Hamilton to 37.8% from the field and just 46.6 ts%(see videos below). In the finals, Wade put up a carry job-victory that was emulated by perhaps only Lebron, Duncan and Shaq. It doesn't make it necessarily better than all the others, just the load carried+lack of offensive help.

Footage is not the best of quality, but you can see it's him.

Watch on YouTube

Watch on YouTube

The argument for 06 is probably stronger than 09, but i think 09 could've gave you all that 06 did with more precise passing and better decision making.

[Mavericksfan]

Wade stuff

Isnt that the same series Danny Green went nuclear thanks to Wade’s poor off-ball D?

Don’t let the highlights fool you into thinking he was anything close to anchor. He constantly ball watched and roamed(with varying results)

No? Danny Green played more SF and LeBron (who got voted in off the strength of his "great" defense in that series ) gave up more of those shots than Wade by my memory (I'll confirm after watching all his 3s).

Watch on YouTube

That's 7 on Wade and 8 on LeBron if we go off who was his primary defender. He wasn't perfect but saying he's why he went nuclear is a gross exaggeration.

Just saw this. Wade was certainly the primary defender the first 2 games but you’re right in that Danny started cooking everyone after that.

That’s a mistake on my part.

[FrogBros4Life]

two players with less impressive box stats were able to "carry" teams as much while taking them much higher that very season.

Curry literally carried the goat team harder than westbrook carried a trash one. Lebron carried a better team harder than westbrook carried a trash one.

Since you like to throw out logical refutations all over the place like you're Bertrand Russell...

Fallacy is fallacious: https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/245/False-Equivalence

The operative word here is "carry", which itself is imprinted with certain connotations. No reasonable person would agree that Curry, by any means, "carried" as much of a load in 2017 as did Westbrook. Curry was playing with another top 25-ish player of all time, arguably the 2nd best shooter ever, one of the best defenders in the NBA and probably one of the best intangibles guys of all time in Draymond, and another extremely valuable glue guy who also happened to be a former All Star, All Defensive team selection and FMVP in Igoudala. Nobody on that team carried anything individually that season. That was a collective effort. And for the sake of consistency, no Lebron also did not carry a load to the same degree as Westbrook that year (but he did shoulder slightly more of a load than Curry). In the 20 games Durant missed, all but 1 of which Curry played in, Curry still had a USG ~10% less than what Westbrook "carried" that season. When Durant and Curry both play, Curry's USG drops another ~4% points, so again the notion that Curry was "carrying" anything like Westbrook was having to is pretty inaccurate. Curry's team can have the better season without Curry necessarily having to do more individual heavy lifting.

Even if you want to include things like Curry's "gravity" into your argument, to me, the team playing better because of something like Curry's gravity is more of a byproduct of his presence on the court, as opposed to being directly related to anything he did that is quantifiable in a box score. For example, Curry's gravity might improve the play of the team around him whether Curry even attempts a shot for the entire game (or even touches the ball for that matter), simply because of the threat that exists by ignoring his gravity. Like the Sword of Damocles, sometimes the threat is mightier than the execution. But Curry's gravity being effective, by way of "threatening" opposing teams, whether he subsequently follows through with the actual execution, isn't quite the same thing as Westbrook rolling up his sleeves and getting in the mud game in and game out. I think the discrepancy here might be your application of the word carry vs how others are using it. Maybe it's better to say Curry was a better leader, than it is to say he had more to carry than Russ, or performed better in a "carry job".

Now, just because Westbrook was carrying a heavier load, I'm not necessarily arguing he was a better player that season either, only that you can't necessarily prove one of those things true by proving the other. If Curry and Westbrook both ran a sprint, and Westbrook had 250lbs of weight strapped to his back, and Curry had 150lbs of weight strapped to his, and Curry crossed the finish line first....that doesn't mean Westbrook was actually the better runner....but it also doesn't mean that Curry performed as well (or better) under the same conditions.

If you want to argue that Curry's season in 2017 was "better" than Westbrook's season that same year....I think that is a fair point in and of itself. But it's not one you should try and prove on the grounds of a comparison between what they were each tasked to carry.

As for this....

Ah yes, lets casually dismiss that malone was a weak passer as we say he was better than arguably the best facilitator in the league.

I'm more or less ok using this to compare Curry to Lebron to Westbrook to Kawhi etc ( I see a few flaws, but nothing so bad to discredit it altogether). But with KD vs Curry specifically, I find it a bit problematic.

