How to Read an NBA Box Score: Every Stat Explained

A box score is the most complete record of a basketball game that fits on one screen, and most people who look at one read about four of its columns. That is a shame, because almost every argument fans have about a game has already been settled somewhere in that table. Did he actually have a good night, or did he just score a lot? Was the bench the problem? Did the rebounding gap decide it? The answers are all there, but they are hidden behind twenty abbreviations, several of which mean nearly the same thing and one of which is routinely quoted to mean the opposite of what it measures. This guide walks the whole line, column by column, explains which numbers are trustworthy and which need context, and flags the ones that did not exist for the first three decades of NBA history.

Stylized illustration for How to Read an NBA Box Score

The Two Halves of a Box Score

Every box score splits into two parts, and knowing which half you are in prevents most misreadings.

The first half is counting stats: points, rebounds, assists, steals, blocks, turnovers, fouls. These are literally tallies of things that happened. A rebound is a rebound. Nobody has to interpret it.

The second half is rate and efficiency stats: field goal percentage, true shooting, plus/minus, and the derived ratings you see on Basketball Reference. These are calculated from the counting stats, and every single one of them encodes an assumption. That is not a criticism, it is just the thing to remember when someone quotes one at you as though it were a measurement.

The most common analytical mistake in basketball is treating a rate stat as though it were a count, or a count as though it were a rate. A player who grabs 4 rebounds in 12 minutes and one who grabs 9 in 38 are not doing the same job, and neither raw number tells you that on its own.

Minutes: The Number That Contextualises Every Other Number

MIN (or MP on Basketball Reference) is minutes played, and it is the first number you should look at, not the last. It has been tracked since the 1951-52 season.

Minutes turn every counting stat into a rate. Eighteen points is a good night in 22 minutes and a quiet one in 41. When you see a line that looks strange in either direction, minutes usually explain it before anything else does.

Minutes also encode information the coach is giving you for free. A starter who plays 14 minutes was either in foul trouble, hurt, or benched, and the fouls column will usually tell you which. A reserve who plays 34 has been trusted with something.

Points, and Why FG% Lies

PTS is points. FGM and FGA are field goals made and attempted, and FG% is simply the first divided by the second.

Field goal percentage is the most quoted efficiency number in basketball and one of the worst, because it treats every made shot as equal when they have not been equal since the 1979-80 season, when the NBA adopted the three-point line. A guard who goes 5-for-12 with four threes has scored 14 points. A centre who goes 6-for-12 with none has scored 12. The centre has the better field goal percentage and the worse night.

That single flaw is why the two stats below exist.

eFG% and TS%: The Two Fixes

eFG%, effective field goal percentage, adjusts for the fact that a three is worth more. The formula is straightforward: (FG + 0.5 × 3P) / FGA. A made three counts as one and a half made shots, because it is worth one and a half times as much.

Basketball Reference gives the cleanest illustration of why this matters. Suppose Player A goes 4-for-10 with two threes, and Player B goes 5-for-10 with none. Both scored 10 points from the field. Their field goal percentages are 40% and 50%, which makes B look better. Their effective field goal percentages are both 50%, which is the truth.

TS%, true shooting percentage, goes one step further and folds in free throws. The formula is PTS / (2 × TSA), where true shooting attempts are FGA + 0.44 × FTA. That 0.44 is an estimate of how many free throw attempts actually end a possession, since some come in pairs and some come from and-ones.

True shooting is the single best one-number answer to "did he score efficiently tonight". It is the number to reach for when someone tells you a player who took 25 shots had a great game.

Rebounds: Two Different Skills in One Column

REB or TRB is total rebounds, tracked since the 1950-51 season. It is split into OREB (offensive) and DREB (defensive), and these two are not the same skill wearing different hats.

Defensive rebounds are mostly a matter of position and size, and on many possessions they are uncontested. Offensive rebounds require going back into traffic against players who are already between you and the rim, and a team that generates them is buying itself extra possessions.

The split is available separately only from the 1973-74 season onward. Before that, box scores carry a total and nothing else, which is worth remembering the next time you see a rebounding leaderboard that seems to have forgotten an entire era.

Assists, Turnovers, and the Ratio That Matters

AST is assists. TOV is turnovers, and this is the one that surprises people: turnovers were not tracked as an official statistic until the 1977-78 season. Every box score before that is silent on them.

Individually, both numbers are noisy. An assist depends on whether a teammate makes the shot, which is not entirely in the passer's control. A turnover can be a bad pass or an ambitious one.

