PDA

View Full Version : Ranking of GM's draft picks



Major Cold
08-12-2007, 10:22 AM
The link is to another message board. If this is not appropiate then delete thread. It is interesting even if we do not agree with assessment. Converstation piece beyond JO....

http://sonicscentral.com/apbrmetrics/viewtopic.php?t=1456
The table is messed up someone can work their magic do so.


Here is a system I have developed from evaluating the drafting of 17 GM's. Please give me some feedback.

Here's my GM rating (best to worst):
+4.1 - Rod Thorn of New Jersey
+4.0 - Geoff Petrie of Sacramento
+3.9 - Randy Pfund of Miami
+3.8 - John Paxson of Chicago
+3.1 - Bryan Colangelo of Phoenix
+3.0 - Don and Donnie Nelson of Dallas
+2.6 - Mitch Kupchak of the LA Lakers
+1.6 - Indiana
+1.1 - Rick Sund of Seattle
-0.4 - Kevin McHale of Minnesota
-0.7 - RC Buford of San Antonio
-1.4 - Joe Dumar of Detroit
-2.1 - Carroll Dawson of Houston
-3.6 - Jerry West of Memphis
-5.2 - Kevin O'Connor of Utah
-8.3 - Elgin Baylor of LA Clippers

I expect for each pick in the draft order to net a certain level of talent. If a GM drafts better talent with lottery picks than another GM does with late first round picks, that GM isn't necessarily doing a better job. So, I defined what level of talent I expect with each pick, then I will give GM credit (positive points) or blame (negative points) if he the player he drafts turns out better or worse than I expected. Here is the scoring chart:


----------------------------------------------------------------------
| | | | | | Deep | Never | never
Pick |All-Star| Star |Starter |Rotation| Bench | Bench | in NBA |
======|========|========|========|========|======= =|========|========|
1-2 | 1 | 0 | -1 | -2 | -3 | -4 | -5 |
3-4 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 | -3.5 | -4.5 |
5-8 | 2 | 1 | 0 | -1 | -2 | -3 | -4 |
9-12 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 | -3.5 |
13-16 | 3 | 2 | 1 | 0 | -1 | -2 | -3 |
17-20 | 3.5 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 | -2.5 |
21-24 | 4 | 3 | 2 | 1 | 0 | -1 | -2 |
25-28 | 4.5 | 3.5 | 2.5 | 1.5 | 0.5 | -0.5 | -1.5 |
29-32 | 5 | 4 | 3 | 2 | 1 | 0 | -0.6 |
33-40 | 5.5 | 4.5 | 3.5 | 2.5 | 1.5 | 0.5 | -0.2 |
41+ | 6 | 5 | 4 | 3 | 2 | 1 | 0 |
----------------------------------------------------------------------

All-Star - Selected as an All-Star
Star - Not an All-Star, but one of the team's two best players
Starter - Not a star, but starts when not injured
Rotation - Plays when not injured, gets 15 - 24 mpg
Bench - Plays 10 - 15 mpg, occasionally has a DNP-CD
Deep Bench - Gets erratic minutes, will generally play in the NBA three seasons or less
Never in NBA - Not made it on to to his team's roster or played just a handful of gamesMy assessment is based upon the player's current year unless he has played more than 4 seasons in the NBA, in which case it is based upon his fourth year. What matters is how the player did on the team that drafted him. The fact that George Lynch went on to be a starter on other teams doesn't change the fact that he didn't do well as a Laker draft pick. If a team trades a player, I assess his value in the trade (which will be based mostly on his performance to date) and then don't evaluate him further. If a team lets a draft pick walk, the draft pick gets a "Deep Bench" rating for the rest of his career. If a player who walked made the rotation for a season for his team before walking, I am going to give the GM a +0.5 bonus. The fourth year cut off is because most players are what they are after their fourth season and the chance of trade, career-ending injury, being beat out by a newer draft pick, etc. greatly increases after that.