For starters, there is an extra data point on the With/Without Steph chart than there is for the With/Without KD chart, so you are starting this by comparing more pieces of data for Curry to less pieces of data for Durant. This might not necessarily change the outcome of the analysis, but it's bad practice. This graph also fails to account for the fact that Curry in 2017 played SEVENTEEN more games than Durant. That changes the scope and dynamics of using on/off as a barometer of impact. If they both played roughly the same number of games, then sure, looking at the on/off can give you some insight. But 17 more games? C'mon now....Of the 20 games Durant missed that year, the Dubs won 16 of them. Of those 16 wins, 8 of them came against teams that failed to make the playoffs. So, level of competition, relative to when each player missed entire games, is also something not accounted for in that graph. These players appearing to do better with Curry than with Durant is at least in part likely due to them playing better against inferior competition over a 19 game stretch (Durant and Curry both sat out in a late season game against the Spurs) where Durant didn't play a single minute, as much as it is Curry "elevating" them to greater heights.

Also, did you double check your sources? Most of these values seem correct but a handful of them are off compared to the values I get when calculating individual on/off splits myself, some slightly, a few a bit more noticeably.

But let's compare the numbers anyway. According to what's represented on this graph: Klay shot worse on a higher USG without Durant than he did with Durant. Klay also shot worse on a higher USG without Curry than he did with Curry. Klay shot relatively better without Durant than he did without Curry. It's fair to infer on a surface level here that Curry positively impacts Klay's shooting efficiency more so than Durant does.

Draymond shot about as well with Durant as he did with Curry. Draymond shot worse without Curry than he did without Durant, but it appears he took hundreds of shots less without Curry than without Durant, so it's not exactly an apples to apples comparison here. It's not clear what to infer from this, especially since (quizzically), Draymond's USG appears a bit lower without Curry than it without KD.

And if we take [(Draymond + Durant) vs (Draymond - Durant)] compared to [(Draymond + Curry)] vs (Draymond - Curry)] and extend that to [((Draymond + Curry) - Durant )] vs [((Draymond + Durant) - Curry)]....we see Draymond's efficiency skyrockets to 62.5TS% with KD only, up from 52.4TS% when he's on the floor with just Curry and not KD. Not an ideal sample size, but if you're going to ignore the 20 games KD missed, this seems just as valid.

Durant shot better with a lower USG when Curry played than he did without Curry. Curry shot better on a lower USG when Durant played than he did without Durant. Durant's % without Curry is better than Curry's % without Durant, despite them both maintaining a similar USG without the other on the court. From this, one might infer than Durant actually makes the game easier for Curry than vice versa.

Of Curry, Durant, Klay and Draymond....Durant had the best efficiency on the highest USG in any on/off scenario.

Durant, Thompson and Green all shot worse whenever Curry was out. Curry, Thompson and Green all shot worse whenever Durant was out.

We have no information from this graph about how Durant and/or Curry play when Thompson and/or Green (and Iggy) were out.

So, this is hardly an iron clad case to support the point you are trying to make. I'm not even saying what you are arguing here is in fact wrong (re: that Curry positively impacted the team to a greater degree than Durant), only that what you are providing as evidence doesn't give us enough corroboration to reach that conclusion. In the case of Thompson, it does seem that Curry's presence more positively correlates to his shooting efficiency. But to contrast that, in the case of Curry it seems that Durant's presence more positively correlates to Curry's shooting efficiency than Curry's presence does to Durant's. And Draymond + KD together with no Curry has Draymond at ~10% better efficiency than Draymond + Curry with no KD.

Also, none of this is really a good way to back up a claim about Curry being a great "facilitator" as we have no information about raw assist numbers, AST%, number of assisted baskets for each individual player vs unassisted, etc shown. The differences in efficiency could just as easily be due to ball movement as a 5 man unit in a set offense (which is more anti KD than it is pro Curry fwiw), double teams, spacing, home vs away performance differences for certain players etc. as it could that Curry is solely responsible for manufacturing "easier shot opportunities" for his teammates/better at it than Durant. Most teams are going to play better when their superstar(s) are on the court, but not all superstars are facilitators, simply because their team plays better with them in the game. Most players who get open looks due to another player drawing double teams/help D are probably going to see a spike in their efficiency. Spot up shooters who played around Shaq for instance, are generally going to shoot at a better efficiency when Shaq is in the game garnering extra defensive attention. But this doesn't make Shaq a facilitator. In the lexicon of basketball the word "facilitator" has a pretty widely agreed upon definition, and that does not include simply having teammates shoot more efficiently when you are on the floor. I'm not saying Curry isn't a facilitator...just that claiming he's the best one in the league based on what that graph represents is kind of strange. You have to be using a fairly loose interpretation of the word facilitator to pin your argument on something this imprecise.

