2012-13 Player of the Year Discussion Thread

[sp6r=underrated]

Thank you. Harden looks clearly superior in prior informed and there is a negligible difference in non-prior informed. Are you the one who applies the formula or are you getting from a website?

I'm not calculating. I just find the RAPM on the internet and store it on a website.

Thanks for the response. Do you mind sharing where you get it? I understand if you won't. Thanks.

[SideshowBob]

Thank you. Harden looks clearly superior in prior informed and there is a negligible difference in non-prior informed. Are you the one who applies the formula or are you getting from a website?

I'm not calculating. I just find the RAPM on the internet and store it on a website.

Thanks for the response. Do you mind sharing where you get it? I understand if you won't. Thanks.

It's in his sig, as well as here

[colts18]

Thanks for the response. Do you mind sharing where you get it? I understand if you won't. Thanks.

I have a link with the data in my sig. I got it from using google caches that saved the RAPM data before it was taken down from the web.

[sp6r=underrated]

Thanks for the responses. I was more interested in whether you knew the formula to calculate RAPM.

[colts18]

Thanks for the responses. I was more interested in whether you knew the formula to calculate RAPM.

Here is an article on how to calculate APM.

http://www.countthebasket.com/blog/2008 ... lus-minus/

[therealbig3]

James Harden has a better season than Curry and it really isn't close.

It would be good if you elaborate here.

I think Curry's a very interesting case, because many people jumped Curry like crazy in their rankings based on the playoffs, but statistically he wasn't much different from what he'd been in the regular season. One might be inclined to say that that's evidence that people overrated his playoff run, but the dude did finish 7th in Win Shares in the regular season with various pockets of people proclaiming his performance as miraculous.

I won't deny I had Harden ahead of Curry before the playoffs, and I won't say I don't see an argument for Harden now, but Curry debates now very much have my attention.

One of the big things to think about for me: How serious was the GS threat to the Spurs? Early in the series it just looked like they had no answer for Curry. Then they won. How impressed should we be with the Curry-based Warrior performance?

Harden to me had a substantially better regular season than Curry because he produced at a higher level. He was a more dynamic scorer who could score at a higher volume than Curry and his penetration generated a ton of open shots for teammates all years through his play on the pick and roll.

A plus/minus individual such as yourself will be ready to point out Curry's superiority in unadjusted +/- but that is the result of McHale's coaching strategy. The Rockets used balanced lineups all year which makes it very difficult to stand out. I haven't seen adjusted plus/minus or RAPM (is that stuff still available online?) for this season and I still consider box score production to be more important in player evaluation than +/-.

As to the SAS-GSW series, the first two games to me were a case of SAS doing a terrible job on perimeter rotations leaving tons of empty jumpers along with Curry being hot. It is telling that Thompson was annihilating SAS from the outside during those first 2 games.

The final 4 games weren't that close. GSW one win was a game they were down in for a majority of the action. Once they tightened up on the perimeter they handled GSW with ease because they shut down their offense more than SAS offense improving.

when I engage in skill-set analysis I think the 2013 Harden would have improved more teams than Curry. Harden is elite at running the most important play right now in the NBA: pick and roll. He can generate a significant amount of high percentage shots for himself and his team playing basically anywhere in the league due to his skill in this area. By contrast, I do not consider Curry an elite ball handler nor elite off the ball player. I think there are teams in the league were his impact would be less.

A few more elite playoff games isn't going to cause me to discard the regular season 82 game sample size. This is why I am still on the Wade bandwagon and why I still prefer Harden.

If you put an emphasis on box score production, then I don't see why you have this substantial difference between the two in the regular season. Their box score production is really quite similar, and Curry looks more impressive in the playoffs by quite a bit. I like using ORating when I'm comparing lead guards, and Curry and Harden are pretty similar in the regular season (Harden at 116, Curry at 115), while Curry is at 113 in the playoffs and Harden is at 105.

Secondly, I think you're vastly underrating Curry's abilities as an on-ball and off-ball player, and his ability to run the PnR. His outside shooting, ball handling, passing, and off-ball game are better than Harden's. In fact, that's why Curry actually seems MUCH more portable than Harden: Curry can play off-ball and still retain his impact, because his outside shooting is always there. Harden seems to be ineffective when you take the ball out of his hands. If you paired them up with another ball-dominant player, I would think Curry looks better and would fit in more comfortably.

[SideshowBob]

Thanks for the responses. I was more interested in whether you knew the formula to calculate RAPM.

http://www.apbr.org/metrics/viewtopic.php?f=2&t=8232&start=30#p15849

That's for running it in R, (as well as plain 1YR APM, which I'd do first, as its just running a simple weighted OLS regression).

[sp6r=underrated]

Thank you for the responses. I appreciate it. Like I said I am not as big a believer in plus/minus as many people on this board but it does have value.