Second round picks that make the bench for a year or two don't really matter for a team's fortunes. Also, a second round pick can get playing time one season because of injuries, shortness of bench, etc. and that shouldn't inflate a GM's score. To reflect that, here are some special rules for players picked 33 or later:
1. If the player plays only one season and then is waived, the GM gets a 0.1
2. If the player never played more than 500 minutes, then the GM gets a 0.1
3. If the player is at least a Bench player during his fourth season, the GM gets the full score
4. If the player is at least a Rotation player for two seasons, then GM gets the full score
5. If the player is at least a Starter player, then GM gets the full score
6. If the player doesn't meet #3, #4 or #5, then the GM gets 0.3 if the player is a Bench player and 0.5 if the player is a Rotation player

Alternatives considered
Assessing players for years 2-4 of their career
I tried this but realized it didn't work after I did Atlanta to Indiana. This alternative is considering the years 2-4 of a draft picks career, giving more weight to the most recent years. A draft pick's second year would have a weight of 1, third year a weight of 2, and fourth year a weight of 3. Then, the GM would be credited the weighted rating for the player. For most players, this didn't move the rating that much. For some players, this alternative didn't work. What to do about Eddie Griffin, the #7 pick who was a rotation player his second season, and then released after the season? What to do about Mehmet Okur, who was a rotation player his second season and then left Detroit through free agency? Isn't Orien Greene, who played 15.4 mpg in 80 games his rookie year before being cut a better late second round pick than someone who never played a game for the team?

Rating players how they did regardless of their team
Mehmet Okur was an absolute steal at #38. However, Detroit didn't recognize that and let him go as a free agent. Does his stellar play in Utah redeem the pick? Some think yes, but I don't

Adjusting scores for C's, PF' and PG's
Centers are extremely hard to draft. SG's and SF's are relatively easily. I played with the idea of adding a point for Centers that earn at least a Bench rating and 0.5 points for PF's and PG's that make at least a Bench rating. However, I don't know the positions of all NBA players and didn't know what to do with PF/C's and combo guards.

Seeing how each pick rated out of the available talent
For each draft pick, assess all of the players taken after that pick and how well the player ranked relative to the other players. For example, Chris Duhon would have better a pick that Sasha, but Sasha was the second best option. To do this would require evaluate and rank every player drafted for 4 years, which is more work than I want to do.

Known Flaws to This System
No adjustment for depth in draft
The 2003 draft has produced 5 All-Stars so far, so drafting one then was much than in the 1997 draft, which produced 2.

GM's of bad, young teams are overrated
Philidelphia, Boston and Atlanta score very high in these rankings because those teams have gotten rid of most of their veteran players and are playing most of their draft picks regardless if their are NBA quality or not. For example, of the 12 players who played the most minutes for Boston last season, 7 were drafted in the last 3 years and one more was drafted the year before. As a consequnce, players like Salim Stoudamire, Ryan Gomes Green and Willie Green get large positive ratings when they aren't NBA quality players. I have moved the rankings for those three teams out of the overall rankings.

Tenures to short to evaluate
Bernie Bickerstaff of Charlotte (3 seasons)
Danny Ferry of Cleveland (2 seasons)
Mark Warkentien of Denver (1 season)
Chris Mullin of Golden State (3 seasons)
Larry Harris of Milwaukee (3 seasons)
Jeff Bower of New Orleans (2 seasons)
Isiah Thomas of New York (3 seasons)
Otis Smith of Orlando (2 seasons)
Kevin Pritchard of Portland (New)
Bryan Colangelo of Toronto (1 season)
Ernie Grunfeld of Washington (3 seasons)

Ratings pulled because they are for bad, young teams
+12.0 - Billy King of Philadephia
+10.8 - Danny Ainge of Boston
+3.7 - Billy Knight of Atlanta