[freethedevil]

s fallacious: https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/245/False-Equivalence
The operative word here is "carry", which itself is imprinted with certain connotations. No reasonable person would agree that Curry, by any means,

I don't really see the value of those connotations.

"Carry" is to lift something. The operative application of "lift" in basketball is making said team better. So how good a team is before the lift takes in does not preclude carrying. Carrying is about the lift, not the heights the lift started from. If curry and westbrook's cast played at a similar level when both were "carrying them" you may have a point, but they didn't. Westbrook had worse teammates and did worse. Curry had better teammates and did better. That alone does not tell us who "carried" more.

Nobody on that team carried anything individually that season. That was a collective effort.
And for the sake of consistency, no Lebron also did not carry a load to the same degree as Westbrook that year (but he did shoulder slightly more of a load than Curry). In the 20 games Durant missed, all but 1 of which Curry played in, Curry still had a USG ~10% less than what Westbrook "carried" that season. When Durant and Curry both play, Curry's USG drops another ~4% points, so again the notion that Curry was "carrying" anything like Westbrook was having to is pretty inaccurate.

I mean, if we're only defining lift by on ball play, i guess. But that seems an odd definition given curry is arguably the most valuable player in the league off the ball while westbrook...

isn't

Even if you want to include things like Curry's "gravity" into your argument, to me, the team playing better because of something like Curry's gravity is more of a byproduct of his presence on the court, as opposed to being directly related to anything he did that is quantifiable in a box score. For example, Curry's gravity might improve the play of the team around him whether Curry even attempts a shot for the entire game (or even touches the ball for that matter), simply because of the threat that exists by ignoring his gravity. Like the Sword of Damocles, sometimes the threat is mightier than the execution. But Curry's gravity being effective, by way of "threatening" opposing teams, whether he subsequently follows through with the actual execution, isn't quite the same thing as Westbrook rolling up his sleeves and getting in the mud game in and game out. I think the discrepancy here might be your application of the word carry vs how others are using it. Maybe it's better to say Curry was a better leader, than it is to say he had more to carry than Russ, or performed better in a "carry job".

Thats' semantics really. I'm not seeing why box score improvement vs non-box improvement would matter in a discussion regarding lift.

Now, just because Westbrook was carrying a heavier load, I'm not necessarily arguing he was a better player that season either, only that you can't necessarily prove one of those things true by proving the other. If Curry and Westbrook both ran a sprint, and Westbrook had 250lbs of weight strapped to his back, and Curry had 150lbs of weight strapped to his, and Curry crossed the finish line first....that doesn't mean Westbrook was actually the better runner....but it also doesn't mean that Curry performed as well (or better) under the same conditions.

Again, this is why I haven't argued "curry is the goat peak" based on being on the goat team. I am not using that curry was faster. I'm using that he outperformed where he should finish with the weight he carried relative to westbrook. My case is not that he won more. It's that he lifted his team more

If you want to argue that Curry's season in 2017 was "better" than Westbrook's season that same year....I think that is a fair point in and of itself. But it's not one you should try and prove on the grounds of a comparison between what they were each tasked to carry.

Again arguing who had more help is not the same as arguing who carried more. that doesn't mean the gap in team performance or the lift was heavier.

Also, I'm aware you aren't arguing this, but I'd realize like to reemphasize. Winning with a better team is easier. Carrying or impacting a better team is not:

It is harder to carry a better team by as much as you would carry a worse team. Questioning a player's winning based on his teammates makes sense. Questioning a players imapct because of winning is, generally, looking at things backwards.

I'm more or less ok using this to compare Curry to Lebron to Westbrook to Kawhi etc ( I see a few flaws, but nothing so bad to discredit it altogether). But with KD vs Curry specifically, I find it a bit problematic.

That's how i was using it tho...

