Comments on: Who Are the Most Similar Players to This Year’s All-Stars?

By: Bob M.

Bob M. — Mon, 21 Feb 2011 02:45:20 +0000

Carmelo's comparisons are not that impressive. Do you all think his low win share numbers over his career are accurate or misleading? i don't watch the nuggets that often personally to know. If they're accurate, i'd stay away from him I were the Knicks or Nets, especially the Nets. they're giving up just about everything but Brook Lopez in their offer.

By: Brian

Brian — Mon, 21 Feb 2011 00:06:41 +0000

More valid comparisons (based on style)

Amare Stoudamire -- Shawn Kemp, Spencer Haywood
Kevin Love -- Wes Unseld, Jerry Lucas, Dolph Schayes (?)
Kevin Durant -- George Gervin
Tim Duncan -- Elvin Hayes
Kobe Bryant -- Michael Jordan (do we have to wait for Kobe to hit 30,000 points before we can make this comparison)

By: Nick

Nick — Sun, 20 Feb 2011 20:36:03 +0000

How similar is LeBron to the closest match? I wouldn't think it would be close at all, much like Jordan.

By: Willie

Willie — Sun, 20 Feb 2011 05:25:06 +0000

Off topic but my opinion on the current ELO rankings:

[ELO Most Overrated]
Tim Duncan
Steve Nash
Dennis Rodman
Bernard King
Bill Walton
Sam Bowie (WTF?)

[ELO Most Underrated]
Bill russell
Julius Erving
George Mikan (Below Domonique?)

By: DWarner

DWarner — Sun, 20 Feb 2011 04:45:31 +0000

Interesting and hilarious! Gotta love Amare's comps! Good laugh...

By: AHL

AHL — Sat, 19 Feb 2011 22:03:16 +0000

I think that's why systems such as Basketball Prospectus's SHOENE were developed.

By: Neil Paine

Neil Paine — Sat, 19 Feb 2011 18:25:07 +0000

#1 - Right, I suppose it's not "totally useless"... but it's also isn't really finding similar players in the traditional Bill James sense, and I wanted to convey that it's probably less useful than a system which does that. #3 - First, I took every player-season in NBA/ABA history and pro-rated their OWS/DWS to 82 team games. Then, for a given pair of players at the same position, I calculated the squared difference between their prorated OWS & DWS at each age, and summed those differences across their entire career. The historical player with the smallest squared differences is the one who had the most similar career arc in terms of shape and quality. Here's Michael Jordan, for instance:

			WS by Age
Player	Pos	Type	21	22	23	24	25	26	27	28	29	30	31	32	33	34	35	36	37	38	39
Michael Jordan	SG	OWS	10.3	1.0	11.9	15.2	14.6	14.7	14.9	12.1	12.0	0.0	1.2	14.2	13.3	10.4	0.0	0.0	0.0	1.2	2.9
		DWS	3.7	0.5	5.0	6.1	5.2	4.3	5.4	5.6	5.2	0.0	1.1	6.2	5.0	5.4	0.0	0.0	0.0	2.1	3.4

Then you try to match up the SGs with the most similar career paths (in order from most to least similar): Clyde Drexler Ray Allen Reggie Miller George Gervin Kobe Bryant Eddie Jones Joe Dumars Hal Greer Sidney Moncrief Jeff Hornacek Vince Carter Jason Terry Dwyane Wade Hersey Hawkins Walter Davis It's not trying to find players who were necessarily stylistically similar to MJ (other than being SGs & having a similar OWS/DWS split); instead it just wants to find players who were roughly as good at him at the same ages. And obviously in the case of someone like Jordan, even the best comparison is going to be weak, because he was such a major outlier as a player.

By: JTaylor21

JTaylor21 — Sat, 19 Feb 2011 16:57:42 +0000

Can someone explain to me how this comparisons work?

By: Ed

Ed — Sat, 19 Feb 2011 16:37:32 +0000

Love the Dirk Nowitzki/Charles Barkley comp!!!

By: Mike Goodman

Mike Goodman — Sat, 19 Feb 2011 14:02:29 +0000

Neil, maybe you shouldn't file this under 'Totally Useless', as it could evolve into something useful.
Maybe even something like - http://sonicscentral.com/apbrmetrics/viewtopic.php?t=2721
Good luck!