True shooting percentage and the death of FG%
For most of basketball history, field goal percentage was the cleanest thing about a box score. You shot, you either made it or you didn't, and the column at the bottom of the page told you how often you made it. The number was simple, it was honest, and for fifty years it was approximately right. Then the three-pointer arrived, and then it gradually took over the sport, and somewhere along the way FG% stopped meaning what it used to mean. Today, looking only at FG% to evaluate a scorer is roughly like judging a quarterback by completion percentage without knowing how far the passes traveled.
The replacement statistic, true shooting percentage, has been around for a while and is now standard on every analytics site that takes itself seriously. It is not complicated. It is, however, easy to get wrong if you don't understand what it's adjusting for.
What FG% is actually measuring
FG% answers one question: of the shots a player took from the floor, what fraction went in. That's it. A made two-pointer counts the same as a made three-pointer; a missed layup counts the same as a missed thirty-footer. Free throws don't enter the calculation at all. When the only shots available were two-pointers and the occasional free-throw line jumper, this was fine. When some players take half their attempts from beyond the arc and some take none, it stops being a fair comparison.
Consider two guards who each shoot 45% from the field. The first takes 90% of their shots inside the arc; the second takes 60% from three. They have the same FG%, but the second player is scoring at a meaningfully higher rate per shot. The column doesn't know the difference. It treats them as identical scorers, which is the one thing they emphatically are not.
What true shooting fixes
True shooting percentage tries to answer a slightly different question: how many points did a player generate per shooting possession, accounting for the fact that three-pointers are worth more and that free throws are also shots that produce points. The formula is short enough to fit on a napkin: points divided by twice the sum of field goal attempts and 0.44 times free throw attempts.
The 0.44 multiplier on free throws is the part that confuses people. It's there because not every free throw represents a possession. Some come in pairs from a shooting foul, where the two attempts together represent one possession that ended in a foul. Some come as part of an and-one, where the player already used a possession on the made field goal. The 0.44 is empirical — it's the historical ratio that converts free throw attempts into the number of possessions they actually consumed, on average. League play has nudged it up and down by a couple of hundredths over the decades, but 0.44 is the canonical number.
The result is a percentage you can interpret directly: league-average TS% sits in the mid-fifties, anything in the low sixties is excellent, and anything north of .650 on real volume is elite. The relationship between TS% and offensive value is much tighter than the relationship between FG% and offensive value, especially across players who take very different shot profiles.
The Steph Curry effect
Curry's emergence is the cleanest empirical case for why FG% had to be replaced. He has had multiple seasons where his FG% looked unremarkable — high forties, occasionally just over .500 — while his TS% sat near .660 and he was, by every credible plus-minus measure, the most valuable offensive player alive. The FG% column was reading him as a good-not-great shooter. The TS% column was reading him correctly, because it knew that two of every five shots he took were worth fifty percent more than a two-pointer, and that he was hitting them at a rate the rest of the league could not.
The same logic, in reverse, applies to a different kind of player. Big men who shoot 60% from the field have always sounded efficient, and traditional FG% rewards them lavishly. But many of those same players have FT% under .700 and never shoot threes. Their TS%, which accounts for the free throws they miss but ignores the threes they don't take, often lands a couple of ticks below their FG% — closer to the league average than the headline number suggests.
What TS% still doesn't capture
Shot quality, mostly. TS% knows the value of a make but not the difficulty of the shot. A wide-open corner three counts the same as a contested step-back from twenty-eight feet. A dunk on a switched center counts the same as a putback on a defensive rebound. Modern tracking-data measures — most notably "shot quality" models that estimate the league-average conversion rate of a given shot from its location, defender distance, and shot clock — are the next layer beyond TS%, and they are mostly proprietary. The public box score still doesn't include them.
TS% also says nothing about creation. Two players can shoot the same TS% with one of them doing it on catch-and-shoot threes assisted by a teammate and the other doing it on contested pull-ups they had to create for themselves. The second skill is rarer and more valuable. TS% knows neither.
And, like every shooting stat, TS% rewards volume only indirectly. A bench player shooting .640 TS% on six shots a night is not adding the same value as a star shooting .610 on twenty. Per-possession rate stats need to be paired with usage to mean anything in a valuation context.
How to use it
The rule of thumb is short. When you're comparing two scorers, compare their TS%, not their FG%. If you only have FG%, mentally adjust upward for high three-point volume and downward for a player who never gets to the line. League average TS% for the current era is in the high fifties; anyone consistently below it is hurting their team's offense regardless of what their points-per-game looks like, and anyone consistently above it is helping more than the raw scoring number suggests.
FG% isn't useless. It's still the right column to look at for finishers around the basket, where almost every shot is a two, and it's a useful diagnostic when a player's TS% suddenly diverges from their FG%. But as a single summary number for offensive efficiency, it has been obsolete for at least a decade. The league knows it, every front office knows it, and increasingly the broadcasts know it. The box score, slow as ever, will catch up eventually.