Just in time for the 2018 NBA Draft, it’s my 2018 Draft (Mini)series! For those of you who follow me on twitter and the like, you may have noticed that I did a 5 part NFL Draft Series. As a result of that, I spent a lot of time digging into NFL theory. Much like how the Super Bowl champion Eagles’ GM Howie Roseman spent a year talking with executives in other sports and came away with a better understanding, so too have I spent time digging into another sport and come back with new insight.
Maybe we’re all going about this NBA Draft prospect projection stuff all wrong. See, when I dove into NFL QB analysis, one fact was so glaring, so completely undeniable, that it became a necessary asterisk next to every prospect: a QB’s chance of success in the NFL is so significantly dependent on the head coach and offensive coordinator that with the exception of the truly elite, the best QB prospect in a bad system had a lower chance of success than a mediocre prospect in a great system.
In other words, with the exception of the best of the best (Wilson, Rodgers, Brees, younger Tom Brady/Peyton Manning), there isn’t so much a “true talent level” as much as a baseline set of mental and physical tools which can manifest extremely differently based on coaching, scheme, and system. What made this so obvious was QBs going from good coaches (Sean McVay, Kyle Shanahan) to bad coaches (Matt Cavanaugh, Steve Sarkisian) and vice versa (most notably, not having Jeff Fisher as your coach).
Jared Goff had a truly dreadful rookie season under Fisher, but put up great numbers under McVay. Matt Ryan and Kirk Cousins both significantly declined after losing their great OCs. Nick Foles put up one of the best seasons ever under Chip Kelly before the rest of the league caught up to that offense, declined when Kelly didn’t adjust, cratered under Fisher, and then blossomed again when the Eagles built an offense specifically for him.
In the NBA, there are a few truly can’t miss prospects – LeBron, Durant, Anthony Davis. These guys come around once every five years or so. The problem is figuring out the other 99.9% of prospects.
Typically, draftniks approach analyzing prospects in a combination of two ways. One is developing and using a stat-based model to map out what stats have typically predicted success at the next level, then seeing which prospects best fit the model. The other is seeing which prospects have the best tools and projecting that the better the tools, the higher the ceiling and therefore the better the prospect. Both methods use historical data to project in different ways – the former using historical statistical data, the latter using historical biometric data, but both are ultimately just projecting, which is a fancy way of saying educated guessing.
Under this two pronged projection strategy, the onus of a player being successful is on the player – either he was good in college and therefore will be successful in the pros or he has all the tools to be successful and therefore given time to develop, will be successful. But either way, the assumption is that the player will succeed or fail primarily based on his own merit. And if that’s not true, then, well, everybody is doing it really really wrong.
See, some strange things happened in 2017. Jayson Tatum had the best 19 year old season since LeBron James. Jaylen Brown and Dejounte Murray both significantly outperformed every college statistical indicator. Donovan Mitchell turned into an offensive star. Going back to 2012, none of these things had happened. 19 year olds have been bad, bad college players have been bad pro players, and college defensive dynamos haven’t been pro offensive dynamos. So what happened?
One explanation is that there is something unique mentally about these players, something that online draftniks would never be exposed to or have access to that explains why certain players develop better or quicker than others. But every team would have access to that, and other than Jaylen Brown, there’s been no indication that these guys are somehow completely different than everybody else. Another explanation would be that these guys are special athletic specimen, but that’s hard to judge because 3 of the 4 did not attend the combine. Mitchell’s measurements don’t seem to indicate truly out of the ordinary – he had an elite sprint time and leap but below average agility and shuttle, and other guys with great sprint and leap didn’t pan out, so I’m not sure that’s a real indicator either.
As you can tell from the beginning of this article, I came to a different conclusion: that coaching and scheme matters much more than we give it credit for. Basically, when it comes to evaluating NBA Draft prospects, we typically do so with a passing wave that, yeah, somebody coached by Pop will turn out better than somebody in Sacramento, but it’s not treated as a big factor. Should it be though? What if, in fact, for all but the absolute best prospects, the most important factor in their success is what coaching staff they end up with?
Um…well…digging in, that’s actually not right either. While it’s easy to remember the hits, there’s plenty of misses for every coach too, even Pop and Stevens. And quite frankly, most players just follow the standard model of moderate progression year by year regardless of coach. Which is rather frustrating really. While there is certainly some element of scheme and coaching present, there’s no consistent effect. Much like the other explanations, it’s a factor, but not a defining one.
That leaves us with one final explanation. The most likely explanation. The most unsatisfying explanation: projecting most prospects is little more than pure, random variance combined with opportunities to play NBA minutes over a span of many years with quality players. That’s not an explanation that can be used for any useful purposes going forward. But sometimes, shit happens for no apparent reason and there’s nothing more to it.
Back in 2016, I did a Big Board that was a little more detailed in the breakdown of chances of success at the next level. While that breakdown wasn’t perfect, the larger overall point holds as true today as it did then – most prospects will fail. And after spending 5 years trying to refine and analyze my way to answers, I am no closer now than I was 5 years ago.
I am not saying that the draft is truly random, because it’s not. But in terms of determining which players will blossom into actual starters or better? Predicting failure for every player is still highly effective. By my count, players drafted in the lottery have blossomed into unquestioned starters at about a 35% rate. Some drafts are a little higher or lower, but that’s the rough rate. And while top 3 picks tend to be superstars more, in terms of finding starters, they’re pretty evenly distributed throughout the lottery. The 5-9 range has been less productive recently than the 10-14 range, but again, that appears to be variance more than some sort of projectable trend.
In the scientific community, there is often pressure to come up with some conclusion, some result so that the research was not a “failure”. I have no such pressure here. I have no need to fabricate results for any reason. I will continue doing big boards and big picture analysis, because again, it’s not completely random. But I have come to accept that if there is a way to be super accurate projecting prospects, it will not come from me. I have come to accept that is simply a fool’s errand.
Ultimately, take big boards and prospect analysis as what they are and nothing more – educated guesses on who is most likely to hit on that 35% or less chance to become a starter. Hitting more than 35% is a success. Hitting 50% is a resounding success. That’s just the way this goes.