RealGM Top 100 LIST- 2014

[Texas Chuck]

In 03 Nash absolutely was a top 5 PG in the game. He was in 02 and 04 too.

[ThaRegul8r]

Re: Recency bias. An understandable fear. I think Duncan and his placement is the most telling thing to look at.

Which was why I said in my prediction about Duncan's place on this list:

Duncan will rise, benefiting from the drop of Magic, Wilt and Bird, and also from the fact that his fifth title win is fresh in everyone's minds.

This list came at the perfect time for him.

to be fair I have been championing Duncan on this board since I registered. Before either 13 or 14 happened. Some of us thought extremely highly of Duncan as a player all along.

I'm on record as saying—on numerous occasions—that Duncan's my favorite active player. I've seen his entire career. I was glad to see him win this year after the frustration of last year, which I erroneously thought was likely his last opportunity to win another title. So your telling a perceived Duncan naysayer that "some of us thought extremely highly of Duncan as a player all along" falls flat.

If you can't see that this voting—coming on the heels of Duncan's fifth title (which places him behind only Russell, Jordan and Kareem as far as superstars go) in which he was their best player through the playoffs overall, which itself comes on the heels of a season in which he was a legitimate Defensive Player of the Year candidate and would have been Finals MVP had the Spurs won, came at the perfect time, then I don't know what to say.

Guess Lebron is only high because of 4 straight Finals trips(oh wait maybe that stuff matters a little).

Shaq is only high because of his visible presence on TNT.

Mike is high because of commercials and he owns a team

Russell is only high because a trophy got named after him.

Magic is high because 3 guys saw movies in his theatres last night.

Rodman will soon get a big bump thanks to his recent Korean diplomacy

Quite frankly, stuff like this makes it hard for me to take you seriously. This is exactly how not to make a point. From reading your posts, you have a continued tendency to get emotional at times. You'll want to rein it in in order to have serious discussions. I have to filter emotions out of some posters' posts I read in order to get to content of substance.

[Texas Chuck]

Quite frankly, stuff like this makes it hard for me to take you seriously. This is exactly how not to make a point. From reading your posts, you have a continued tendency to get emotional at times. You'll want to rein it in in order to have serious discussions. I have to filter emotions out of some posters' posts I read in order to get to content of substance.

Thanks for the advice.

[ThaRegul8r]

deleted due to a misunderstanding on my part.

[G35]

edite: sigh and now the Dirk was only thought highly of last time because of the title.

Keep in mind I'm not saying Dirk's placement was wrong! It's just that I make note of narrative shifts.

In 2008 Dirk ranked 41st in this project BEHIND LeBron.
In 2011 Dirk ranked 17th in the project AHEAD OF LeBron.

There's no plausible way to argue that Dirk should have surpass LeBron based on those 3 years, but he did. No way it would happen if not for the 2011 finals.

Thing is, I think the later ranking makes much more sense than the earlier one. Dirk's reputation suffered mightily as a result of his performance in the '06 finals and '07 playoffs. Too much. The '11 results helped correct that overreaction, and I don't think acknowledging this cheapens what Dirk did.

Cmon its obvious rankign Lebron ahead of Dirk in 2008 was the greater wrong. But Dirk did develop his post game starting in 08 so its not like there wasnt a significant growth in Dirk's game and not just the title.

Yes I think it was wrong when posters boil things down to net production on the court for who is greater. Then try and denigrate rings as a poor or second rate way of judging a players career or rating. Then the recency bias is huge, as is the perception bias e.g. Jordan is best ever without question, the "Immortal Six", one peak year trumps a career...etc, etc.

I disagree with some of Regul8trs feelings on how to judge players but that's natural, however he was dead on about many poster's biases and how it would reflect in the the voting. You could predict that Lebron would jump into the top 10 based on those two title runs, Duncan would also be seen more favorably because of the past two seasons of success, Kobe would drop because of the past 3 years of poor Laker success, and the RAPM warriors KG, Nash will all be represented well EXCEPT for David Robinson who will be the exception to the rule.

It's well documented that the statistics aren't the problem...it's the individual poster and how the decide to interpret and apply them to individual players.....

[Doctor MJ]

edite: sigh and now the Dirk was only thought highly of last time because of the title.

Keep in mind I'm not saying Dirk's placement was wrong! It's just that I make note of narrative shifts.

In 2008 Dirk ranked 41st in this project BEHIND LeBron.
In 2011 Dirk ranked 17th in the project AHEAD OF LeBron.

There's no plausible way to argue that Dirk should have surpass LeBron based on those 3 years, but he did. No way it would happen if not for the 2011 finals.

Thing is, I think the later ranking makes much more sense than the earlier one. Dirk's reputation suffered mightily as a result of his performance in the '06 finals and '07 playoffs. Too much. The '11 results helped correct that overreaction, and I don't think acknowledging this cheapens what Dirk did.

