Menu

Predicting NBA Draft Success and Failure through Historical Trends

Predicting NBA Draft Success and Failure through Historical Trends
Jun 21, 2006, 10:42 pm

Paul Gearan
Senior Consultant
Rexer Analytics
www.rexeranalytics.com


The surfeit of discussion about how teams should best utilize their relative position in the draft begs the question: what do historical trends tell us about the probability of successful outcomes given even broad parameters? Obviously it still comes down to the evaluation of the individual, but even broadly, are there types of players who tend to do better than others? What is the wisdom of moving down and for what purpose? What have been some of the biggest errors made and who has made them? What can an analysis of player outcomes tell us about these and other issues?

Method of Analysis:

Five hundred seventy-six (576) players drafted in the first round from 1980 to 2001 (22 drafts) were assessed for their “5-year outcomes” (i.e. what kind of player were they by their fifth year in the league). Five years was selected because it was a long enough time frame to allow for a reasonable evaluation of most players. A handful of players became bigger stars later – Steve Nash and Jermaine O’Neal for example – but for most the 5-year mark was a solid indicator on long-term NBA performance.

The 2001 draft was selected as the end point, of course, because that is the last draft with a 5-year outcome. Starting at 1980 was fairly arbitrary but based on it being the beginning of a decade, post-NBA-ABA contraction in 1977, and after the insertion of Bird and Magic into the league in 1979, which heralded the modern age of basketball in this writer’s opinion. Draft years such as 1966 when there were only 10 teams in the league or 1976 when there were 17 (and after several years of some of the best players going to the ABA) just did not have similar enough contexts to the current era. Certainly, with spread of early-entry among college players and the inclusion of high school and international players, the flavor of the draft has changed considerably over those 22 drafts as well, but even given these substantial changes this time frame had enough stability in league structure to justify these parameters.

Every player’s NBA career was coded - by this writer who acknowledges the subjective nature of that process, but suspects except for minor categorization differences here and there, most raters would generate similar findings - at the point of five years after they were drafted using seven primary categories which were then collapsed into the following four categories:

I. Out of the league (category 1) & Barely in league/End of bench guy (2)
II. Solid bench player (3) & Top bench player/Marginal starter (4)
III. Solid starter (5)
IV. Star: All star caliber player (6) & Superstar (top 10-15 player in the league) (7)


Player positions were distilled into these three categories:

I. Point guards (and combo guards)
II. Wing players (2s & 3s)
III. Big Men (4s & 5s)


Players were also categorized by whether they were an international or directly out of high school player; however, as there were only 22 international players and 13 high school kids drafted in those years, there are not enough to make generalizations and therefore are not included in the following analysis.

Given that the draft grew from 23 to 29 from 1980 to 2001, relative draft position was controlled for by roughly dividing each year’s first round into “quarters.” This also helped level off the effect of the influx of foreign players which grew the talent and for the increased number of opportunities created by the increased number of teams.

Initial Findings:

Finding I: Yeah, it’s pretty much where you draft. Just to get the obvious point out of the way: if you draft in the top quarter of the draft your odds are much better in finding an impact player or at least a solid starter. How much better? The chart below displays the dramatic drop in the rates of star players and the rise of marginal players as the first round proceeds.

1395





If a player is selected among the top 7 picks in an average year, the odds are about 70% he will be a solid starter or better by his fifth year (and of course, substantially higher odds [86%] if picked in the first two spots as opposed to lower in the first quarter). From 8-15, those odds go down to 43%, and in the lower half of the round the odds go under 25%. In fact, you’re doing well if you get a guy who is even a solid bench contributor or a marginal starter. About half the guys drafted in the second half of the first round, are out of or barely hanging in the league five years out.

