Can We Predict Fantasy Success from College Metrics?

7 min read

Any time you begin to learn about a new hobby, a new activity, a new job, you will encounter shorthand terminology that has gone from being casual slang to validating shibboleth (a distinguishing characteristic of a group or subculture) to vital technical language in that group.

In the world of gaming and game design, the term “grok” isn’t a nonsense word that sounds like a goblin’s name; instead, it’s the process of a player understanding a game mechanic. A term that has grown beyond (or possibly been appropriated from) its drag subcultural roots is “spilling the tea”, or sharing the gossipy truth about something. In the post office, the word for the junk mail that can’t be forwarded or returned to sender is either “skosh” or “skulch”. This jargon carries implied meaning to people who have the same experiences as each other on a regular basis, and therefore allows them to communicate more efficiently.

This is the same reason we in the fantasy basketball field study statistics and watch game tape: we hope to find facets of these players’ games that will translate to pro success or indicate ahead of time that development is on the horizon.

So, which traits, measurables, and metrics will help us find “blue chips”, and which will prove to be scouting thirst traps?

Trust The Process

As I mentioned in a previous article, the more we can identify general qualities that indicate value, the easier it becomes to find players who will pan out in our dynasty basketball leagues. Nothing we identify will give us ironclad guarantees that a player who checks all the boxes will turn into a fantasy superstar, but if we can figure out what those boxes are, we know what to look for and what to avoid in talent evaluation.

To help create those boxes, I compiled the NBA Combine data and career college production from 2000 to present and compared success in a variety of statistical categories to a simple metric of fantasy success: NBA Fantasy Points per Minute. I ran a simple R correlation to see how closely each college stat or athletic measurement was related to NBA fantasy production, and give us a relative sense of which metrics matter most.

If you’re not familiar with R correlations or the neat little number it spits out (the coefficient), the R correlation is simply a mathematical way of depicting the strength of a relationship between two sets of variables. While I’ll caution you to never confuse correlation for causation, this test can show us a benchmark of the strength of a statistical relationship. The closer the value of the R coefficient is to 1 or -1, the stronger the relationship between variables; the closer the value is to 0, the more chaotic or random the relationship.

You Can’t Teach Size

In terms of Combine anthrometrics (physical measurements), this is what I found:

Absolutely none of these criteria crosses the threshold for statistical significance to NBA success, and that’s no surprise. Players as tall as Yao Ming and as short as Allen Iverson have made immense production in the league, and the same goes for players as immobile as Shawn Bradley or as quick as Derrick Rose (fun note: it’s interesting that hand length has the strongest correlation, since maybe the only constant among NBA physique demands is that one must be able to hold the ball).

Even when we break these down by position, there is little to hang our collective hat on. The chart below shows this; none of the results signal even moderate significance.

Pulling down the position-specific strength and agility drills does give us a little more insight at points. This is what we see when we do so.

Again, there’s not much here. But the correlation between fantasy scoring and agility drills for centers is astronomically higher than any other relationship. Taken all together, it’s clear that while players of many shapes, sizes, and persuasions can thrive in the 1-4 positions, the most successful centers will be big, muscled, have a long reach, and quick to respond and move.

With athletic and physical measures mainly debunked, let’s turn to pre-pro production to help us out.

Numbers Don’t Lie

One more interesting factor remains from the Combine, but this isn’t a measure of athleticism; instead, it’s the shooting drills that help simulate different play situations.

Prospects shoot from the NBA three line (23-foot-9), the college three line (20-foot-9), and the high school three line (15-foot), from five different spots around the arcs, and in three ways: spot-up, off-the-dribble, and on-the-move. Each of these helps to indicate different qualities in a shooter, but taken together, can they provide a compelling picture of a future fantasy star? The chart below shows how each grouping of drills correlates to Fantasy Points per Minute.

This is a much more intriguing chart for our purposes, because – while an R correlation of ~0.30 is still fairly weak in a vacuum – relatively speaking, this is the most indicative set of correlations we’ve found yet. You’ll notice that the 20-foot shots tend to be less indicative than either the 23- or 15-footers, and that seems to follow the trend of those long-range two’s dying out in the NBA game (for a visual, check out this Jamal Murray long-range layup). You’ll also notice that spot-up jumpers are a hair more indicative than the off-the-dribble, which suggests that the simple act of shooting says more about a player’s pure scoring talent than drills that factor motion in.

Produce Section

Now, what about an even deeper set of data: college production. Can any of these stats tell us anything about a player’s NBA fantasy future?

There’s a lot in this table, but the strongest correlations to pro fantasy production come from these: Field Goal Percentage, Blocks per Minute, Total Rebounds per Minute, and – perhaps unsurprisingly – college Fantasy Points per Minute.

Let’s start with the easiest to decipher. Fantasy Points per Minute is a fairly sticky stat between college and the pros, and for good reason: if one is productive and efficient in college, it makes sense that they will probably remain so in the NBA. Field Goal Percentage also makes a lot of sense, since a great deal of being an effective fantasy asset is a player’s ability to score points. Defensive mavens can certainly stuff the stat sheet enough for to be fantasy-usable, but the easiest way to rack up faux points is by bucketing real ones.

Speaking of defense, Blocks and Total Rebounds per Minute seem to be important also because they are some of the more individual-centric stats. An assist is dependent on the assisted player scoring after getting the pass, while steals just aren’t as common, and free throws can be affected by how tight the defense plays a player. Blocks and boards are a testament (often) to a player’s size and vertical athleticism, and the old saying goes “you can’t teach size”.


This is all to say that there is no ironclad, easy-to-follow blueprint to predicting which players will be successes in the NBA; if there was, you and I would most certainly be coaches and GM’s. That said, we can identify the most significant statistical factors and form a sort of stat scouting “constellation” for our dynasty players.

I also found it fascinating that there is a fairly relatively significant negative correlation between college three-point shooting and NBA fantasy production, especially since the three-point shot has become such a significant part of the NBA game.

I took that correlation conundrum to our team, and our own Rhett Bauer had these thoughts, which seem to make a lot of sense to me: “Someone whose primary skill is to shoot deep in college probably struggles to find a consistent role in the league. The top prospects are usually superb athletes that need work on their shot, so they aren’t attempting nearly as many three pointers as they are getting to the rim and finishing because they’re way more athletic than the guys around them. Combine that with a shorter three-point line and more inconsistent wing defense in college, and you have a greater likelihood that those guys bust.”

If you’re able to adapt your process to include the strongest of these statistical factors, I’d bet your evaluation process goes from “brick” to “making it rain”.