For starters, there is an extra data point on the With/Without Steph chart than there is for the With/Without KD chart, so you are starting this by comparing more pieces of data for Curry to less pieces of data for Durant. This might not necessarily change the outcome of the analysis, but it's bad practice. This graph also fails to account for the fact that Curry in 2017 played SEVENTEEN more games than Durant. That changes the scope and dynamics of using on/off as a barometer of impact. If they both played roughly the same number of games, then sure, looking at the on/off can give you some insight. But 17 more games? C'mon now....Of the 20 games Durant missed that year, the Dubs won 16 of them. Of those 16 wins, 8 of them came against teams that failed to make the playoffs. So, level of competition, relative to when each player missed entire games, is also something not accounted for in that graph. These players appearing to do better with Curry than with Durant is at least in part likely due to them playing better against inferior competition over a 19 game stretch (Durant and Curry both sat out in a late season game against the Spurs) where Durant didn't play a single minute, as much as it is Curry "elevating" them to greater heights.

This is a fair reservation, however the data offers the same conclusion when we look at three years of data. I am curious what a opponent weighted version of the metric would look like.

That said, while i realize the sample disparity increases the chance for variance, it's not like the ts split is accumulative, so it's not automatically skewed towards a player who plays more. That said, we get similar data looking at the 2017-2019 warriors as a whole. I used this here, because the point wasn't really durant, it was just making the case curry is the best in the league at making his teammate's offense more efficient, which is how i define "playmaking" and "facilitation". This is supported by other things such as

-> Curry leading or being close to leading the league in opportunities created over a consistent stretch of time
-> Curry having a playoff playval(which regresses bpm with passer rating/box oc/layup %, ect relative to the league) that ranked top 15 all time despite limited longetivity(note that playval does not account for spacing).

Looking at that along side his effect on teammate effiency, he has a strong case for best "play-maker" in the league.

Also, did you double check your sources? Most of these values seem correct but a handful of them are off compared to the values I get when calculating individual on/off splits myself, some slightly, a few a bit more noticeably.

But let's compare the numbers anyway. According to what's represented on this graph: Klay shot worse on a higher USG without Durant than he did with Durant. Klay also shot worse on a higher USG without Curry than he did with Curry. Klay shot relatively better without Durant than he did without Curry. It's fair to infer on a surface level here that Curry positively impacts Klay's shooting efficiency more so than Durant does.

Draymond shot about as well with Durant as he did with Curry. Draymond shot worse without Curry than he did without Durant, but it appears he took hundreds of shots less without Curry than without Durant, so it's not exactly an apples to apples comparison here. It's not clear what to infer from this, especially since (quizzically), Draymond's USG appears a bit lower without Curry than it without KD.

And if we take [(Draymond + Durant) vs (Draymond - Durant)] compared to [(Draymond + Curry)] vs (Draymond - Curry)] and extend that to [((Draymond + Curry) - Durant )] vs [((Draymond + Durant) - Curry)]....we see Draymond's efficiency skyrockets to 62.5TS% with KD only, up from 52.4TS% when he's on the floor with just Curry and not KD. Not an ideal sample size, but if you're going to ignore the 20 games KD missed, this seems just as valid.

Eh...The sample's much smaller here so it would be higher variance. Also, since it's one player instead of the average of a set, the variance is probably much higher. Still, it's worth considering.

Durant shot better with a lower USG when Curry played than he did without Curry. Curry shot better on a lower USG when Durant played than he did without Durant. Durant's % without Curry is better than Curry's % without Durant, despite them both maintaining a similar USG without the other on the court. From this, one might infer than Durant actually makes the game easier for Curry than vice versa.

I mean, sure, but the drop off for durant was bigger. It's the drop off which gives an idea regarding how players affect each other.

Of Curry, Durant, Klay and Draymond....Durant had the best efficiency on the highest USG in any on/off scenario.

Sure but usage isn't inherently valuable. And that aside, the argument was about effect on teammate effiency, not the effiency in vacum. Not that i've argued curry vs durant on this thread...

Durant, Thompson and Green all shot worse whenever Curry was out. Curry, Thompson and Green all shot worse whenever Durant was out.

We have no information from this graph about how Durant and/or Curry play when Thompson and/or Green (and Iggy) were out.

Okay, but the main thing here is by how much they shot worse. As both are scoring threats, they would be expected to make their teammates more effecient, That doesn't mean the extent they do it to is similar.

So, this is hardly an iron clad case to support the point you are trying to make. I'm not even saying what you are arguing here is in fact wrong (re: that Curry positively impacted the team to a greater degree than Durant), only that what you are providing as evidence doesn't give us enough corroboration to reach that conclusion.

I wasn't really trying to argue curry vs Durant....