Billy Knight's picks (Atlanta) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Boris Diaw (21) - Rotation then traded => 1
Travis Hansen (37) - Never in NBA (-0.2)
2004
Josh Childress (6) - Rotation (-1)
Josh Smith (17) - Star (2.5)
Donta Smith (34) - Deep Bench (0.1)
Royal Ivey (37) - Bench (0.3)
Viktor Sanikidze (42) - Never in NBA (0)
2005
Marvin Williams (2) - Starter (-1)
Salim Stoudamire (31) - Rotation (2)
Cenk Ayol (59) - Never in NBA (0)
Overall Score
+3.7

Danny Ainge's picks (Boston) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Marcus Banks (13) - Bench then traded => -1
Kendrick Perkins (27) - Rotation (1.5)
Brandon Hunter (56) - Deep Bench (0.1)
2004
Al Jefferson (15) - Star (2)
Delonte West (24) - Starter then traded => 2
Tony Allen (25) - Rotation (1.5)
Justin Reed (40) - Deep Bench then traded => 0.1
2005
Gerald Green (18) - Rotation (0.5)
Ryan Gomes (50) - Starter (4)
Orien Greene (53) - Deep Bench => 0.1
Overall Score
+10.8

John Paxson's picks (Chicago) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
Kirk Hinrich (7) - Starter (0)
Mario Austin (36) - Never played in NBA (-0.2)
Tommy Smith (53) - Never played in NBA (0)
2004
Ben Gordon (3) - Star (0.5)
Luol Deng (7) - Star (1)
Chris Duhon (39) - Rotation (2.5)
2005
No picks
Overall Score
+3.8

Don and Donnie Nelson's picks (Dallas) and my scoring of them:
2001
Kyle Hill (43) - Never in NBA (0)
2002
Mladen Sekularac (54) - Never in NBA (0)
2003
Josh Howard (29) - All-Star (5)
2004
Pavel Podkolzin (21) - Never in NBA (-2) Note: Played 1 game
2005
No picks
Overall Score
+3.0

Joe Dumar's picks (Detroit) and my scoring of them:
2001
Rodney White (9) - Bench then traded => -1.5
Mehmet Okur (38) - Rotation then lost to FA => 1.0
2002
Tayshaun Prince (23) - Starter (2)
2003
Darko Milicic (2) - Deep Bench then traded => -4
Carlos Delfino (25) - Not signed for a year, Rotation (0.5)
Andreas Glyniadakis (58) - Never in NBA (0)
2004
Rickey Paulding (54) - Never in NBA (0)
2005
Jason Maxiell (26) - Bench (0.5)
Amir Johnson (56) - Deep Bench (0.1)
Alex Acker (60) - Deep Bench (0)
Overall Score
-1.4

Carroll Dawson's picks (Houston) and my scoring of them:
Note: Daryl Morey is now GM for Houston
2001
Eddie Griffin (7) - Rotation then released => -2.5
2002
Yao Ming (1) - All-Star (0.5)
Bostjan Nachbar (15) - Bench then traded => -1
Tito Maddox (38) - Deep Bench (-0.1)
2003
Malick Badiane (44) - Never in NBA (0) => 0
2004
Vassilis Spanoulis (50) - Took until '06-'07 to sign
2005
Luther Head (24) - Rotation (1)
Overall Score
-2.1

Indiana Pacer's picks and my scoring of them:
2001
Jamison Brewer (40) - Deep Bench (0.1)
2002
Fred Jones (14) - Rotation (0)
2003
James Jones (49) - Rotation then traded for 2nd round pick => 1.5
2004
David Harrison (29) - Deep Bench (0)
Rashad Wright (59) - Never in NBA (0)
Overall Score
+1.6

