Statistical Analysis on Potential Stardom of NBA Draftees

Now that the NBA draft is over, fans and analysts are pouring over team selections. Lotto teams are thinking they have their next star. Contenders are thinking they have the next complementary pieces to boost their team. But it won't be a few years until we know the next busts and steals of this draft.

In the meantime, I wanted to do some quick analysis on the draft and the players in the NBA.

There are only 23 players drafted since 2006 (the 1st draft that restricted the drafting of players who just graduated from high school) who have been selected to more than 1 All-NBA team. To put that into perspective, 840 players have been drafted since (although 60 would have been from the most recent draft). So fewer than 3% of all players drafted since 2006 are multiple time All-NBA selections (fewer than 2 per draft).

The following are the list of All-NBA players to turn the numbers into basketball terms.

As you can see, there are a few 1st time selections that the majority of NBA fans would never include as stars and who might be better categorized as role players. In addition, all players with more than 1 selection have been offered max contracts, while that's not necessarily the case for 1 time selections. Of course, these are not necessarily the most perfect definitions of what constitutes a star, but I wanted something quantitative and this'll do for now.

But how does one identify a potential star during draft time? Obviously play is a big deal. But I was wondering if the NCAAB All-NBA equivalent would be a good indicator. And it turns out, it's not. Of the 20 multi-time All-NBA players who played in college (3 did not - Marc Gasol, Giannis Antetokounmpo, and Rudy Gobert), only 11 were All-America selections in college.

What's interesting is that the majority of the 9 players who were never All-Americas in college is that they all played for major programs. And even Stephen Curry (Davidson), Kawhi Leonard (San Diego St), and Damian Lillard (Weber St) were noticed by the All-Americas selection committee.

But what about All-America selections who never turned out to be stars? Well rather than divide players solely into 2 categories of stars and non-stars, I created 4 other categories:

1. Solid Contributor: career average of >= 10 MPG, >0 BPM, >0 VORP
2. Rotation Player: career average of >= 10 MPG, >0 VORP
3. NBA Washout: everyone else with >0 G
4. Never Played in NBA: everyone with G=0

Excluding the most recent draftees, there were 152 college All-Americas players. Eleven (7%) became NBA stars. 35 (23%) became solid contributors, 37 (24%) became rotation players, 64 (42%) became NBA washouts, and 5 (3%) never played in the NBA. Nearly half of all college All-Americas players become nobodies after college.

I took it a step further and checked how many NBA stars were high school All-Americans (either McDonald's or Parade) and the percentages were actually the same. Only 11 of the 20 "stars" were selected for any All-American high school honors.

And what about the other players that were high school All-Americans? Excluding the most recent draftees, 243 were high school All-Americans. Eleven (5%) became NBA stars, 51 (21%) became solid contributors, 54 (22%) became rotation players, 117 (48%) became NBA washouts, and 10 (4%) never played in the NBA despite being drafted. Overall, high school All-American awards seem just as predictive as college All-America awards.

But what if we applied combinations of college and high school all-American selections? And to provide some sense of comparison, I also calculated how draftees split in general.

Based on this analysis, it does seem sensible that the greatest percentage of NBA stars come from the draftees that contains both high school and college stars (10% vs 2%/5%/2%).

What's interesting is that between draftees that had only 1 of either college or high school All-American awards, you're more likely to find either NBA stars and solid contributors from the draftees of only high school All-Americans (42% vs 12%). And even more damning against college basketball success is that a greater percentage of draftees who had no awards in either college or high school become NBA stars or solid contributors when compared to the percentage of draftees who had just college All-American status (15% vs 12%).

Overall, this seems to indicate:

• A draftee's college All-American status is a positive sign for NBA teams only if he was a high school All-American as well as that draftee pool has the greatest percentage of stars.
• If the prospect was a college All-American but not a high school one, then that might actually be a negative sign for NBA teams as the draftee pool of non-college All-Americans as a whole (high school All-Americans + non-All-Americans) has a better chance of finding solid contributors or better (20% vs 12%).

For this year's draft, that would be great news for Zion Williamson, R.J. Barrett, and Kyle Guy. All 3 were both All-Americans in high school and college.

Darius Garland, Coby White, Cam Reddish, Romeo Langford, Darius Bazley, Nassir Little, Keldon Johnson, Bol Bol, and Jaylen Hands were all high school All-Americans, but not college ones. They might fare better in the pros than Ja Morant, Jarrett Culver, Rui Hachimura, Brandon Clarke, Grant Williams, and Carsen Edwards - all of whom where college All-Americans but not high school ones.

This raises a few questions:

Why does college All-American status without high school All-American status seem to have a negative impact on NBA contribution? One would think that if a player played well in college, even if he were not a star in high school, that would be at minimum equal to if not a better indicator than players who were never stars at all or only high school stars!

Is it possible that NBA basketball is more similar to high school basketball, with more emphasis on athleticism, natural talent, and/or unequal teams?

Are college basketball programs or coaches just able to overperform with their players? If so, why? Is it because they are managing their players more as opposed to NBA teams who provide more freedom?

Does the fact that NBA money come into play make college stars perform worse (or non-college stars better)?

---

Regardless, in order to tie this back to the NBA draft, I wanted to consider draft position: how well do higher draft picks correlate with future NBA contributions?

Overall, pretty good I'd say. Your chance of finding stars and solid contributors drops the later you pick (dramatically), so much so that it seems picks 56–60 are pretty much worthless.

But what's interesting about this chart is that is seems draft position (and consequently scouting and GM decisions) is quite good and may be an even better predictor of future contributions.

So I tried to see the draft positions of our college and high school stars.

And despite there being an overwhelming number of NCAA stars, they never once outnumber by percentage the number of high school stars selected in the same draft positions.

Unsurprisingly, the 1st pick is almost always a NCAA + HS star and >50% of all NCAA + HS stars were selected by the 11th pick.

What's interesting though is that this chart looks strikingly similar to that of the NBA stars draft percentage chart.

Overall, it would seem that whatever methods teams and GMs use to draft, they are on average pretty accurate (or perhaps it is just easier to draft NBA players than expected).

What do you think? Do these numbers surprise you? What other data points would have been interesting to capture?