Also, none of this is really a good way to back up a claim about Curry being a great "facilitator" as we have no information about raw assist numbers, AST%, number of assisted baskets for each individual player vs unassisted, etc shown. The differences in efficiency could just as easily be due to ball movement as a 5 man unit in a set offense (which is more anti KD than it is pro Curry fwiw),

Okay, i never really brought up durant here, so I don't know why you think i was arguing that. The point was curry's effect on his team's offense effiency comapred very well with the best playmakers in the league, not how he compares with durant.

double teams, spacing, home vs away performance differences for certain players etc. as it could that Curry is solely responsible for manufacturing "easier shot opportunities" for his teammates/better at it than Durant. Most teams are going to play better when their superstar(s) are on the court, but not all superstars are facilitators, simply because their team plays better with them in the game. Most players who get open looks due to another player drawing double teams/help D are probably going to see a spike in their efficiency. Spot up shooters who played around Shaq for instance, are generally going to shoot at a better efficiency when Shaq is in the game garnering extra defensive attention. But this doesn't make Shaq a facilitator. In the lexicon of basketball the word "facilitator" has a pretty widely agreed upon definition, and that does not include simply having teammates shoot more efficiently when you are on the floor. I'm not saying Curry isn't a facilitator...just that claiming he's the best one in the league based on what that graph represents is kind of strange. You have to be using a fairly loose interpretation of the word facilitator to pin your argument on something this imprecise.

[/quote]

Im defining facilitator and playmaker as "making your teammates more effecient". I realize it's not how most people define it, but imo, if you're not going to look at the end game(making it easier for your team to score), the term loses practical value. I've shifted to "faciliator" because i've been told it's been suggested as more appropiate than playmaker. If you have a single word that would better describe "making your teammates more effecient", I'm open to suggestions.

And yes, I consider shaq a great playmaker. Because, when it came down to what mattered, making it easier for his teammates to score, he was, arguably, the best in the league.

I also oppose the notion that definitions should be used based on wide acceptance, rather than utlility.

[freethedevil]

I'm going to ignore the various unfounded assumptions of my character you've made. you seem to think Ad hominem(the fixation of an arguer's credibility rather than the argument) is a good thing. I don't feel like writing an essay explaining why I think policing people's thinking processes isn't good practice, so I'll just get to the heart of the matter. The tldr of the hypotethetical is: as people should be defined by what they do, not what they think... arguments should be judged as they are, not based on what you assume motivated the crafting of said arguments. With that out of the way...

Because I do not value your "forest".

Reminder... you... have posted with the explicit goal of showcasing my inconsistency. So whether you value what my forest is does not matter. You are arguing I'm inconsistent. To show I'm inconsistent, you'd actually have to show, my conclusions contradict with my forest. Unless off course you're using a different definition of inconsistent, in which I'd ask you define inconistent as you want us to use it.

Oh, sorry, I should have been more sensitive to your arbitrary evaluation that says "impact metrics matter until they go against the guy I like, in which case I can subjectively declare him more portable and thus advantaged."

Except that I cited "portability" from the get go:

Top ten.
is one of the most impactful players ever by any standard.
Is one of the most portable players too.

Was the most important player on a championship winner well past his prime.

Typically the first argument you see against him is an accusation of "analytics bias" from posters or a "why are we even considering this" which indicates to me the case against him as a top ten player is a weak one. Once you get that you have people cherrypicking his obpm in the playoffs instead of using his overall bpm because...

reasons?

If you use factors outside of play to evaluate greatness, fair enough. Greatest doesn't have to equal best. But if you're going to equate the best, the data says he's top ten, and context tells you his game is highly portable(aka, he fits well on better and better teams), something which is backed up with a past prime kg being the clear cut #1 on winning celtics side.

This post was ages before i started arguing for curry. Impact #'s supported my conclusion and i still went on to highlight portability AND winning.

Whether you value it or not, it's nonsensical to assume I use portability when impact #'s contradict my conclusions. Since I have used them in the absence of such a contradiction, you asserting otherwise doesn't make sense.

Rejecting the premise of your "dismissals" is not missing them.

Whether you reject the premise is not relevant. The premise was presented. You claimed the premise i presented was one thing, but the actual provided premise was another. Whether you reject such a premise or not, complaining about a premise that was not given is disingenuous.

in which case I can subjectively declare him more portable and thus advantaged."

You have an issue with "subjective" arguments now? Well, then, I'm sure you won't use a "subjective" argument then...

adversity

Oh.
It's also amazing you've accused me of being reliant on formulas but go out of your way to paint anything not formula based as bad for being "subjective."

Oh, look, more disingenuous framing. Yes, whine about having Klay as a second option, as if none of us can remember when he was enough of a second option to win 140 regular season games over two years.

