First, before getting into the post, just a happy note that this is not a statistically heavy post and understanding is as simple as looking at a few graphs (yay!!!). So before reading a couple mentions of my draft model or following links to some of Layne Varsho’s (VJL’s) work on CanisHoopus, don’t be alarmed or think this is going to be some wonky read that takes a degree in econometrics to keep up with. Now, on with the show.
For the last two years, I have worked on a draft model similar to Dalt99’s and Kevin Pelton’s. Anyone who read my piece / predictions this year may remember that one core finding was being called a high IQ basketball player by scouts was a very strong predictor of success (even more so when combined with steals per 36 minutes). And I experimented with a ton of variables and high basketball IQ was both consistent and persistent in them – basically, showing it is very likely to be a big factor and not just a fluke based on a limited model. I was further intrigued when I read a post by Layne Vashroo where he looked at how young players improve and found that players with high skill levels improve at faster and more steadily than those with superior athleticism.
This made me curious, what is it about high IQ players? What does their performance look like compared to other players? And can we build on Layne’s ideas and show that perhaps NBA teams place too much important on athleticism and not enough on basketball intelligence?
One caveat here is I am specifically speaking about basketball IQ or feel for the game. It’s certainly possible that a player could be highly intelligent, but simply not have a passion for the game and therefore not possess high basketball IQ. Or that creative intelligence instead of analytical intelligence is more important for NBA players, so a potentially great investment banker might not be seen as having a great basketball IQ. I don’t presume to know how "smart" a player is. Most of what the scouts and I can observe is how their intelligence translates onto the basketball court.
So how do we know a High IQ player and how does their performance compare?
Before diving into some graphs and numbers, I think it’s only fair to touch on methodology. Specifically, how does a player get classified as having a high basketball IQ? When putting together my draft model, I had a theory that there were probably some elements traditional scouts could identify better than a number and these could be incorporated into the model. For consistency, I used all of the ESPN scouting reports, which are available dating back to 2005. Whatever your opinion of Chad Ford, he does speak with a number of NBA scouts and uses some fairly consistent terminology. So finding the high IQ players was as simple as reading through hundreds of scouting reports and crediting the players ESPN claimed that scouts said had a high basketball IQ. The same methodology was used to identify low basketball IQ players.
So first, one more brief callback to the model. High IQ and age were the two simplest, most consistent variables even as the rest of the model changed. Being identified as a high IQ player by scouts was enough to add about 2 points to a player’s projected peak Win Score. However, I wanted to dig a little deeper into player development. So with the help of some data provided by Layne, I was able to chart the average Win Score over time for a player’s first 7 years. The results are below:
It’s worth noting that beyond year 4-5, the sample sizes start getting pretty small. Frankly, I could have cut this chart after 3-4 years, but was curious if there was a major change over time (e.g. low basketball IQ players "clicked" in year 6). Still, the chart is pretty clear, high basketball IQ players tend to improve more over time, whereas low basketball IQ players tend to fluctuate up and down and neutral players showed some improvement, but remained pretty flat. I am sure there are 100 more ways to cut this data and dive into it and I am happy to provide the data for anyone who wants to dive further into this analysis and add to the discussion.
But the next natural question here should be "so what?" Why does this matter? After all, it isn’t a surprise that smart players perform better. If you asked a NBA GM if they would prefer to draft a player with a low or high basketball IQ – every single one would answer a high basketball IQ. However, I think this is important because basketball, much like life is about trade offs. To borrow some learnings from marketing, let’s say you are looking to buy a car. And old school marketing survey would ask you to rate the importance of features such as color, gas mileage, speaker system and price on a scale of 1-5. Well, in a vacuum, almost all of those features would get a 4 or 5. I mean, how many people don’t want great speakers! However, now most firms use some type of a conjoint analysis to make people choose between features. They ask, would you rather have a red card with amazing speakers for $40,000 or a blue car with poor speakers for $15,000. After asking enough of these questions, you start to get a sense of what is really important to people. And frankly, while speaker quality matters in a vacuum, most people would trade down in speaker quality to pay a lower price or have a more beautiful looking car.