Cmon its obvious rankign Lebron ahead of Dirk in 2008 was the greater wrong. But Dirk did develop his post game starting in 08 so its not like there wasnt a significant growth in Dirk's game and not just the title.

The first sentence is precisely what I'm saying.
Second sentence isn't wrong either.

[Gifted Mind]

This is clearly looking like a BOAT list rather than GOAT

[JeepCSC]

Speaking of Ben Wallace on a different thread, where does RAPM place him? I'd imagine he'd do well on that, even in spite of his offensive, well, issues.

[Doctor MJ]

Speaking of Ben Wallace on a different thread, where does RAPM place him? I'd imagine he'd do well on that, even in spite of his offensive, well, issues.

From my spreadsheet, scaled numbers to try approximate actual points of impact, his peak year was +5.69 and his best-5-years average was 4.73. These aren't jaw dropping numbers but they are certainly nice. By comparison, in terms of 5 years peaks, Garnett's is over 10 and Duncan's is over 9.

If you just look at defense, Ben's peak is +6.80 and his 5-year is 5.47, where both Garnett & Duncan's 5-years are in the 6s giving them a small edge there.

So basically, Ben's good enough to be DPOY (in '03-04 I believe), though RAPM would not have given him 4 DPOYs...and that weak offense really makes it so there's a huge gap between him and the top interior defenders of his time that were great on both sides of the ball.

Does that help you? Anything else you're curious about?

[penbeast0]

We have 4 posters who have requested to participate -- ShaqAttack, lukekarts, Ryoga Hibiki, and FJS . . . I have to go back and check to see if they have been active in two threads (I know at least 2 of them qualify). I will try to do it today. For now, go ahead and post/vote/etc. as if you were active -- after all, that's how you get qualified to be active.

Going back and reviewing the last 3 threads (and that's a lot of posts so I could have made mistakes), I find that ShaqAttack, FJS, and to a lesser extent lukekarts have been active in at least 2. Ryoga, despite my having seen his posts for years here, only seems to have been active in the one thread where he first asked to join. Therefore, I will declare the first 3 posters to be eligible for the current thread (#12) and if Ryoga Hibiki becomes consistently active or . . . if I missed good, information rich posts by him in this project, hopefully someone will PM me links to those posts and I will move him to the active list as well.

[HurricaneKid]

Speaking of Ben Wallace on a different thread, where does RAPM place him? I'd imagine he'd do well on that, even in spite of his offensive, well, issues.

From my spreadsheet, scaled numbers to try approximate actual points of impact, his peak year was +5.69 and his best-5-years average was 4.73. These aren't jaw dropping numbers but they are certainly nice. By comparison, in terms of 5 years peaks, Garnett's is over 10 and Duncan's is over 9.

If you just look at defense, Ben's peak is +6.80 and his 5-year is 5.47, where both Garnett & Duncan's 5-years are in the 6s giving them a small edge there.

So basically, Ben's good enough to be DPOY (in '03-04 I believe), though RAPM would not have given him 4 DPOYs...and that weak offense really makes it so there's a huge gap between him and the top interior defenders of his time that were great on both sides of the ball.

Does that help you? Anything else you're curious about?

I believe the most widely used model is:
http://stats-for-the-nba.appspot.com/

Which is FAR more kind to Wallace. From 01-06 he was top 5 in RAPM all 5 seasons, peaking at #1 with a 9.1. Mighty impressive run really.

And he ran away with the Def RAPM every year during that spell.

[nonjokegetter]

Highly entertaining process to check in on when it's a slow day at work, I'll tell you boys that much.

[Doctor MJ]

Speaking of Ben Wallace on a different thread, where does RAPM place him? I'd imagine he'd do well on that, even in spite of his offensive, well, issues.

From my spreadsheet, scaled numbers to try approximate actual points of impact, his peak year was +5.69 and his best-5-years average was 4.73. These aren't jaw dropping numbers but they are certainly nice. By comparison, in terms of 5 years peaks, Garnett's is over 10 and Duncan's is over 9.

If you just look at defense, Ben's peak is +6.80 and his 5-year is 5.47, where both Garnett & Duncan's 5-years are in the 6s giving them a small edge there.

So basically, Ben's good enough to be DPOY (in '03-04 I believe), though RAPM would not have given him 4 DPOYs...and that weak offense really makes it so there's a huge gap between him and the top interior defenders of his time that were great on both sides of the ball.

Does that help you? Anything else you're curious about?

I believe the most widely used model is:
http://stats-for-the-nba.appspot.com/

Which is FAR more kind to Wallace. From 01-06 he was top 5 in RAPM all 5 seasons, peaking at #1 with a 9.1. Mighty impressive run really.

