Layups: Draft Pick Value Research
Posted by Neil Paine on May 19, 2010
Over at APBRmetrics, longtime BBR reader/commenter DSMok1 has been putting together a great thread about the expected value of a draft pick, expanding on a post Justin made last year. The graphs are really fascinating, especially with regard to the possible range of production you might get at each draft slot, as well as the odds of getting a good player by pick#, and the chart of the surplus values of each pick#. With the lottery being held last night and the draft quickly approaching, that thread is a must read.
May 20th, 2010 at 9:27 am
Thanks for the shoutout, Neil. I thought you had done the research, not Justin. I changed my citation over at APBRmetrics.
May 20th, 2010 at 9:49 am
Good call, I'm sure Justin appreciates it. I loved the graph w/ the Likelihood of Pick Results, btw -- after the top 10 picks, you have less than a 50-50 shot at getting a solid player, which is sobering because a late lottery pick always sounds like a slot in which you would get a decent player most of the time, but that's not the case. If you don't have a top-3 pick, you're probably not getting a star; if you don't have a top-10 pick, you're probably not even getting a solid player.
May 20th, 2010 at 10:25 am
I'm going to try to create a smoothed version of that chart, because it is so pertinent. Unfortunately, it is exponential in 2 directions...
May 20th, 2010 at 11:17 am
That was my thought exactly, it's going to be a pain to smooth that thing.
May 20th, 2010 at 7:50 pm
My first effort at smoothing that is up on the APBR thread. I used exponentials in both directions and multiplied together and quadratics in both directions and multiplied together. And a min value of 0. In other words, a lot of coefficients.
May 20th, 2010 at 10:58 pm
Neil Paine Says:
"Good call, I'm sure Justin appreciates it. I loved the graph w/ the Likelihood of Pick Results, btw -- after the top 10 picks, you have less than a 50-50 shot at getting a solid player, which is sobering because a late lottery pick always sounds like a slot in which you would get a decent player most of the time, but that's not the case. If you don't have a top-3 pick, you're probably not getting a star; if you don't have a top-10 pick, you're probably not even getting a solid player."
If you had started drafting in 1969 with just the guys in the 2nd round or after (so guys who should have been available to everyone), by 1976 you could have fielded a team of
Bob Dandridge
Alex English
Tiny Archibald
George Gervin, and
Dan Issel
May 20th, 2010 at 11:48 pm
That's why I said "probably", not "certainly"... But the odds against getting that haul would be astronomical. Picking each of those would be like finding a needle in a haystack of Cliff Shegoggs and Narvis Andersons.
May 21st, 2010 at 11:01 am
The odds of any given second rounder turning out to be a star (20+WSoRP) over the player's first 4 years is about 1.25%. The 31st pick has about a 3% chance; 45th pick is below 1.5%, and below 52nd has basically no chance.
May 22nd, 2010 at 6:54 pm
The odds of drafting and signing Gervin and Issel after the ABA already had them is pretty slim.