Finding II: Point guards are drafted less, but stay around more: 255 big men were drafted in the first round over the 22 years, but after 5 years, 79 were out, or nearly out, of the NBA (31%), leaving 176 contributors. Only 98 point (or combo) guards were drafted (well less than half the 4s and 5s), but only 19 were out or nearly out of the league (19%), leaving 79 contributors. Wing players are drafted slightly less frequently than big men, but have similar rates of success and failure. Do not wait too long for those point guards though, 12 of the 19 who were out/nearly out of the league were in the bottom quarter of the first round.

Finding III: Ergo, guess who the biggest busts tend to be? Those big guys. Robert Traylor, Doug Smith, Bill Garnett, James Ray, Chris Washburn, Russell Cross, Sharone Wright, William Bedford and Melvin Turpin: all drafted between 3 and 6 in the first round, all out of basketball (or barely hanging on) in 5 years. Special mention to Michael Olowokandi as perhaps the most disappointing number one pick in that span (Kwame Brown is making a charge though). Danny Ferry and Stromile Swift take the prize as the worst #2 picks. Lesson: big guys are risky even at the top of the draft. For further evidence, of the 77 big men drafted in the top quarter of the round, 36% have been marginal starters or worse, 30% have become solid starters, and 34% have become stars. Of the 17 point guards drafted in the top quarter, none have been complete busts, 24% have become solid bench players/marginal starters, 29% have become solid starters, and 47% have become stars. Of the 41 wing players drafted in the top quarter of the round, 22% have been marginal starters or worse, 44% have become solid starters, and 34% have become stars. So if you think a point guard is worth a top pick, he probably is. Drafting big men is like playing Russian Roulette for GMs.

Finding IV: Good and great wing players can be found in the second quarter of the first round: Exactly half of the 62 wing players selected in the second quarter of the first round go on to be stars (26%; not that much lower than the 34% in the top quarter) or solid starters (24%). Star point guards (8%) are hard to find at this stage, but solid starters are not (46%). Less than a third of big men become stars (7%) or solid starters (25%). So if you are Golden State sitting with the 9th pick and Sheldon Williams, Patrick O’Bryant and Rodney Carney are there, history says go with Carney. But if you want a solid starting point guard (not necessarily a star) then Marcus Williams, Randy Foye or Rajon Rondo get into the picture. In the third quarter (from 16-22), stick with point guards, where 50% (13 of 26) have been solid starters or stars, compared with only 22% of wing players and only 15% of big men. The last quarter of the first round? Best of luck there.

Finding V: Worst drafting team over those 22 years is, by far, the Atlanta Hawks. Of the 23 first round picks they have had, only 1, Pau Gasol, became a star and they immediately traded his rights away (along with Brevin Knight and Lorenzen Wright) for Shareef Abdur-Rahim. They have drafted only 3 solid starters (Jason Terry, Stacey Augmon, and Kevin Willis) and the three picks they had in the top quarter of the first round were Al Wood, Jon Koncak, and Demarr Johnson (Ouch!).

Finding VI: There is no clear cut best drafting team. Golden State has drafted the most star players (7), but that is mostly due to the large number of high draft picks they have had. Four of the seven players Charlotte drafted in the top half of the round became stars, but they traded away the rights to a pretty important one right away: Kobe Bryant. Seattle has made very few mistakes (only 2 of the 20 players they drafted were out of the league after 5 years), and have done well in the second half of the first round with one star player (Shawn Kemp) and 5 solid starters (by far the most of any team’s second half selections), but other than Gary Payton and Kemp (and for a brief time Xavier McDaniel) theirs is a legacy of solid, but not great players.

Finding VII: The worst draft years were 1986 and 2000: Sadly the death of Len Bias gives this an asterisk, but the 1986 draft produced only one star (Brad Daugherty) and, after five years, 46% of the picks were out of the league or just hanging on, a high over the 20 years. Even among the 7 guys who became solid starters there are some long-range complications (e.g., Roy Tarpley). The draft of 2000 is the only other draft to produce just one star player (Kenyon Martin – and he was only narrowly categorized as such), and is filled with a lot of mediocrity even among the 8 solid starters (Primo Brezec, Mike Miller, Joel Pryzbilla, Jamal Crawford, etc.).

