From the Archives: Shorting the Special Snowflakes

The following is a lightly edited/updated piece which I wrote for Ian Levy’s old site, Hickory High which is sadly no longer available on the intrawebs. As we enter NBA Draft prospect season, it’s worth reposting if only because other Nylon Calculus and former Hickory writers have asked if it’s still online, so here goes with small updates footnoted:

One of my nerdiest hobbies of all time was my early 2000s obsession with an online game called Hollywood Stock Exchange. HSX, as it is called, imagines that one could buy and sell stock in actors and new release movies.  The prices would float up or down based on whether players of the game were buying or selling, and the values would then adjust upon the announcement of each film’s opening weekend box-office take.

One viable strategy for “playing the openers” in HSX was fairly simple: short ‘em all. The aggregate expectations of the market of players summed to a larger amount than the movie going public as a whole was likely to spend. Thus the combined values of the movies’ “stocks” would fall, and if you bet against every film, you were nearly guaranteed to win overall even if some movies exceeded expectations.

I tend to feel the same way about the NBA draft, and if I could short every incoming draft class, I probably would. Of course, no team thinks their pick is going to be a bust, but the sad fact is more players wash out than make it, fewer lottery picks become stars than don’t, and those second round “steals” are most likely never to be heard from again[1. The chart is largely unscientific as it involved me placing players in rough “outcome” categories, but I think the rough breakdown is in line with most other analysis on the topic.]:

Now, much like the disclaimer actual financial advisers are required to give that past performance is no indication of future returns, any individual draft might be stronger or weaker than the historical trend. This does not invalidate the “short ’em all” strategy though, as a touted draft class like this one[2. This was written ahead of the 2014 draft, on which the jury is still out, especially relative to those lofty pre-draft expectations.] already has those higher expectations built in.

Still, a one-size fits all approach only gets us so far.  Some of these lauded prospects live up to and even surpass these high expectations. In recent years, Anthony Davis has not only justified the early hype, but has tantalized with how lofty his eventual ceiling might be. John Wall appears ready to join the elite, alongside much phenoms such as LeBron James and Kevin Durant. Still others have exceeded any rationale expectations for their progress. Certainly very few thought Kevin Love[3. I still believe there is an All-Star level player there, but perhaps not in Cleveland.]  or Russell Westbrook would be the perennial All-Stars they have turned into.

Casual analysis doesn’t lead to any easy conclusions as to who those breakout players might be, though teams are spending millions and amateur analysts spending thousands of hours to look into just that. However, there might be some simple heuristics that help identify who they won’t be.  One of these is what might be called the “Special Snowflake Player Comp.”

In this sort of comparison, a prospect’s weaknesses are mitigated by looking at an NBA player who has succeeded despite those limitations. Got a slightly heavy, slightly unathletic big man who can score and rebound? The next Zach Randolph.  A wing shooter with questionable defensive or ball-handling skills? Why that’s Kyle Korver reincarnate. A great athlete who plays hard but lacks touch or finesse might be a Tony Allen-level defender. And of course, the athletic guard who can’t actually shoot could always be the next Dwyane Wade.  The reasons these comps should be red flags is that those players are sui generis.

Take Wade for example, since 2000, there have been 22 individual player-seasons (min 1500 minutes) where a guard has posted a PER of 20 or higher, while taking under 100 three-point attempts. EIGHT of those seasons have been turned in by Wade. One was Michael Jordan’s last hurrah at 38 another was Grant Hill’s one relatively healthy season in Orlando, and the other 12 are from all point guards, who can make up in play-making for others what they lack in shooting (6 from Tony Parker, two each from John Stockton and Andre Miller, with Sam Cassell and John Wall rounding out the list).

The comparisons to Korver are equally inapt, though that’s mostly because Korver the player-archetype is a much different (and lesser) creature than Kyle Korver the actual player, whose usefulness  isn’t in his deadeye shooting, but rather that he does everything else decently well too,[4 No link to sadly lost Hickory High article goes here :(] allowing him to be on the court much more than your Steve Novak/James Jones model of spot-up specialists.  Thus, while Doug McDermott is often compared to Korver because it’s quick and easy due to shooting ability, complexion and alma mater, the knocks against Korver (defense, passing, rebounding) are all things Korver has learned to do more than adequately over his career, rather than his defining characteristics as a player.

Randolph is benefits not only for unusual nimbleness and instincts around the ball for a player of his size and build, but also has an exceptionally wide wingspan, measuring 4-5 inches wider than Julius Randle’s. Finally, if the comparison is Tony Allen, the subtext is that if the player doesn’t turn into one of the very best on-ball defenders in the league, he’ll be borderline unplayable.[5. And even Allen himself has been rendered of questionable utility at times, most notably by Golden State’s refusal to even pretend to guard him in the 2015 Playoffs.]

While eventually, one of these players will turn into “the next Wade” or a reasonable facsimile thereof, if a more realistically attainable comparison isn’t readily apparent, I’m probably selling.