Similarly, every GM would love a player who is a highly intelligent, freak athlete, who is a born leader and a hard worker. Basically, if every draft pick or potential free agent signing was LeBron James, being a NBA GM would be easy. However, there is only one LeBron. So in the real world, GMs have to look at potential draft picks and signings and make trade offs and many GMs and fans still seem to like "swinging for the fences" and drafting a highly athletic player. The problem is that other than combining athleticism with high IQ, at no point in any of my potential models was athleticism very important. Combine that with Layne’s findings and you might start to wonder if GMs and fans have the wrong idea about how to "swing for the fences." Instead of drafting Joe Alexander or Hasheem Thabeet and hoping you can teach them to play basketball, you are better off looking to draft Stephen Curry or Brook Lopez and helping them to develop (or drafting James Harden or Jordan Adams and helping them to lose weight). Or for Pete and Vivek, to draft Nik Stauskas with Noah Vonleh still available. And what’s interesting is that looking at athleticism as we did with basketball IQ, we see the following win scores over time:
Everything is just sort of pushed towards the middle compared to the basketball IQ graph. Now, clearly not every player is average. So this does not mean you need to target 8th year veterans who were called high IQ players when they entered the league. With veterans, you have a body of work to look at. The scouts may have been wrong (e.g. they do have a high basketball IQ) or they may have studied and learned over time. But in a situation where you are gambling on the relative unknown like the draft or signing/ trading for a young player – it appears you are better off gambling on basketball IQ instead of athleticism.
Ok. Forget statistics. Do any NBA teams seem to get this?
Sometimes statistics shock you and give you new, crazy insights that make you rethink the way you view the world (or more likely rethink your model and look for errors). However, statisticians are typically more happy or relieved when the model validates some hypothesis that seems to link to real world behavior. In other words, if you have 5 potential ways you think the world might work, it’s nice if the model seems to back one of them up instead of giving you some bizarre, funky result that you can’t explain.
What’s nice here is that a quick look at the NBA shows that many of the best run teams already seem to appreciate this concept and some of the worst ones do not. As Kings fans, we watched some very talented teams crater under the own inability to play fundamental basketball, as detailed in this post on a 4th quarter collapse. We watched guys like Hickson, Green, Evans, Thornton, JT, etc. play ISO offense and get lost on rotations. We heard every coach talk about how they were "simplifying the playbook" and "making things easier." The Pistons have had similar issues, compiling highly talented players who seem to look lost on the basketball court. And it’s no accident that the Wizards made the leap after ditching McGee and Blatche and bringing in Nene and Beal.
Conversely, most winning teams seem to thrive off of high IQ players. The Spurs specifically target them. They run a complex system and constantly bring in smart veterans and target high IQ rookies, such as Danny Green and Dejuan Blair. The Hawks brought in Spurs disciples and immediately grabbed Korver and Millsap, while benching Sweet Lou. The Pacers have rebuilt around high IQ draftees like Hibbert and Paul George. The Blazers did the same with Aldridge and Lillard.
The Warriors are almost the epitome of this system. They have built their team from the high IQ draft list – Curry, Klay, Barnes and Draymond Green. It’s not a surprise a NBA scout I had the chance to meet said the Warriors were trying to trade into the draft this year for Stauskas. Or that they jettisoned Ellis to bring in Bogut. Looking back, the Warriors never seemed to be building a very athletic team. Frankly, a few years ago, analysts were predicting our team with Evans, Cousins and Hickson would be the better one. But the Warriors drafted players who intuitively understood the game and have made drastic improvements on both sides of the basketball.
So what does this mean for the Kings?
Bringing this back to our beloved team, it’s one of the reasons I am excited for this season. We have brought in two heady PGs and drafted a rookie off of the high basketball IQ list. And frankly, there’s a lot of reasons to question the Stauskas pick. The 8th pick seemed high for him. No draft model (mine included) had him as a top 20 player. Yet, of the 6 players who far outperformed my models predictions (or retrodictions to be fair), 4 of them were high IQ players. Basically, at every level it helps to be a high IQ basketball player who can run more complex sets, better synthesize feedback from your coaches and potentially even self-diagnose problems if you wind up in a disastrous situation (e.g. the Maloof era Kings).
So as the new season gets ready to kick off, there’s a reason for excitement. Our front office seems to appreciate the value of basketball intelligence. And our rotations are starting to get peppered with veterans who know the game and young players who have an intuitive feel for it. After years of watching a team under-perform compared to the talent on the roster, we seem to be building a team like the Warriors, Spurs, Blazers and Pacers where the whole is greater than the sum of the individual parts. This might not be evident from game 1, but by game 82 there’s a good chance we see a much more cohesive and savvy team than we have seen in years.