I'm sorry, are you using an achievement curry contributed to, to argue the extent of credit of curry recieves for a later achievement? You realize that's circular reasoning?

Like, this is the equivalent of me saying "gasol was a good enough second option that the lakers won b2b"... therefore "kobe leading good offenses vs great defenses" doesn't mean kobe overcame adversity."

A couple of years ago Elgee did this neat little calculation, with PIPM I believe, on relative team quality (http://www.backpicks.com/2017/07/06/supporting-casts-are-more-important-than-stars/). Wade never led a positive supporting cast. Kobe led a strongly positive supporting cast once. I wonder, if they were available for Curry, what do you think those numbers would indicate? Especially given what we know of Draymond's and Durant's impact scores.

What #'s. Team #'s? Or induvidual #"s? Team #'s would plummet. The notion that impact #"s would go down rather than up isn't supported tho:

Why do players on bad teams generally have higher impact #"s then players on good ones? Is it possible it takes better players to life better ones?...

Regardless, you've yet to make the case peak kobe was better at "overcoming adversity" than peak curry, whatever that means. The existince of greater adversity does not mean such adversity was overcome. Kobe faced greater adversity, and he was less successful facing it. On what basis was he better at overcoming it? Since both curry and kobe have burned good defenses, i'm not really sure what your basis for asserting there was a disparity here.

The case for Wade here is certainly better, but Kobe's defensive reputation among media, analysts, and other players is not wholly invented just because he was a popular guy on the Lakers.

And of course it is a lot easier for Curry to basically play into a limited role behind Draymond than it is for Kobe to replicate the same on a team like the 2005-07 Lakers.

You're appealing to authority?

I'll take it you consider pierce more important to the wolves beating the lakers in 2008 than the media analysts did?

Also not sure what gives you the impression kobe was any more than a complimentary defensive piece. He was never the centerpiece of his team's defenses. I also don't know why you think having a bigger role would hurt your defensive impact. Assuming you're decent at it, a bigger role gives you more chances to improve your team's d. How does curry's role being "limited" help him here?

Oh, so VORP.

You've acknowledged pipm is different from bpm in your post. Not sure what the point of this was, snark aside.

Distinction to the point of extenuating eight different types of three-pointers is not "nuance".

Huh? I didn't point out 8. I'm also not sure what your issue with this distinction is. Most players are significant better or worse depending on where the three is taken from. Bledsoe for example sucks on catch and shoot threes. Raptors won a championship partially because they adjusted for that.

And what does this have to do with spacing?

A team scoring on a possession from you making a three is a distinct event from a team scoring on a possession because a defender went up to guard your shot. Hence they're listed separately.

How do you claim to be a proponent of seeing the whole story while getting sarcastic whenever you're presented with specific breakdowns?

At no point was this ever about me wanting clarification

Huh?

I present you... asking for clarification on how information relevant to my criteria affects my judgement:

Alright, side-stepping that, given that I do not think we are being especially clear on the comparative numbers (generally speaking, if something is not immediately available on a player’s basketball-reference page, I would prefer a link), what do you make of this (http://www.backpicks.com/2018/06/10/aupm-2-0-the-top-playoff-performers-of-the-databall-era/)?

Here's you doing it again:

? Going back to Draymond, is he right to say he is maybe the greatest defender ever because of how scalable he is? I mean, if we gave Russell a top eight spot for his defence and smart offence, how soon will we see Draymond?

And again:

Why was his impact higher? (Alright, Embiid was better this postseason; are you arguing him as the best in the game now because he led the postseason in RAPM?) Has Curry fallen off? Without Durant, and sometimes Klay, why did his impact metrics not explode the way you keep saying is true of solo stars with great metrics?

You've literally been asking a **** ton of questions these posts about how I apply my criteria. My criteria is the focus of these posts because you're making them so. You keep whining about me being fixated on my criteria, and then repeatedly shift the focus of the discussion back on my criteria. If you don't want me to discuss my criteria, don't ask about it. I certainly didn't force you too...

what is their argument? What parts of the story do kobe, robinson, and wade's season have that overrules the winning, impact, and fit of 2017 curry

After you asserted you don't care about my criteria, i adjusted accordingly asking you to set the table.

Then you did, asserting that curry can't do as much. So, I broke down what each he and kobe can do.

You've also wanted to focus on the sample size of the postseason, so i used, postseason data. Fun fact, corp is almost exclusively considering postseason performance. I provided it to you to answer your qualms regarding 2017 being an isolated incident. You randomly assumed it was an rs stat, when again, it was provided to you considering issues you brought up.