[Doctor MJ]

A plus/minus individual such as yourself will be ready to point out Curry's superiority in unadjusted +/- but that is the result of McHale's coaching strategy. The Rockets used balanced lineups all year which makes it very difficult to stand out. I haven't seen adjusted plus/minus or RAPM (is that stuff still available online?) for this season and I still consider box score production to be more important in player evaluation than +/-.

RAPM with Priors
Through March 30th:

Rank Name ORating DRating Rating Pos
5 James Harden 6.6 -0.9 5.6 5472
38 Stephen Curry 1.7 1.1 2.8 5326

2013 RAPM without any prior info:

104 Stephen Curry 0.9 1.5 -0.6 5326
134 James Harden 0.4 2.1 -1.7 5472

Appreciate that clarification. I saw the first set of numbers on your page and it was pretty hard to believe based on non-prior.

Looking at both sets of data, it's tough for me to really rely on the data. I've got a long history talking about the pros and cons of priors in this setting. More stat minded people than me tend to look at my objections as silly.

Clearly missing the prior makes it more noisy which is bad. Am I really going to give Harden a big vote of confidence based on something that totally disappears if you go only by what he did for his team this year?

[Doctor MJ]

If you put an emphasis on box score production, then I don't see why you have this substantial difference between the two in the regular season. Their box score production is really quite similar, and Curry looks more impressive in the playoffs by quite a bit. I like using ORating when I'm comparing lead guards, and Curry and Harden are pretty similar in the regular season (Harden at 116, Curry at 115), while Curry is at 113 in the playoffs and Harden is at 105.

Secondly, I think you're vastly underrating Curry's abilities as an on-ball and off-ball player, and his ability to run the PnR. His outside shooting, ball handling, passing, and off-ball game are better than Harden's. In fact, that's why Curry actually seems MUCH more portable than Harden: Curry can play off-ball and still retain his impact, because his outside shooting is always there. Harden seems to be ineffective when you take the ball out of his hands. If you paired them up with another ball-dominant player, I would think Curry looks better and would fit in more comfortably.

I'll focus on an aspect of portability: Scalability.

I think you all know I like Harden a good deal. I found him very impressive this year, and this was after him playing very different in OKC, also to great value. So there's no doubt he can port his game productively to a variety of settings, however if we're really talking about a top tier guy, part of what we're asking is how well Harden can port THIS primacy heavy game on to contenders.

Regardless of how he can fit himself in with other talents, if his current style of play can't scale well to a better team, then I do consider it's value to be fool's gold to some degree. What crossed my mind watching the team play toward the end of the year was that it just seemed like the Rockets would need to re-format their attack in future years, and that while Harden could probably learn to do amazing stuff in that new format, that's not the same as being able to literally transplant your game to the greater scale.

It's thus not a criticism of Harden as a talent but it's a wariness to champion his current play as too much of an ideal.

By contrast, Curry seems to be much closer to that transplant scalability. With his probably GOAT on-ball shooting ability and proven ability to go back & forth between the shooting focus and the passing focus, it's easy to imagine him just adjusting his balance a little bit here and there as other talent emerges next to him.

Getting back to the question about the Warriors performance in the playoffs: What the Spurs appear to have experienced this post-season is two absolute dominations of their opponents...and one series where they really had no easy answer. They got through it, but the weakness that Curry appeared to expose in their defense seemed pretty legit...and seemed like it could happen to pretty much any defense by virtue of it's simplicity: It's basically impossible to keep a point guard from getting a sliver of an opening for a shot on the outside, so if a guy can shoot that quick shot accurately enough, the defense will be in trouble.

But as I say all this, I'm really looking for rebuttals. As someone already mentioned in this thread, it feels naive to champion Curry too hard given that there were times where he seemed to disappear. I can say that he gets a Top 5 vote from me despite these times simply due to the weak competition, but before I do that I really want to hear all the negative arguments.

[therealbig3]

But as I say all this, I'm really looking for rebuttals. As someone already mentioned in this thread, it feels naive to champion Curry too hard given that there were times where he seemed to disappear. I can say that he gets a Top 5 vote from me despite these times simply due to the weak competition, but before I do that I really want to hear all the negative arguments.

So I'm clearly pro-Curry, and so just to respond to that possible rebuttal you're pointing out:

To hold Curry disappearing at times against him moreso than anyone else seems unfair, because outside of LeBron, and really, even including LeBron, there are plenty of blemishes and disappearing acts that we can hold against anyone.

Also, as praiseworthy as Mark Jackson's motivational efforts have been with this young team, he's also pretty oblivious to facts that stare him right in the face. David Lee imo has at least partly been exposed as an expendable piece who doesn't really have much impact as long as Curry simply takes on a bigger role. And in fact, the Warriors were playing better without him because they had to use defensive players more, while Curry kept their offense clicking almost by himself. But Mark Jackson decided to use David Lee out of nowhere for important minutes in the playoffs once he came back, and the other team would almost always go on a big run as a result. Also, Mark Jackson calls for Jarret Jack ISOs as a huge part of his offense, WHILE Curry is out there.

