Testing SMILODON: 2013 Draft

A couple weeks ago, I introduced SMILODON, a stats model that uses prospects’ college numbers to suss out their combination of skills rather than assign them an overall number to predict their NBA success. Last week, I used SMILODON to judge the NCAA prospects from the 2012 draft, the first for which I have the needed data. After reviewing the past five drafts, I will use the model to evaluate the 2017 draft, supplemented with my own scouting observations.

In contrast to the 2012 NBA draft, this set of players quite clearly illustrates one of SMILODON’s current limitations: the lack of information concerning international prospects. What does it matter if the NCAA players are judged correctly, but you pass on Giannis or Gobert? Well, for the purposes of this article, let us pretend it matters a little.

I was satisfied with SMILODON’s performance evaluating the 2012 draft and, for the most part, continue to be satisfied when looking at the 2013 draft. The most important false positive was for a player type I already subjectively discount - small point guards. And even that player, Trey Burke, would not have been the model’s choice at any slot at which he was available.

One note before getting to the results. C.J. McCollum’s scores are based off his junior year statistics due to injury. His senior year flags, based off only 13 games, would have been: shooting - ELITE, driving - MEH, passing - MEH, defense - OK, age - MEH. Not as good as his previous year, but still worthy of consideration, especially when projecting his driving, passing, and defense to progress toward his career means. Any player with an elite skill and no red flags is probably a prospect, especially when projected as a top ten pick by traditional scouts.

If a prospect is flagged as a “steal”, that means that they have either a blue flag, multiple green flags, or all of their skills save one are at least at the level of a yellow flag. A prospect flagged as a “bust” will usually have multiple red flags without the corresponding green flags for balance.

Some flags are more important than others. Age is crucial; a player with a red flag for age needs proficiency in at least two of the three “s” categories: size, shooting, and shotblocking to be taken seriously as a prospect. Undersized 22 year old seniors posting good numbers should be heavily, heavily discounted. Shooting is also crucial for perimeter players. Guards who can’t shoot need to be All-NBA defenders to be worth a first round pick. Point forwards who can’t shoot are also likely to be far less valuable in the NBA.

As in the last article, I sorted players by Draft Express’ mock draft, so that Anthony Bennett merely looks like a mediocre top ten prospect instead of a laughable number one. Doing this exercise really reminded me of how crazy that pick was in real time.

Nerlens Noel looks like the clear number one pick by this method. His profile is reminiscent of Anthony Davis, with the notable exception of shooting. The fact that Noel’s shot looked fairly hopeless played a big role in limiting his offensive potential. The fact that Noel was injured for the most difficult portion of Kentucky’s schedule also should have slightly tempered the enthusiasm surrounding him.

Victor Oladipo, Otto Porter, C.J. McCollum, and Trey Burke all were pegged as prospects by this model. McCollum and Porter both look like good starters with balanced skillsets, Oladipo looks like a defensive specialist who can shoot a little, and Burke is probably the weakest of the group. Even without the benefit of hindsight, driving is a really bad red flag for a point guard prospect to carry, and he is undersized for the position. (A player under 6’3 can be thought of as carrying an additional red flag.)

Neither Alex Len nor Anthony Bennett look like lottery picks by these categories. Ben McLemore is projected as a one dimensional spot-up shooter that would likely come off a bench. Kentavious Caldwell-Pope and Cody Zeller both look like capable rotation players, but would be less exciting than Noel & McCollum in the range in which they were actually drafted.

Excluding the international players, this is a fairly weak group. SMILODON was bearish (one might say cave bearish) on the prospects of Shabazz Muhammad, Tim Hardaway Jr, Shane Larkin, and Michael Carter-Williams. (MCW is a Syracuse prospect whose only green flag is based on his steals. Pass.) Steven Adams projected as a capable defensive big, while Kelly Olynyk projected as an effective offensive big. The former is probably more valuable, so Adams being drafted first makes sense. Adams has been the most valuable of these NCAA players at this point in their careers.

Reggie Bullock is the only player in this range given two green flags and he has had a less than stellar career thus far. He actually received the exact same flags as Darius Miller from the 2012 draft, so this might be a player type overrated by SMILODON: older, less athletic wings who can shoot and pass, but not drive or defend. There is one more player in this draft who fits that archetype, albeit with slightly better driving ability, and his career has been fairly mediocre up until this point.

The only two players picked out as “steals” from this group both have the dreaded red flag for age, meaning that they would each be fine picks in this part of the draft, but maybe shouldn’t be seen as lottery picks, unlike Draymond Green from the 2012 draft.

Jeff Withey was a great shotblocker at Kansas and has been an above average player by every box score based advanced stat you can think of in a four year career backing up Anthony Davis and Rudy Gobert. Gorgui Dieng learned to shoot free throws and midrange jumpers sometime between his final year at Louisville and his debut for the Timberwolves and has become a very effective rotation player.

Mason Plumlee and Allen Crabbe might qualify as misses here. Each has turned into a respectable rotation piece despite lackluster college profiles.

Tony Snell and Andre Roberson both stand out as players worth a flier. Roberson could defend, but do little else. Snell fits the same mold as Bullock and Miller, but could handle a little more than either of them. Mike Muscala also could be a steal at this point in the draft as a very capable rotation big without any weaknesses. Nate Wolters and Solomon Hill both look like capable bench filler despite not quite hitting the “steal” criteria. There are no players in this group that have made a poor rating look silly.

From SMILODON’s perspective, Ryan Kelly is clearly the most interesting of these prospects. While that may seem to be an incredibly sad sentence, I think there is value in confirming that the model has a relatively low incidence of false positives. Kelly has outperformed his draft position, by actually receiving playing time, but has not exactly lit the world on fire. He is also an example of a type of big who receives a green flag for awareness through limiting turnovers and fouls. A player who looks good in AWARE through limiting negative events (Kelly, Shurna from 2012) is probably a less valuable player than one who does so by creating positive events (Draymond, Ben Simmons).

Archie Goodwin and James Southerland both receive two green flags. Goodwin never learned to shoot, a problem for a guard, while Southerland was a Syracuse product with three red flags. If we perform the Syracuse Subtraction on his defensive numbers, he doesn’t look like a steal. Grant Jerrett seemed to have NBA potential, largely on the basis of age and a promising jump shot, but he has not stuck in a NBA rotation.

We get a few blue flags among the notable undrafted from this class! Robert Covington looks like the ultimate 3&D role player. His driving, passing, and upside were all severely limited to the point I wouldn’t advocate drafting him in the first round, but by the end of the 2nd, why not? Ian Clark was an older, undersized shooting guard, but showed enough to warrant a second round selection. Arsalan Kazemi has a very strange profile. It would have been worth a shot. Among players without elite flags, Dewayne Dedmon has the most interesting profile. He was very old, but had the two “s”, size and shotblocking, needed by a big man and a good enough shooting touch to suggest that he wouldn’t be hopeless on offense. It took Dedmon a few years, and a few teams, but he seems to be living up to that potential now.

Next time, join me as I analyze the 2014 draft, and watch things get messy as I sort through a bewildering array of one-dimensional players.