Explore/Exploit

The genius of Victor Wembanyama

Victor Wembanyama is the NBA’s first eight-footer. Not literally, he’s listed at 7’4”. But you see what I mean.

Wemby next to Lebron

Unlike many of the other tallest players in the league’s history, however, he’s not awkward or gangly. He moves fluidly and athletically. He has a good handle and strong passing instincts. He can shoot from anywhere on the court. A true unicorn.

Yet what is really mind-boggling about Wemby is that his mental gifts are even more impressive than the physical ones. Last year, in an interview, he said, “I’ve been told never to skip steps my whole life, but it didn’t stop me from running up the stairs.”

Imagine saying that as a 20-year-old. English isn’t even his first language!

Wemby was drafted by the Spurs in 2023. It was a tremendous stroke of luck for the organization: the order is determined randomly (although you have a better chance of picking highly the more games you lost) and whichever team got the number one spot would have taken Wemby. Everyone knew he was a generational, perhaps all-time talent.

The Spurs are a highly respected organization, but they did not have a good team at the time. As such, Wemby spent the first two years of his NBA career not playing for much, competitively. During these years he spent a lot of time around the perimeter of the court launching threes, and not much inside catching alley-oops. This frustrated a lot of the commentators on the NBA podcasts I listen to.

Well, I’m pretty sure Victor Wembanyama knows he can dunk. He is 7’4”, after all. In fact, I’m quite certain Wemby knew exactly what he was doing. He was exploring.

Let’s take a step back. In machine learning, there’s a dynamic called the explore-exploit dilemma. It’s about striking a balance between extracting value from what you already know, vs. learning more about your environment. That might sound confusing, but an example will make it clear.

Suppose you’re going out to dinner. You could try a new place (explore) or go to a place you already know you like (exploit). Most likely, the new place wouldn’t turn out to be as good as one of your existing favorites, but it could be, and then you’d have another great option next time you go out.

It’s not an easy choice. In fact, there is no known mathematical solution to the explore-exploit dilemma. But it seems to make sense to explore more earlier in the process, when you know less and have more runway in front of you; and exploit as you reach the end, when you know more and don’t have much time left anyway.

In those first two seasons, Wemby was in an explore phase. Even if he went all-out to win, he couldn’t have accomplished anything; the rest of the team wasn’t good enough. So instead he probed the limits of his abilities, trying different shots, moves, strategies. By staying out of the paint he also preserved his body – Wemby is undoubtedly very strong, but at 7’4”, you wouldn’t call him stout.

He’s also a chess player

Here’s what you can take from this as a chess player. Whenever you’re playing, it’s very important that it’s clear in your mind what your priority is. Are you exploring or exploiting? Another way of saying the same thing is, are you maximizing for learning or winning?

Most players probably maximize for winning too often. When you’re playing practice games online, for example, you should clearly be maximizing for learning. Here are some ways to prioritize exploration:

• Play a move that looks intriguing or has potential, even if you’re not sure if it will work.

• Decline draw offers (you’ll learn more by continuing the game).

• Learn a new opening to gain experience with different pawn structures and plans.

I’m sure you can think of more examples. The point is, you should know what you’re trying to accomplish. There’s also a time and place to go all-out to win; for example, at a tournament you care deeply about or has a big prize.

I’m guessing some of you are now thinking, “I’ll just maximize for learning all the time. That’s what I’m in it for.” I would still recommend going all out to win, at least occasionally. Paradoxically, this may teach you the most of all, since you don’t really know where you’re at until you give a full effort.

It’s not that one of these is better than the other. It’s actually really important to have both. You need a place to experiment, and a place to bring together the best of what you can do in a pressurized environment. The important thing is to know which is which.

This season, the Spurs have a better team. They’re in a position to contend. And Wemby is going absolutely nuclear. He spent two years exploring. Now he’s exploiting everything he learned.