James Harden is a “Thirsty Scorer.” What is a Thirsty Scorer?

On ESPN Thursday morning, Max Kellerman referred to the Rockets’ James Harden as “the thirstiest scorer, but not the best scorer.”

The quip set NBA Twitter on fire—Kellerman wasn’t the first to use the term, but he has certainly at least for the moment popularized it—but what it didn’t do was explain just what the heck a thirsty scorer is and what separates one from a great scorer.

Fortunately, we can invoke Sheed’s Law around here—ball don’t lie—and devise a stat which we’ll call Thirst Points to separate efficient scoring (there are plenty of stats for that) from pure, selfish, “I want my 50 points even if I have to shoot 38 shots and force my way to the line for 24 free throws to do it” thirsty scoring.

With that assertion, we’re going to need to decide on some good diagnostic criteria for thirst.

Let’s start with points.

If you only score 4 points on 13 shots, you’re not a thirsty scorer. You just suck. There’s a difference.

Along the same lines, let’s make sure that every measure of thirst pushes the stat in the same direction; we’ll be adding and multiplying here, not subtracting and dividing, since thirst should get downright recursive the more a guy like Harden keeps chucking the ball at the basket inefficiently while somehow still managing to score a ton of points.

So let’s try this here formula. (PTS+USG%) * (1/eFG%) * (1/1-FTR) to create a baseline. Yes, I know you can simplify the math, but I wanted to hold to that “adding and multiplying” principle, and every math teacher I ever had since fifth grade is now either facepalming or turning in their grave depending on if they’re still alive.

And you’ll notice that turning the ball over plays into this and gives a bonus to guys who cough up the ball while trying too hard to get space to score, committing offensive fouls, traveling, or trying to draw a foul, failing, and having the ref give them a Nelson Muntz non-call.

Harden scored 50 points, had a usage rate of 43.7 (for a total of 93.7) while putting up an eFG% of .342 (for a multiplier of 2.92) and a FTR of .632 (a multiplier of 2.72, and I’m ignoring rounding because excessive precision irks me.)

So 93.7 * 2.92 * 2.72 = 744.2 Thirsty Points.

Is that a lot? Well, let’s go find some other inefficient scoring performances from the annals of NBA mania and see what we come up with.

Like Harden’s own teammate, Russell Westbrook, who gets to test the “just sucks” assertion above with his 19 points on 7 of 30 in the same game.

So 19 plus Westbrook’s usage of 30.6 provides us a baseline of just 49.6. His multipliers are 4 (eFG% of .250) and 1.15 (he attempted only four free throws for a FTR of .133.)

So 49.6 x 4 x 1.15 = 228.16. Still pretty thirsty, but you have to score some actual points to really move the needle.

Let’s get away from the Rockets-Spurs game and look at some other truly thirsty performances from NBA history.

Like Kobe Bryant‘s 81-point outburst on January 21, 2006.

In that game he scored 81 on a USG% of 56.8 for a raw score of 137.8 thirst points.

However, he shot the ball quite well, accruing a multiplier of just 1.46 for his .685 eFG% and 1.77 because 20 FTA isn’t quite as impressive against 46 FGA as you might think.

That raw 137.8 becomes 356.10 Thirst Points, less than half Harden’s total in that 50-piece.

Which broadly squares with Kellerman’s point about a “thirsty” scorer not necessarily being a “good” scorer, while also reinforcing my point that there’s a certain butter zone that involves accruing thirst points purely by being your team’s high usage scorer.

Let’s further illustrate this by taking another 50-piece that was nothing at all like the 50-piece Harden barfed out Tuesday night.

On February 25, 2016, the Golden State Warriors beat the Orlando Magic 130-114, raising their record to 52-5.

Stephen Curry had 51 points on 20-of-27 shooting and 10-of-15 from long range while shooting just one free throw (which of course he made.)

He had just a 39.8 USG%. That’s 90.8 raw Thirst Points, below Harden’s total and miles below Kobe’s.

He also achieved multipliers of just 1.08 for eFG% and 1.04 for FTR.

So your total Thirst Points for a 51-point outburst? Just 102. Double. Harden’s 744.2 was 14.88 times his point total, 7.44 times less efficient than Steph’s performance.

And there…I do believe a stat is born. Two stats, when you really get down to it.

First, we have Thirst Points. That’s the formula we’ve been playing with up to this point.

But second, and much more to Kellerman’s point about “the thirstiest scorer but not the best scorer”, we have a way to compare thirst independent of actual number of points scored (and bring Westbrook’s 12x multiplier of Thirst Points over actual points into perspective) and make Steph’s 2.00x multiplier comparable to the 2.58 that Kobe put up in 2006.

Raw Thirst Points are fun, but its derived stat of Thirst Factor makes comparisons possible.

For reference, Thirst Factor is calculated as ((PTS+USG%)/PTS)/eFG%/(1-FTR). That removes all that multiplicative linguistic gimcrackery from my rhetorical flourishes above.

So raw Thirst Points x Thirst Factor = true Thirst Points.

And what’s better, this can be done for individual games…or per-game season averages, or per-36 minute or per-100 possession or whatever you want to throw it into. As long as you have a point total, a usage rate, a free throw rate, and an effective field goal percentage (those last three being independent of however you decide to figure the point total since they’re always rate stats), you have what you need to determine how thirsty a scorer is.

So someone blast Max Kellerman and get him to read this. All I ask is attribution if it gets out into the wider world.

Stay tuned to Pace and Space, and if you’re new here, I highly recommend the “Analysis” tab (clickable on desktop, in the drop-down menu if you’re on mobile) at the top of this article if you like stuff like this, or click on the “Statistical Tests” tag below for more specific stat experiment type stuff.

Thanks for reading!