Draft Theory: On Patience and Probability

Unknown probabilities are a finicky thing. But to approach the NBA Draft without an understanding of probabilities is to fail right from the start. There are no 100% or 0% probabilities when it comes to the NBA Draft. Using words like “guaranteed” or “always” or “never” should be completely refrained from.

The tricky thing about unknown probabilities are that it becomes increasingly difficult whether a certain occurrence was likely to happen and did or was unlikely to happen but did. That is a massive distinction. A player with a 90% chance to hit can miss, and a player with a 10% chance to miss can hit. Does that mean the initial evaluation was wrong? The chances of those two things happening at the same time is 1 in 100, or 1%. That may sound small, but a 1% chance should happen once every two years in the NBA Draft. That doesn’t sound so small, right? #1 overall pick Michael Beasley failed. #60 overall pick Isaiah Thomas succeeded. These things were exceedingly unlikely, but they happened.

Danger of Instant Gratification

The NBA is a results-based league, and NBA Draft projection is a results-based analysis. Let’s be real – nobody cares why you were right or wrong, only if you were right or wrong. Which means we have to deal with one absolutely massive problem right off the bat: NBA fans (and teams) declare players to be busts or stars right away because there has to be an immediate right or wrong, but NBA player development doesn’t work that way.

As explained yesterday, player development isn’t linear. A player could be terrible at age 21, a bench player at age 22, and a starter at age 23. Or they could be mediocre at age 20, mediocre at age 21, mediocre at age 22, and great at age 23. Judging players too early leads to a lot of bad judgments. While this is understood to a large extent in the MLB and NHL, where the average fan doesn’t even pay attention to the draft or prospects until they are potential call-ups, it is completely lost in the NBA. And it’s not just a fan problem – it’s a team problem as well. The league is littered with good or even great players who were discarded by their initial team because they weren’t good enough fast enough.

If a team drafts a player under the belief that he has an 80% chance to pan out, they need to stick with him long enough to make sure he has actually failed. If they are not willing to stick it out, they should not draft the player. Yet teams do this every single year.

Two Prime Examples

Two players that most recently demonstrate this are D’Angelo Russell and Bruno Caboclo. Russell was a playable player at age 19, which is always a great sign for overall upside. By 20, he was a good backup. And then he was traded as a salary dump because he wasn’t a superstar yet. Russell developed into an actual all-star this year at age 22. Which is what was expected when he was drafted #2 overall! Why draft a player #2 overall if you’re not willing to see him out through at least his rookie contract? Shockingly, Russell isn’t considered a lock to get a max contract this offseason. A 22 year old lead guard who is already an all-star and has a ton of growth left in him, and people still aren’t buying him because…he wasn’t good immediately as a super young player? I don’t get it.

Bruno Caboclo is a different version of the same story. Caboclo was famously “two years away from being two years away” when he was drafted, and everybody laughed about it because it was a funny joke. Only it wasn’t a funny joke, it was a statement of reality. Despite being drafted in 2014, Caboclo is less than five months older than 2019 Draft prospect Cameron Johnson. So it turns out the super young Caboclo was five years away instead of four years away, but if you draft a guy who you know has upside but could take a long time to reach it, why are you abandoning him right as he’s starting to hit the period where he might pan out? Caboclo will likely be starting for the Grizzlies next season and could be in line for a very nice contract when he hits free agency thereafter.

Importance of Probability

And this brings us back to probability, and why it is so important. Almost every player drafted is younger than 23, and most are younger than 22. The top players and many international players are 20 or younger. That means that for many players drafted, teams will not know if they are good or not for years after they draft them. While being patient is of critical importance, of equal importance is not wasting a roster spot, time, and money on a player who is unlikely to develop into anything of value.

Waiting on a player who has a 10% chance to develop into a nice backup is a waste. Waiting on a player who has an 80% chance to develop into a star is necessary. What about a player who has a 40% chance to be a starter, 40% chance to be a bench player, and a 20% chance to wash out? Or the player who has a 50% chance to be a star and a 50% chance to bust? And how do we know these probabilities to begin with? And is a player who is “safe” better than a player who has a chance to hit a higher ceiling? At what point do the probabilities tilt in favor of one or the other?

This quickly turns into a tangled mess. How it manifests is something like “there are 10 guys who each have a 10% chance to hit.” None of them are good prospects, because 10% is low. But one of them will hit, or sometimes two or maybe even three, or maybe they’ll all miss. And as it’s a results-based business, the larger response isn’t “wow, a low percentage player hit,” it’s “why did you not know that this player was going to hit?”

So Many Unknowns

As stated in the first article of this series, this unpredictability is maddening. Watching this year’s playoffs, it’s easy to ask how everybody missed on players like Draymond Green, Pascal Siakam, and Malcolm Brogdon. And while different people could say different things, the most reasonable explanation is that those guys hit their top 1% outcomes. Again, once every two years, a player will hit their top 1% outcome. You can drive yourself mad trying to figure out what links all of them together, but the answer is frustratingly probably nothing at all.

So in the draft, you’re dealing with unknown probabilities, with unknown outcomes, that won’t be known for sometimes four or more years, and you are judged only on actual outcomes, often in far less than four years. This is pure masochism.

What’s the point of all this? It’s twofold.

First, it’s a really long plea to everybody to stop judging players so soon in their careers. NBA prospects are much more akin to MLB prospects than NFL prospects, who all come out after their junior or senior college seasons. If you believe in a prospect when he’s a 19 year old, believe in him when he’s 21 or 22 and hasn’t quite gotten over the hump, and stick with him long enough to allow him to do that.

Second, it’s a reminder that nobody is going to project the draft completely correctly, and once you get outside the top few prospects who have reasonably high probabilities, there’s just a lot of educated guesswork. If a player has a 20% hit chance and hits, a lot of people are going to look stupid based on the result. But over the long term, a good process should lead to better results. The goal of draft analysis should be to find processes that result in overall better accuracy. Not good accuracy, because that may not be possible, but better accuracy.

Previously: On Age, Development, and Value | Next up: There Are Levels To This Game