Correlating Big Men’s College Stats to NBA Success

In past years, I have looked at several draft classes of wing and point guard prospects and compared their college stats to NBA advanced metrics to see what the correlation might be between the two sets of numbers. This year, for the first time, I did the same with big men, running regressions on prospects from the 2010-2015 draft classes.

A little clarity on my methods before jumping into some of the data and analysis. I only looked at players who played at least 600 or so NBA minutes, ignoring those with only brief stints in the league, or those who didn’t play at all. While examining those busts to see if any of their college stats could have predicted their failures would be interesting, their sample size in the NBA is just too small, and their advanced stats in the NBA might be misleading because of it. Maybe in the future I’ll look at all those prospects who never got a shot in the NBA or only a limited one, but for now, this analysis will be on players who really played in the league.

Second, I understand that “big men” is a large bucket to generalize 80 something prospects over half a decade’s worth of time. I tried to include every player in those six draft classes that has played most of their minutes at power forward or center in their time in the NBA (not in college). I understand that a stretch power forward like Patrick Patterson is a very different player than a plodding center like Jahlil Okafor, and that breaking down the “big man” category into smaller sections could prove useful as well going forward. I stayed away from that this year because it gets tough to classify players as the groups get smaller and smaller, and it’s difficult to find the cut-off line. So for this year, power forwards and centers were grouped together.

As always, I only look at the stats from the prospects’ last two seasons in college-- if I’m filling out the spreadsheet for a senior, I ignore his freshman and sophomore years. I think that too many upperclassmen get penalized for their age, and as “age” is a separate category in my regression analysis, I don’t need to drag the stats of the juniors and seniors down further. I also don’t use per-40 stats, mostly because if players are going to make it in the NBA, they should be able to play big minutes in college. Sure, some coaches don’t play their best players as much as they might, but maybe there is some limitation to the player that they know about, hence the lack of minutes.

With all that said, here are the simplified results of the regression analysis. Significant means that the statistic, on its own, explained at least some of the variability in the NBA advanced statistics at some level. Model-usable indicates that while not useful by itself, the inclusion of that stat strengthened the overall model of correlation. Insignificant suggests, simply, that the college stat had no real bearing on the players’ advanced metrics in the NBA.

Now that the relevant statistics have been determined, here’s a brief look at the impact each of them had on the final models. Any stat labeled ‘positive’ means that the prospects had better advanced statistics when that value was higher. ‘Negative’ designates the opposite: the lower values in that category translated to superior advanced numbers.

For the most part, the coefficients of the variables make sense. The counting stats (rebounds, assists, blocks) have positive coefficients, meaning that the higher those numbers are, the better the prospect is. Age has a negative coefficient— younger players generally turn out superior advanced statistics in the NBA. This confirms the general consensus that younger prospects have more upside than older college players. The two coefficients that go against the grain are points per game and three-pointers per game, both of which are negative when common sense would lean towards their being positive.

The three-point variable can be explained away with relative ease. Most big men in the sample (even those who became proficient three-point shooters in the NBA) didn’t take many outside shots in college. That means that a handful of players who took a lot more threes skewed the sample somewhat (Ryan Kelly and Robbie Hummel being prime examples of this). Points per game being a negative factor is somewhat more complicated, and more interesting. Many of the most prolific big men college scorers of the past decade (who played in the NBA) were busts in the league, especially upperclassmen (Luke Harangody, Andrew Nicholson). By contrast, quite a few big men who didn’t score much at all in college have been extremely successful in the NBA. Why is scoring a negative for big men prospects?

I have a couple of thoughts on this. The first is that in college, the physical differences between freshmen and seniors, especially for big men, give more mature players a large advantage. A 22-year old senior, fully developed and muscled, can dominate younger and more talented/skilled players in college. That physical edge will probably be gone in the NBA, especially if it is dependent on strength/size rather than athleticism. Similarly, players who develop physically earlier than their peers might not work on their skills as much as they should, as they don’t need finesse to score. Therefore, some big men who are potent scorers in college might actually be disadvantaged later in their careers.

The second point is related to the first: The college and NBA games are very different from one another. College basketball promotes much less pick and roll offense, and is generally less free-flowing than the NBA, instead featuring pointless passing around the perimeter and post-ups of big men. In the NBA, while certain big men are a consistent matchup advantage in the paint and are fed there frequently, the game is more perimeter-oriented. Most NBA big men score via the pick and roll, putbacks on offensive rebounds, and running the court in transition rather than through bully-ball in the post. That difference in scoring method can mean that some players’ scoring in college doesn’t translate to the NBA. The primary function of big men in today’s NBA, at least for most teams, is not to score in isolation at all, but to provide rim protection on defense, set hard screens and picks on offense, and let shots be created for them, either via the roll or spotting up along the perimeter. High-scoring big men in college, even if they are extremely skilled post players, don’t always check off those boxes.

Rebounds and blocks can also come from a place of physical advantage. However, while sheer height and size can lead to accumulating both stats, athleticism and basketball instincts come into play just as much, if not more. Assists, too, can be the result of a system or unstoppable play in the post. Yet if players don’t have the vision and basketball IQ to move the basketball smartly and correctly, they won’t rack up assists regardless of other factors. This is particularly true for big men, who don’t have the ball as much as point guards, and are therefore reliant on their reading of the court to generate assists rather than piling them up through usage rate and ball dominance alone.

I don’t think that big men’s scoring numbers in college should be ignored, or that players who score more are worse prospects than those who aren’t scorers at the NCAA level. However, just like everything else, a player’s scoring should be examined to see if it is being accomplished in a method that might translate to the NBA. In general, I value a big man’s blocks, rebounds, and assists more than his scoring (or even efficiency), as I believe they are better indicators of a prospect’s functional athleticism and basketball instincts/IQ. In a week or two, I will apply this data to the 2018 draft class to see which prospects stand out in a good or bad way, and determine the direction I think the Clippers should lean if they have a choice to make between several similarly-ranked big men prospects.