Elgin Baylor's picks (Clippers) and my scoring of them:
2001
Traded away draft rights to Tyson Chandler
2002
Chris Wilcox (8) - Bench then traded for Vladimir Radmanovic => -2
Melvin Ely (12) - Deep Bench then traded => -2.5
Traded away rights to Mario Kasun (41)
2003
Chris Kaman (6) - Starter (0)
Sofoklis Schortsanitis (34) - Never in NBA (-0.2)
2004
Shaun Livingston (4) - Injury prone all career. Estimate: Rotation (-1.5)
Lionel Chalmers (33) - Deep Bench (0.1)
2005
Yaroslav Korolev (12) - Deep Bench (-2.5)
Daniel Ewing (32) - Bench (0.3)
Overall Score
-8.3

Mitch Kupchak's picks (Lakers) and my scoring of them:
2001
No picks
2002
Kareem Rush (20) - Traded for 2 high second round picks => -1
2003
Brian Cook (24) - Bench (0)
Luke Walton (32) - Starter (3)
2004
Sasha Vujacic (27) - Bench (0.5)
Marcus Douthit (56) - Never in NBA (0)
2005
Andrew Bynum (10) - Rotation (-0.5)
Ronny Turiaf (37) - Rotation (0.5)
Von Wafer (39) - Deep Bench (0.1)
Overall Score
+2.6

Jerry West's picks (Memphis) and my scoring of them:
Note: Chris Wallace is now the GM of Memphis
2001
Wasn't GM yet
2002
Drew Gooden (4) - Difficult to evaluate because traded mid-rookie season in a multi-player deal. Estimate: Starter (-0.5)
Robert Archibald (32) - Deep Bench (0)
2003
Troy Bell (16) - Deep Bench (-2) Note: Acquired draft day trade
Dahntay Jones (20) - Deep Bench (-1.5) Note: Acquired draft day trade
2004
Andre Emmett (35) - Never in NBA (-0.2) Note: Acquired draft day trade and played 8 games
Antonio Burks (36) - Deep Bench (0.1) Note: Acquired draft day trade
Segei Lishouk (49) - Never in NBA (0)
2005
Hakim Warrick (19) - Rotation (0.5)
Overall Score
-3.6

Randy Pfund's picks (Miami) and my scoring of them:
2001
Ken Johnson (49) - Deep Bench (0.1)
2002
Caron Butler (10) - Difficult to evaluate because traded twice. Estimate: Starter (0.5)
Rasual Butler (53) - Bench then traded => 2
2003
Dwayne Wade (5) - All-Star (2)
Jerome Beasley (33) - Never in NBA (-0.2) Note: Actually played 2 games
2004
Dorell Wright (19) - Bench (-0.5)
Albert Miralles (39) - Never in NBA (0) Note: Acquired draft day trade and rights were later traded
Matt Freije (53) - Never in NBA (0) Note: Was waived by Heat before playing
2005
Wayne Simien (29) - Deep Bench (0)
Overall Score
+3.9

Kevin McHale's picks (Minnesota) and my scoring of them:
2001
No draft picks
2002
Marcus Taylor (52) - Never in NBA (0)
2003
Ndudi Ebi (26) - Deep Bench (-0.5)
Rick Rickert (55) - Never in NBA (0)
2004
Blake Stepp (58) - Never in NBA (0)
2005
Rashad McCants (14) - Rotation, had microfracture knee surgery => 0
Bracey Wright (47) - Deep Bench (0.1)
Overall Score
-0.4

Rod Thorn's picks (New Jersey) and my scoring of them:
2001
Richard Jefferson (13) - Star (2)
Jason Collins (18) - Starter (1.5)
Brandon Armstrong (23) - Deep Bench (-1)
Brian Scalabrine (35) - Bench (1.5)
2002
Nenad Krstic (24) - Starter (2)
Tamar Slay (54) - Deep Bench (0.1)
2003
Zoran Planinic (22) - Deep Bench (-1)
Traded rights to Kyle Korver (51)
2004
Traded rights to Viktor Khryapa (22)
Christian Drejer (51) - Never in NBA (0)
2005
Antoine Wright (15) - Bench (-1)
Mile Ilic (43) - Not signed until '06-'07 season
Overall Score
+4.1