Finding VIII: Best Draft Year: How about doing some time travel and trading and top pick in 2006 pick for one in 1981? Would any of these guys look good to GMs right now: Isiah Thomas, Mark Aguirre, Larry Nance, Tom Chambers, Herb Williams, Rolando Blackman, Orlando Woolridge, Buck Williams, and Kelly Tripucka? Only 3 players in the first round of he 1981 NBA draft were not on the league 5 years later. Other strong drafts include:

1985 (Karl Malone, Chris Mullin, Patrick Ewing, Terry Porter, Joe Dumars, Detlef Shrempf, Xavier McDaniel);
1987 (David Robinson, Scottie Pippen, Reggie Lewis, Reggie Miler, Kevin Johnson and a number of other solid starters) ;
1996 (Allen Iverson, Kobe Bryant, Peja, Shareef, Stephon Marbury, Ray Allen, Antoine Walker and then later bloomers Steve Nash, Jermaine O’Neal, & Marcus Camby); and
1999 (Ron Artest, Kirilenko, Shawn Marion, Steve Francis, Wally Sczerbiak, Elton Brand, Baron Davis, Andre Miller, Corey Maggette, Richard Hamilton, Jason Terry, Lamar Odom).

The prediction of future NBA success is related to factors that go well beyond position played and even draft slot to some degree. Athletic ability, specific basketball skills and the temperament of the individual are being pored over by GMs across the league. Some GMs are better able to integrate this information than others, some get lucky. However, to the fan without this level of insight into players, getting emotionally attached to a center with “upside” who is drafted 14th is likely to leave you sorely disappointed. An even riskier bet is being certain that the “jewel” at the bottom of the first round, being lauded by your team’s GM and TV analysts alike, will be the next Tony Parker. It happens, but not very often.

