As the NBA has grown and the age of analytics has prospered, so have the ways we’ve looked at who’s efficient. The most essential aspect of any play is scoring, and as such, we’ve evolved in the way we look at it — particularly regarding understanding efficiency.

For a while, we just appreciated how *much* a guy scored without taking regard as to how many shots he took to do it. In a two-point age where all field goals counted the same, that was fine to a point. But when we added a three-point line, some people realized that while someone might be shooting worse from there, those shots might be actually scoring more points.

Would you rather have a guy go 4-of-10 from deep or 5-of-10 from the rim? One has a higher percentage; the other produces more points. Thus effective field goal percentage was born, which assess the value of a made trey at 1.5 times that of the two-point variety.

But then, some smart people realized that none of this accounts for free throws, which can be the most efficient shot in basketball. If a 90 percent free throw shooter takes 12 shots from the charity stripe in one game, he’s going to probably make 11 of them. That’s 11 points on roughly six possessions. To match that from the field, someone would have to make 5-of-6 field goals with one make coming from behind the big semicircle.

So we came up with true shooting percentage to measure that. Although, in it, we count a free throw attempt as .44 of a possession instead of .50 because one attempt isn’t really half a possession. There are times when a shooter might get fouled shooting a three, taking a technical foul or completing an and-1.

True shooting percentage is better, but it’s still not accurate because not all players have it worked out to exactly .44. That’s just the overall league average. There’s disparity. It’s estimating instead of getting an exact number.

The problem with all of these numbers is that they’re trying to assess the efficiency of the *shot* and not the *possession*, and everything in basketball these days points to the possession. That’s the best framework from which to understand the game.

So how do you assess that?

There are three ways a player will use a possession. If he makes a shot or draws a shooting foul (though both can happen in the same play) he almost always ends the possession, with one exception I’ll get to momentarily.

The second way a player *can* use a possession (but doesn’t necessarily use one) is if he misses a shot. If his teammates rebound his miss, he didn’t use it, because it’s not over yet. (This can also happen if a player draws a shooting foul and misses his last free throw attempt, which is the rare case a player wouldn’t use the possession on a shooting foul and/or made field goal).

The third way a player uses a possession is on a turnover, though, it should only be accounted to his scoring efficiency if he was actually trying to score when he turned it over. For example, if he gets called for a three-point violation or illegal screen, that says nothing about his attempt to score the ball. Nor does a bad pass. Those should still be counted, just accounted to his scoring endeavors.

Things like charging, double-dribbles, palming, going out of bounds, etc. are the types of turnovers a player makes when actively trying to score. Ergo, those are the ones we should focus on.

So, this becomes our formula for possessions:

*FGA+Shooting Foul+Ball-Handling Turnover-Missed FG Rebound by Teammates. *

That gives us an exact number. Then, by looking at the number of points scored compared to the number of possessions, you can get the number of points per possession to accurately reflect the efficiency of a scorer in a way that supersedes any current standard.

Using play by play data through games of 11/28, here’s how many possessions each player used and how many points they scored in doing so:

The trendline, as you’d expect, says that as the number of possessions go up, so do the points scored. The players to the right of the trendline score more than their possessions used would suggest; the players to the left score fewer. This helps account for the fact that higher usage players will tend to have their efficiency go down because of the increased attention. So, guys like James Harden and Russell Westbrook (at the top, just to the right of the trendline) are actually pretty efficient for what they do.

That’s because of a couple of reasons. First, they both draw a lot of shooting fouls and are very efficient at the line. Harden has drawn 79 and Westbrook 64. They also both see a good number of their misses corralled by their teammates. That’s not coincidental. That they draw so much defensive attention leaves better rebounding opportunities. Westbrook’s buddies have hauled in 48 of his misses and Harden’s have grabbed 42 of his.

But Stephen Curry is on an island here. He’s that Golden State logo way off to the right. He’s used 412 possessions, seven fewer than Harden. He’s also scored 93 more points.

That Pistons logo bumping out way to the left belongs to Andre Drummond. Kobe Bryant and Derrick Rose are also pretty bad. You can see Rose’s logo to the left near the 200 grid-line, but Bryant’s is a little obscured.

Here’s an alternative view. The line is indicative of the median for players, meaning that half are above it, and half are below it:

Again, look how far off Curry is at 1.393 points per possession used (think of that like a 139.3 offensive rating!). He’s far and away the most efficient player with more than 100 possessions used. In fact, the only player more efficient than Curry who’s used 25 is, weirdly, Kirk Hinrich at 1.553. So, yeah, expect some regression there.

Here are the players with at least 1.15 points per possession and those who are under 1.00:

Emmanuel Mudiay is the most inefficient player. Derrick Rose missed the chart on a technicality — he’s only used 196 possession, but he’d be second-worst at .832 if he were on it. Other noteworthy stars include LaMarcus Aldridge and John Wall.

For a more interactive version of the chart where you can see the complete rankings and filter by points, possessions or teams, as well as see all the pertinent data by hovering over a player, click here.