Billy King's picks (Philadephia) and my scoring of them:
2001
Samuel Dalembert (26) - Starter (2.5)
Damone Brown (37) - Deep Bench (0.1)
Alvin Jones (57) - Deep Bench (0.1)
2002
Traded rights to Jiri Welsch
John Salmons (26) - Rotation (1.5)
Sam Clancy (44) - Never in NBA (0)
Randy Holcomb (56) - Never in NBA (0)
2003
Willie Green (41) - Rotation (3)
Traded rights to Paccelis Morlende (50)
Kyle Korver (51) - Rotation (3)
2004
Andre Iguodala (9) - Star (1.5)
2005
Louis Williams (45) - Bench (0.3)
Overall Score
+12.0

Bryan Colangelo's picks (Phoenix) and my scoring of them:
Note: Colangelo is now the GM of Toronto
2001
Alton Ford (51) - Deep Bench (0.1)
2002
Amare Stoudemire (9) - All-Star (2.5)
Casey Jacobsen (22) - Rotation then traded => 1
2003
Zarko Cabarkapa (17) - Deep Bench then traded for 2 second round picks => -1
Leandrinho Barbosa (28) - Rotation (1.5) Note: Acquired draft day trade
2004
Jackson Vroman (31) - Deep Bench then traded => 0
2005
Dijon Thompson (24) - Deep Bench (-1)
Overall Score
+3.1

Geoff Petrie's picks (Sacramento) and my scoring of them:
2001
Gerald Wallace (25) - Deep Bench then selected in expansion draft => -0.5
Maurice Jeffers (55) - Never in NBA (0)
2002
Corsley Edwards (58) - Never in NBA (0) Note: played 10 games for Hornets in '04-'05 season
2003
None
2004
Kevin Martin (26) - Star (3.5)
Ricky Minard (47) - Never in NBA (0)
2005
Francisco Garcia (23) - Rotation (1)
Overall Score
+4.0

RC Buford's picks (San Antonio) and my scoring of them:
2001
Wasn't GM yet
2002
Wasn't GM yet
2003
No Draft Picks
2004
Beno Udrih (28) - Bench (0.5)
Romain Sato (52) - Never in NBA (0)
Sergei Karaulov (57) - Never in NBA (0)
2005
Ian Mahinmi (28) - Never in NBA (-1.2)
Overall Score
-0.7

Rick Sund's picks (Seattle) and my scoring of them:
2001
Vladimir Radmanovic (12) - Rotation (-0.5)
Earl Watson (40) - Bench then let go as FA => 0.1
2002
Peter Fehse (49) - Never in NBA (0)
2003
Nick Collison (12) - Starter (0.5)
Luke Ridnour (14) - Starter (1)
Paccelis Morlende (50) - Never in NBA (0) Note: Acquired draft day trade
2004
Robert Swift (12) - Bench (-1.5)
David Young (41) - Never in NBA (0)
[2005]
Johan Petro (25) - Rotation (1.5)
Mickael Gelabale (48) - Didn't play until '06-'07 season
Overall Score
+1.1

Kevin O'Connor's picks (Utah) and my scoring of them:
2001
Raul Lopez (24) - Two years to sign, Deep Bench, then traded => -1
Jarron Collins (53) - Rotation (3)
2002
Curtis Borchardt (18) - Deep Bench then traded => -1.5 Note: Acquired draft day trade
2003
Aleksandar Pavlovic (19) - Bench then taken in expansion draft => -1.5
Maurice Williams (47) - Bench then lost to FA => 0.1
2004
Kris Humphries (14) - Deep Bench then traded => -2
Kirk Snyder (16) - Deep Bench then traded => -2
2005
Deron Williams (3) - Starter (-0.5)
CJ Miles (34) - Deep Bench (0.1)
Robert Whaley (51) - Deep Bench then traded => 0.1
Overall Score
-5.2