Finally, I've answered your questions on my own inconsistencies in spite of me considering them tangential.

This is, yet another example, of me letting you set the table.

All in all, you have been dictating the terms of the discussion all the while complaining that I haven't given you the reigns.


It isn't my ego that's makin you keep bringing the discussion back to my criteria and me. Now you're mad i'm "fixating" on the table you've set for us. Hence why I'm giving you this:

[Mavericksfan]

s fallacious: https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/245/False-Equivalence
The operative word here is "carry", which itself is imprinted with certain connotations. No reasonable person would agree that Curry, by any means,

I don't really see the value of those connotations.

"Carry" is to lift something. The operative application of "lift" in basketball is making said team better. So how good a team is before the lift takes in does not preclude carrying. Carrying is about the lift, not the heights the lift started from. If curry and westbrook's cast played at a similar level when both were "carrying them" you may have a point, but they didn't. Westbrook had worse teammates and did worse. Curry had better teammates and did better. That alone does not tell us who "carried" more.

If you’re arguing from a logical perspective I don’t think you can simply dismiss the connotations associated with carry. When comparing who carried more it’s not just about “lift” , you also consider the weight of what’s being carried. One wouldnt argue they carried more because they lifted a piece of paper higher than they lifted an elephant.

Same way you can’t say that Curry carried more as a ceiling raiser when GS was the GOAT team because of offense AND defense (the latter of which Curry didnt have much to do with, the former he also has Durant and Klay). By every metric we have Westbrook carried a bigger load in 2017.

[Colbinii]

1. 2015 Chris Paul

Best PG peak behind Magic for me. Could do it all, great scorer, alltime great turnover economy season, great defender.

2. 1995 David Robinson

Best defender left and an elite offensive player.

3. 2016 Curry

Alltime great regular season coupled with a still great though not alltime great post-season.

[freethedevil]

s fallacious: https://www.logicallyfallacious.com/tools/lp/Bo/LogicalFallacies/245/False-Equivalence
The operative word here is "carry", which itself is imprinted with certain connotations. No reasonable person would agree that Curry, by any means,

I don't really see the value of those connotations.

"Carry" is to lift something. The operative application of "lift" in basketball is making said team better. So how good a team is before the lift takes in does not preclude carrying. Carrying is about the lift, not the heights the lift started from. If curry and westbrook's cast played at a similar level when both were "carrying them" you may have a point, but they didn't. Westbrook had worse teammates and did worse. Curry had better teammates and did better. That alone does not tell us who "carried" more.

If you’re arguing from a logical perspective I don’t think you can simply dismiss the connotations associated with carry. When comparing who carried more it’s not just about “lift” , you also consider the weight of what’s being carried. One wouldnt argue they carried more because they lifted a piece of paper higher than they lifted an elephant.

Is it harder to lift a piece of paper on an elephant then it is on the ground?
No.
The supporting cast is the elephant. The impact is the lift on the paper.

In fact, though in this case minimal, being so high up makes lifting the paper harder. Cue the "hard to be great on great teams" graph.

Same way you can’t say that Curry carried more as a ceiling raiser when GS was the GOAT team because of offense AND defense (the latter of which Curry didnt have much to do with, the former he also has Durant and Klay). By every metric we have Westbrook carried a bigger load in 2017.

Scroll up and you'll find a metric that literally says curry carried harder There's actually multiple metrics that say he did.

Or are you talking about usage, which literally only tracks on ball events.

[freethedevil]

Fine.

Westbrook's team sucked. Westbrook suddenly posted amazing box stats when his team started sucking. And yet, two players with less impressive box stats were able to "carry" teams as much while taking them much higher that very season.

the main argument against Westbrook is that he couldn't perform on a better team/contender. To that I say look at the year prior.

In 2016 Westbrook played next to KD and had a great squad around him. He was arguably the best player on the team averaging 24/8/10 for a team that was +7 in the regular season, but beat a +10 team in the second round, and went to 7 with a +11 team in the WCF. Overall in the playoffs the 2016 Thunder are the best team outside of the 09 Nuggets to miss the Finals, the 15th best team ever, and they played at a +13.4 level. Westbrook was the clear best player in their postseason run averaging 26/7/11 on 52 TS% to KD's 28/7/3 on 54 TS% and having clearly better +/- numbers than KD.

If you read my post actually attempting to understand my reasoning you'd understand I already addressed this exact argument.