The ratio is where the signal is. Assist-to-turnover ratio tells you whether a player is creating more than he is giving away, and for a lead guard it is close to a job description. It is also one of the few box score numbers that is genuinely stable across eras, because both halves scale together.

Editorial illustration: How to Read an NBA Box Score

Steals and Blocks: The Stats That Start in 1974

STL and BLK have only been official since the 1973-74 season.

This one fact quietly breaks more basketball arguments than any other. Bill Russell has zero career blocks in the record book. So does Wilt Chamberlain, and both of them were, by every contemporary account, among the most prolific shot blockers who ever lived. The stat simply did not exist while they played.

So when you see an all-time blocks list, you are not looking at the best shot blockers in history. You are looking at the best shot blockers since 1974. The same applies to steals, and to any leaderboard built on either.

Within the modern era both numbers reward watching the whole possession. A block that sails out of bounds returns the ball to the offence. A steal in the passing lane is worth more than the box score line suggests, because it usually starts a fast break.

Personal Fouls: The Column That Predicts the Fourth Quarter

PF is personal fouls, and it is the most forward-looking number on the page.

Six fouls disqualify a player. A starter carrying four early in the third quarter is a starter whose coach is about to make a decision, and the rest of the game will be shaped by it. Read the fouls column alongside minutes and a lot of otherwise baffling rotations become obvious.

Plus/Minus: The Most Misread Number on the Page

+/- is the difference between the team's scoring and the opponent's scoring while that player was on the court. If your team outscored theirs by nine in your 30 minutes, you are +9.

It is the most intuitive number in the box score and the one most often misused, because in a single game it says at least as much about the other four players as it does about the one it is attached to. A reserve who happens to share the floor with the best lineup on the team will post a healthy plus/minus while doing very little. A star dragging a bench unit through the second quarter can post a bad one while playing well.

Over a season, plus/minus and its derivatives become genuinely informative. Over one game, treat it as a description of what happened while someone was out there, not a verdict on him.

Game Score: One Number for the Whole Line

GmSc, or Game Score, is John Hollinger's attempt to compress an entire box score line into a single figure. It weights every counting stat and subtracts the bad ones.

The useful part is the scale, which was designed to feel like points: roughly 10 is an average performance and 40 is an outstanding one. If you want a quick read on whose night was actually the best without comparing eight columns, Game Score is the fastest honest answer.

The Team Row, and the Gaps That Decide Games

Underneath the player lines sits a team row, and it is the fastest way to work out why a game went the way it did.

Three gaps do most of the explaining. The rebounding gap tells you who got more chances. The turnover gap tells you who gave chances away, and each one is worth roughly a possession, which at modern scoring rates is worth a little over a point. The free throw attempt gap tells you who spent the night attacking and who settled.

A team can lose the shooting battle and still win comfortably if it wins all three, because all three are about volume of possessions rather than efficiency within them. This is also why blowouts often look strange in the individual lines: once a game is decided, the closing minutes are played by people whose numbers will be quoted for years without the context that they were accumulated at garbage time.

The other team-row habit worth building is checking whether the points actually add up. In older records they sometimes do not, because the box score you are reading was reconstructed decades later from newspaper accounts.

Reading a Box Score From 1962

Half the fun of old box scores is how much is missing, and knowing what is missing is what stops you drawing a wrong conclusion.

Work backwards through the tracking dates and a clear picture emerges. Before 1950-51 there are no total rebounds. Before 1951-52 there are no minutes, so nothing can be turned into a rate. Before 1973-74 there are no steals, no blocks, and no split between offensive and defensive rebounds. Before 1977-78 there are no turnovers. Before 1979-80 there are no three-pointers, because there was no line. Games started only appear from 1982.

So a 1962 box score gives you points, field goals, free throws, rebounds, assists and fouls, and that is the lot. A player from that era cannot have a recorded steal, and a player from before 1980 cannot have a three, no matter how far out he shot from.

This is the single best defence against a certain kind of bad statistical argument. When someone shows you a list of the top ten in a category and every name played after 1980, the honest question is not who is missing. It is whether the statistic existed for the people who are missing.

What the Box Score Cannot Tell You

For all of that, the box score is a record of events, not of basketball. It has no column for the screen that freed the shooter, the rotation that forced the pass, or the shot a defender talked someone out of taking. Every possession involves ten players and the box score credits at most three of them.

That is the real argument for watching games as well as reading them. The table tells you what happened. It does not always tell you who caused it.

Closing illustration for How to Read an NBA Box Score