As good of a job as he did, that's probably the most unforgivable thing about Mark Jackson and the way he coached the Warriors: he clearly understands how good Curry is, just based on how he talks about him, so why are you taking the ball out of his hands so much and giving it to a far inferior player so much? Not only that, why are you calling so many plays FOR that inferior player at the expense of involving your star?

Curry had his own issues in the playoffs down the stretch, partly due to injury, also possibly due to fatigue, but I think part of his struggles and part of the Warriors struggles as a whole had quite a bit to do with Mark Jackson's inability to see the obvious. And when you look at Curry's struggles in context with the other candidates for 4 and 5, they really don't stand out imo. He had an overall great playoff run as far as I'm concerned.

[CBA]

A plus/minus individual such as yourself will be ready to point out Curry's superiority in unadjusted +/- but that is the result of McHale's coaching strategy. The Rockets used balanced lineups all year which makes it very difficult to stand out. I haven't seen adjusted plus/minus or RAPM (is that stuff still available online?) for this season and I still consider box score production to be more important in player evaluation than +/-.

RAPM with Priors
Through March 30th:

Rank Name ORating DRating Rating Pos
5 James Harden 6.6 -0.9 5.6 5472
38 Stephen Curry 1.7 1.1 2.8 5326

2013 RAPM without any prior info:

104 Stephen Curry 0.9 1.5 -0.6 5326
134 James Harden 0.4 2.1 -1.7 5472

Appreciate that clarification. I saw the first set of numbers on your page and it was pretty hard to believe based on non-prior.

Looking at both sets of data, it's tough for me to really rely on the data. I've got a long history talking about the pros and cons of priors in this setting. More stat minded people than me tend to look at my objections as silly.

Clearly missing the prior makes it more noisy which is bad. Am I really going to give Harden a big vote of confidence based on something that totally disappears if you go only by what he did for his team this year?

I'm confused by the wording of this post. Are you putting credence into the non-prior informed results?

[Doctor MJ]

I'm confused by the wording of this post. Are you putting credence into the non-prior informed results?

The only difference between the two sets of data when it comes to Harden's circumstance this year is that the prior informed model essentially gives him credit for what he did while playing an entirely different role for OKC. I'm not saying people should give the non-prior informed stat a ton of credence, but rather that in this circumstance I'd be cautious with the prior informed stat as well.

[Doctor MJ]

Tangent, a dumb thought, but I'm amused:

This award is basically the basketball equivalent to soccer's Ballon d'Or.

Ballon d'Or + Orange Ball = Ballon d'Orange

Whadya think?

[E-Balla]

Tangent, a dumb thought, but I'm amused:

This award is basically the basketball equivalent to soccer's Ballon d'Or.

Ballon d'Or + Orange Ball = Ballon d'Orange

Whadya think?

Good job sir. Good job.

[CBA]

I'm confused by the wording of this post. Are you putting credence into the non-prior informed results?

The only difference between the two sets of data when it comes to Harden's circumstance this year is that the prior informed model essentially gives him credit for what he did while playing an entirely different role for OKC. I'm not saying people should give the non-prior informed stat a ton of credence, but rather that in this circumstance I'd be cautious with the prior informed stat as well.

Ah okay that makes sense. It's just that NPI RAPM definitely isn't evidence of anything.

[sp6r=underrated]

Aren't you guys saying that RAPM has a real flaw when evaluating single season play in that prior seasons play such a large role in the results?

[colts18]

Aren't you guys saying that RAPM has a real flaw when evaluating single season play in that prior seasons play such a large role in the results?

You have to factor that the RAPM I posted was thru march 30 so it's incomplete. The prior data is there to stabilize the results

[Doctor MJ]

Aren't you guys saying that RAPM has a real flaw when evaluating single season play in that prior seasons play such a large role in the results?

I probably sounded to harsh there. The bottom line is that a prior is in general a really good idea, and if the prior & non-prior disagree significantly in general you'd put more stock in the prior, but in Harden's case I don't think it makes sense to do that.

[sp6r=underrated]

It needs to be said that for all of Wade's struggles, his poor play isn't a big reason for why Miami is doing poorly. The Heat, while they struggled on offense throughout game 6, have played about as well as possible on offense. Their offensive rating games 1-6 is 111.6 which would have been good enough for 3rd in the league against an elite defense.

The Heat are struggling due to the structural flaws inherent on their team. They're bad at rebounding and lack true size.

Indiana isn't hanging in the series due to over-performance in the backcourt which would likely be tied to Wade's poor play/injury. They are winning because they are annihilating Miami on the offensive glass. Bosh and the Heat's front-court lack the size and strength to prevent this from happening.

These flaws are why they are not a GOAT level team and why their season long point differential do not look like that of a GOAT level team and have never looked like a GOAT level team during the 3 years.

They are obviously very good. They likely will win this series and the title but their reputation is above their level of play. Defensive rebounding matters.