Box stats have little to no correlation with winning and westbrook got kickdropped despite his box stats in the first round. Since you care about postseason weaknesses, westbrook wasn't very good off the ball. This means his ability to carry or put up impressive box stats would plummet on a team with a team capable of winning.

I assume you meant dropkicked, but either way this is completely revisionist given what actually happened in 2017 and what Westbrook's whole postseason career post 2012 has been like.

He was horrible in game 1 and they got blown out.

Game 2 Westbrook had 51/10/13 with 4 steals and 4 turnovers on 50 TS% (116 ORTG). They outscored Houston by 11 in the 41 minutes with Russ on the floor. They were outscored by 15 in the other 7 minutes.

Game 3 Westbrook had 32/12/11 with 3 steals and 5 turnovers on 53 TS% (109 ORTG). They won and Houston only had the lead for 13 seconds.

Game 4 Westbrook had 35/14/14 with 3 steals and 5 turnovers on 52 TS% (111 ORTG). They outscored Houston by 14 in the 39 minutes with Russ on the floor. They were outscored by 18 in the other 9 minutes. In the best coaching decision ever Billy Donovan left Norris Cole out there in the start of the 4th quarter while a 7 point lead Westbrook got them turned into a 3 point deficit.

Game 5 Westbrook had 47/11/9 with 7 turnovers on 57 TS% (114 ORTG). They outscored Houston by 12 in the 42 minutes with Russ on the floor. They were outscored by 18 in the other 6 minutes.

To me watching those games it was obvious what was happening. Westbrook had a terrible supporting cast and coach, and he couldn't overcome such gross incompetence. I'm not blaming him and his 41.3/11.8/11.8 with 2.5 steals a game, and outscoring Houston by 10 ppg from games 2-5 for their loss. He clearly carried that team to a level where they should've won. They had a 108.6 ORTG with Westbrook on the floor in that series and a 113.1 ORTG in the last 4 games (Houston had a 108.4 ORTG in the regular season). He was dominant like few players we've ever seen before in losing efforts.

Alright i can get with that. It is an anomaly of a season though, for reasons you've outlined.

Hopefully rockets westbrook can go ham with spacing and use his effecient ass cutting to prove everyone wrong.

Adding to that he went to the CF in 2016 the year right before that as the best player on the team. He's clearly good enough to win with both a terrible squad, and a contender. Postseason weaknesses only matter if they clearly affect the team's ability to play at the highest level. I can't say Westbrook's inability to play off ball as a PG limited his teams to any meaningful extent. At least no more than anyone else on this list's weaknesses did the same. With Robinson (for example) he has no good series against tough defenses as a first option, it's a legitimate criticism to say his weaknesses crippled his team. Everything should have context, you can't just go "player x never won a ring" as if that's a rebuttal to me pointing out why player y didn't win.

Westbrook was better than durant?

I'd like to hear the case.

[Mavericksfan]

I don't really see the value of those connotations.

"Carry" is to lift something. The operative application of "lift" in basketball is making said team better. So how good a team is before the lift takes in does not preclude carrying. Carrying is about the lift, not the heights the lift started from. If curry and westbrook's cast played at a similar level when both were "carrying them" you may have a point, but they didn't. Westbrook had worse teammates and did worse. Curry had better teammates and did better. That alone does not tell us who "carried" more.

If you’re arguing from a logical perspective I don’t think you can simply dismiss the connotations associated with carry. When comparing who carried more it’s not just about “lift” , you also consider the weight of what’s being carried. One wouldnt argue they carried more because they lifted a piece of paper higher than they lifted an elephant.

Is it harder to lift a piece of paper on an elephant then it is on the ground?
No.
The supporting cast is the elephant. The impact is the lift on the paper.

In fact, though in this case minimal, being so high up makes lifting the paper harder. Cue the "hard to be great on great teams" graph.

Same way you can’t say that Curry carried more as a ceiling raiser when GS was the GOAT team because of offense AND defense (the latter of which Curry didnt have much to do with, the former he also has Durant and Klay). By every metric we have Westbrook carried a bigger load in 2017.

Scroll up and you'll find a metric that literally says curry carried harder There's actually multiple metrics that say he did.

Or are you talking about usage, which literally only tracks on ball events.

Honestly this response didnt make any sense to me so I’ll kind bow out. I certainly disagree with the notion that impact = how much you carry a team. But to each his own. I dont think anyone will be able to convince you that Curry didnt have a higher work load than Westbrook in 2017.