Recent articles

7.0 Points
1.5 Rebounds
3.0 Assists
11.7 PER
-->
14.3 Points
8.4 Rebounds
2.1 Assists
20.2 PER
-->
12.3 Points
6.5 Rebounds
2.7 Assists
13.9 PER
-->
9.0 Points
8.8 Rebounds
0.2 Assists
10.0 PER
-->
4.1 Points
2.5 Rebounds
0.4 Assists
8.4 PER
-->
13.0 Points
7.0 Rebounds
0.0 Assists
8.6 PER
-->
2.0 Points
3.0 Rebounds
0.0 Assists
-5.6 PER
-->
1.6 Points
0.6 Rebounds
0.0 Assists
3.0 PER
-->
1.7 Points
2.0 Rebounds
0.2 Assists
4.6 PER
-->
0.0 Points
0.0 Rebounds
0.0 Assists
0.0 PER
-->
1.9 Points
1.2 Rebounds
0.3 Assists
5.1 PER
-->
22.1 Points
11.8 Rebounds
1.6 Assists
28.6 PER
-->
6.6 Points
1.4 Rebounds
2.4 Assists
8.6 PER
-->
5.7 Points
4.7 Rebounds
0.3 Assists
14.7 PER
-->
7.5 Points
1.8 Rebounds
0.3 Assists
3.3 PER
-->
27.1 Points
8.9 Rebounds
4.4 Assists
21.7 PER
-->
5.2 Points
2.2 Rebounds
2.0 Assists
7.3 PER
-->
3.1 Points
2.8 Rebounds
3.7 Assists
6.9 PER
-->
8.3 Points
6.3 Rebounds
1.8 Assists
31.7 PER
-->
2.4 Points
1.2 Rebounds
2.6 Assists
9.6 PER
-->
1.4 Points
1.5 Rebounds
0.2 Assists
4.0 PER
-->
1.7 Points
1.7 Rebounds
0.7 Assists
5.5 PER
-->
2.0 Points
0.0 Rebounds
1.0 Assists
26.3 PER
-->
2.0 Points
1.5 Rebounds
0.6 Assists
7.6 PER
-->
2.4 Points
1.6 Rebounds
0.2 Assists
6.4 PER
-->
6.6 Points
1.7 Rebounds
0.6 Assists
13.2 PER
-->
3.0 Points
4.1 Rebounds
0.8 Assists
8.2 PER
-->
17.6 Points
3.7 Rebounds
2.8 Assists
14.9 PER
-->
6.8 Points
5.7 Rebounds
0.7 Assists
11.7 PER
-->
5.3 Points
1.9 Rebounds
3.0 Assists
9.0 PER
-->
1.3 Points
1.6 Rebounds
0.5 Assists
4.4 PER
-->
17.0 Points
10.2 Rebounds
3.0 Assists
18.4 PER
-->
8.9 Points
5.6 Rebounds
0.4 Assists
12.5 PER
-->
1.8 Points
1.7 Rebounds
0.5 Assists
10.1 PER
-->
1.4 Points
2.0 Rebounds
1.1 Assists
7.8 PER
-->
0.0 Points
0.0 Rebounds
0.0 Assists
0.0 PER
-->
14.8 Points
2.7 Rebounds
6.9 Assists
15.3 PER
-->
10.6 Points
3.0 Rebounds
2.7 Assists
13.1 PER
-->
6.0 Points
2.0 Rebounds
0.0 Assists
32.5 PER
-->
1.7 Points
1.0 Rebounds
0.0 Assists
12.2 PER
-->
15.0 Points
2.4 Rebounds
0.8 Assists
15.3 PER
-->
17.6 Points
5.9 Rebounds
0.9 Assists
18.4 PER
-->
4.9 Points
4.5 Rebounds
0.5 Assists
12.3 PER
-->
7.0 Points
2.3 Rebounds
2.1 Assists
13.6 PER
-->
13.2 Points
8.7 Rebounds
3.9 Assists
17.7 PER
-->
5.8 Points
2.1 Rebounds
1.0 Assists
10.0 PER
-->
6.0 Points
4.0 Rebounds
0.5 Assists
14.4 PER
-->
5.5 Points
2.3 Rebounds
2.8 Assists
14.0 PER
-->
11.3 Points
1.8 Rebounds
3.5 Assists
15.0 PER
-->
8.5 Points
7.9 Rebounds
1.0 Assists
17.8 PER
-->
5.9 Points
3.0 Rebounds
2.2 Assists
12.4 PER
-->
20.8 Points
4.3 Rebounds
3.7 Assists
17.3 PER
-->
6.7 Points
2.7 Rebounds
4.0 Assists
17.4 PER
-->
9.3 Points
3.0 Rebounds
3.0 Assists
14.5 PER
-->
15.7 Points
3.6 Rebounds
4.7 Assists
15.4 PER
-->
9.6 Points
2.8 Rebounds
2.0 Assists
12.8 PER
-->
9.5 Points
4.9 Rebounds
2.3 Assists
7.7 PER
-->
0.0 Points
0.0 Rebounds
0.0 Assists
0.0 PER
-->
2.2 Points
0.8 Rebounds
0.4 Assists
5.9 PER
-->
5.2 Points
4.2 Rebounds
1.0 Assists
10.1 PER
-->
0.5 Points
0.8 Rebounds
0.0 Assists
-14.7 PER
-->
4.1 Points
3.7 Rebounds
1.1 Assists
11.4 PER
-->
3.4 Points
0.9 Rebounds
2.2 Assists
13.7 PER
-->
6.5 Points
3.5 Rebounds
2.0 Assists
1.2 PER
-->
9.8 Points
1.7 Rebounds
2.4 Assists
10.6 PER
-->
0.0 Points
2.0 Rebounds
0.0 Assists
-8.0 PER
-->
2.5 Points
0.8 Rebounds
3.8 Assists
5.7 PER
-->

Twitter @DraftExpress

DraftExpress Shop