And he ran away with the Def RAPM every year during that spell.

That's actually the site I tell people not to use.

It's the most used because it used to have legit RAPM data, and was the only site to have it at one point (the data I'm using from those years is actually from his site).

It no longer does.
The '00s data uses XRAPM, which also factors in box score.
The '90s data uses something the author literally calls Fake RAPM, which is nothing more than a statistical +/-.

All the data on there is legit to be clear, it's just not RAPM. In discussion he'll use the proper labels, but he keeps simply calling it "RAPM" on his site.

[Reservoirdawgs]

All the data on there is legit to be clear, it's just not RAPM. In discussion he'll use the proper labels, but he keeps simply calling it "RAPM" on his site.

That's what confused me about RAPM when I first started looking into it...I googled "JE RAPM NBA" and that's what came up. I saw it was JE so I started using that data, only for people to tell me that wasn't actually JE's RAPM data. It was very confusing and I had to do more work than I initially thought (including perusing the APBR message boards) just so I could get a better understanding.

Do you plan on updating your spreadsheet to include the 2013 and 2014 seasons?

[HurricaneKid]

From my spreadsheet, scaled numbers to try approximate actual points of impact, his peak year was +5.69 and his best-5-years average was 4.73. These aren't jaw dropping numbers but they are certainly nice. By comparison, in terms of 5 years peaks, Garnett's is over 10 and Duncan's is over 9.

If you just look at defense, Ben's peak is +6.80 and his 5-year is 5.47, where both Garnett & Duncan's 5-years are in the 6s giving them a small edge there.

So basically, Ben's good enough to be DPOY (in '03-04 I believe), though RAPM would not have given him 4 DPOYs...and that weak offense really makes it so there's a huge gap between him and the top interior defenders of his time that were great on both sides of the ball.

Does that help you? Anything else you're curious about?

I believe the most widely used model is:
http://stats-for-the-nba.appspot.com/

Which is FAR more kind to Wallace. From 01-06 he was top 5 in RAPM all 5 seasons, peaking at #1 with a 9.1. Mighty impressive run really.

And he ran away with the Def RAPM every year during that spell.

That's actually the site I tell people not to use.

It's the most used because it used to have legit RAPM data, and was the only site to have it at one point (the data I'm using from those years is actually from his site).

It no longer does.
The '00s data uses XRAPM, which also factors in box score.
The '90s data uses something the author literally calls Fake RAPM, which is nothing more than a statistical +/-.

All the data on there is legit to be clear, it's just not RAPM. In discussion he'll use the proper labels, but he keeps simply calling it "RAPM" on his site.

I would love to see a superior data set.

Got a link?

[Doctor MJ]

All the data on there is legit to be clear, it's just not RAPM. In discussion he'll use the proper labels, but he keeps simply calling it "RAPM" on his site.

That's what confused me about RAPM when I first started looking into it...I googled "JE RAPM NBA" and that's what came up. I saw it was JE so I started using that data, only for people to tell me that wasn't actually JE's RAPM data. It was very confusing and I had to do more work than I initially thought (including perusing the APBR message boards) just so I could get a better understanding.

Do you plan on updating your spreadsheet to include the 2013 and 2014 seasons?

Cool.

The delay in the last two seasons is due to the JE issue. The data from 2002 to 2012 comes from JE...and then he changed over to XRAPM. It took a while for anyone else to pick up the slack, and now there are several of them. If I can become convinced that one of them is clearly the best implementation, I'll not only update but see about using that guy's studies as far back as they go in the past.

[Doctor MJ]

I would love to see a superior data set.

Got a link?

Here's my spreadsheet which takes us up to 2012:

https://docs.google.com/spreadsheet/ccc ... _web#gid=3

The 2002 to 2012 data is from JE before he switched to his new stuff. The "SD" studies normalize for the variance in each data set, the "Scaled" studies then my best estimate of the "right" variance to estimate actual points of impact.

In terms of people making real RAPM now:

http://www.gotbuckets.com/statistics/rapm/2014-rapm/

http://shutupandjam.net/nba-ncaa-stats/pi-rapm/

[lorak]

http://shutupandjam.net/nba-ncaa-stats/pi-rapm/

What is that exactly? Because sometimes it's basically like JE's "real" RAPM (for example 2001 and 2002 NPI) and sometimes quite different.

[Doctor MJ]

http://shutupandjam.net/nba-ncaa-stats/pi-rapm/

What is that exactly? Because sometimes it's basically like JE's "real" RAPM (for example 2001 and 2002 NPI) and sometimes quite different.

If you're asking how his methods differ, I don't know. This is something I'm hoping to talk to these guys about actually.

[JeepCSC]

I hope Dirk rides the RAPM rocket past Kobe.