GrangerRanger
08-12-2007, 12:15 PM
If that guy gave 0.1 on Deep Bench for Jamison Brewer, how couldn't he give at least 2.0 on Fred Jones. I always thought Freddy was better then James Jones. :mad:

JayRedd
08-12-2007, 12:42 PM
If that guy gave 0.1 on Deep Bench for Jamison Brewer, how couldn't he give at least 2.0 on Fred Jones. I always thought Freddy was better then James Jones. :mad:

Fred was 14th overall, a place where this guy considers getting a "rotation" level talent a non-achievement.

Jamison was 40th overall, a Second Round pick where just getting a "deep bench" level guy is somewhat of a positive.


All in all, this is a good analysis. Of course there is gonna be some debate on which players fall into which category (for example, someone brought up on the APBR site that Leandro Barbosa is clearly better than Ryan Gomes, but Ainge gets more points because Gomes is a starter and Leandro only fits the "Rotation" category here). And I also think you'd have to do a historic statistical analysis of what the mean category would be at each pick rather than just arbitrarilly assigning the point values.

But the whole undertaking is great in theory and I would love to see someone develop it more fully/mathmatically and go back a lot further in years.

Naptown_Seth
08-12-2007, 01:56 PM
This has already been done, but not applied to individual GMs. A couple of studies (which I've linked to here when we discussed the draft...check the "tank" thread(s)) used an EXPECTED vs BECAME analysis to discuss risk/reward of different points in the draft or by player position.

In those cases they typically use something like the Hollinger player rater rather than just a subjective "AS/starter/rotation/deep bench" thing.

And while he does attempt to correct some odd results for bad teams, the fault to me seems to lie more in something like not weighting "starter/rotation" for a GOOD TEAM as higher than for a BAD TEAM. See Salim vs Josh Smith vs Marvin Williams. Of course Salim is playing, Atlanta has nothing. Why do they have nothing? Because Knight used his #2 pick on Williams.

So giving more reward for drafting a rotation player at 31 than the penalty for ONLY drafting a starter at #2, thus helping to make room for a late first/early 2nd pick in your rotation makes no sense. You do better than Childress and Williams at 6 and 2 and you aren't using your #31 pick quite as much.


Better to perhaps assemble teams in groups of 10, like first team all-NBA (with rotation of 10), 2nd team, etc. Still lets you be subjective in your evaluation, but helps offset the errors that pure stats induce by not discounting stats compiled on a bad team. But it forces the rankings to be more comparitive, hopefully leveling some of the good team vs bad team stuff.

Did Knight draft a rotation player for the 30th best all-NBA team, the 15th or the 5th with pick 31. Plus, what caliber player did other GMs get in that range of picks, take the average over several years and for a window of about 2-3 picks either direction.


These things will never be perfect because we haven't defined explict, quantifiable traits or outcomes that establish what a "good" player is. Does he pass a lot, do assists = smart/great passing on a 1 to 1 ratio, is scoring more important than shooting efficiency, do you measure impact by how well his teammates shoot with and without him on the floor....it goes on forever.



And when people say "I don't like stats, I just use my opinion and awareness of the game", all they are doing is the exact same thing I just described but hide their formula in their head, perhaps even from themselves. Your mind is processing stats (because these are just outcome based facts compiled together) when you decide a player is better than another.

"Tougher rebounder" is just you remembering moments when you defined something as a tough rebound due to the circumstances and then saw that player come out with the ball. That stuff could be spelled out and compiled as a stat if only you thought about it and verbalized what makes it "tough" to you, but people cringe at doing so. Maybe it's the effort, the difficulty of defining it in words, maybe its a hidden fear of accountability.



What this means is that I actually have no problem with these kinds of studies, of personally comparing players "subjectively". I just also support the effort to bring out the explicit definitions and objectivity that lie behind that subjectivity, also known as "looking at stats and formulas". :D