You have seen it a hundred times. Tatum finishes with 28 points and the broadcast flashes +14 next to his name. Your friend texts you that he is plus-twenty-two tonight and that is crazy. You nod like you know exactly what that means. Most people do not.
Plus-minus is the most cited and most misunderstood number in basketball. It looks simple. It sounds definitive. And it is hiding more than it reveals. If you want to get smarter about the game, or if you are trying to predict player stats before the tip, this is one of the first traps you need to avoid.
The Simple Version
Plus-minus measures point differential while a player is on the court. Nothing more. If the Celtics score 102 points and give up 94 during the 34 minutes Tatum plays, his plus-minus is plus-eight. His team scored more than the opponent while he was out there. That is the entire calculation.
The stat does not care who he played with. It does not care who guarded him. It does not care whether he hit three clutch threes or stood in the corner while his teammates went on a 12-0 run without touching the ball. Tatum gets the same plus-minus either way.
This simplicity is why broadcasters love it. It is one number that sounds like a verdict. But definitive and accurate are different things. A verdict that ignores evidence is just a loud guess.
Why Plus-Minus Lies to You
The problems stack fast. Lineup context is the biggest one. A bench player on the Warriors might play almost exclusively with Stephen Curry. His plus-minus looks incredible because Curry warps the floor. Defenders rotate toward him on every possession, leaving open lanes and clean looks that the bench player did not create. Put that same bench player on a lottery team with no gravity threat and his plus-minus collapses. The stat gives him credit for something Curry did.
Opponent context matters too. If a coach rests his star every time the opposing star is on the floor, the backup gets padded numbers against weaker second units. The player did not change his game. The schedule did the work. Plus-minus treats both situations the same.
Then there is the sample size problem. Single-game plus-minus is basically noise. A player can finish plus-eighteen because his team caught fire from three during an eight-minute stretch he happened to be sitting on the court for. He might have played his worst game of the month. The shots just fell while he was nearby. Over a full season the randomness smooths out some, but the context problems never go away. Season-long plus-minus still blends lineup quality, opponent scheduling, and coaching patterns into one misleading headline.
The Basketball-Reference glossary is careful to note this limitation. Even the most respected statistical databases treat raw plus-minus as descriptive, not evaluative. It describes what happened to the team. It does not evaluate what the player did.
The Stats That Actually Work
Serious analysts moved past raw plus-minus years ago. The modern replacements try to isolate individual contribution from team circumstance. Box Plus/Minus, commonly called BPM, starts with box score stats and estimates a player's impact per 100 possessions. It is not perfect, but at least it attributes credit to the player who actually scored, assisted, stole, or rebounded. You can find BPM data for every NBA player on Basketball-Reference.
Value Over Replacement Player, or VORP, extends BPM into a season-long counting stat. It asks a clean question: how much total value did this player provide compared to a generic replacement-level player? It is useful for comparing players with different minutes loads.
On-off splits take a different angle. They compare how the team performs with the player on the court versus on the bench. This is closer to the original plus-minus idea but explicitly accounts for lineup context. If the team is plus-twelve per 100 possessions with Tatum and plus-two without him, the ten-point gap tells you something real about his impact. Of course, on-off numbers can also be noisy if a player's minutes are heavily correlated with other starters.
The gold standard for front offices is Regularized Adjusted Plus-Minus, or RAPM. It uses multi-season samples and ridge regression to separate individual impact from lineup noise. You will not see RAPM on a broadcast because it requires computing power and years of historical data that live models handle behind the scenes. But it is what teams actually use when they are deciding who to pay. For a deeper technical breakdown, the NBA.com stats glossary covers the methodology behind adjusted plus-minus models.
If you want to understand how stats like these relate to prediction, our breakdown of regression to the mean in the NBA explains why even good stats can mislead you if you treat a hot streak as a permanent upgrade.
How to Use Plus-Minus Without Getting Fooled
Raw plus-minus is not useless. It is just a team stat wearing a player costume. The trick is knowing when to use it and when to ignore it.
Use it to describe lineups, not individuals. If the Celtics starting five is plus-twelve per 100 possessions over 400 minutes together, that is a real signal about lineup chemistry and fit. It does not tell you which of the five is driving the success, but it tells you the combination works.
Use it over large samples, not single games. A seventy-game plus-minus tells you something about a player's season. A single-game number tells you almost nothing. If someone tweets that a rookie was minus-fourteen last night, that is weather, not climate.
Always pair it with context. Who was on the floor with him? Who was on the other side? Was the game already decided when he got his minutes? Garbage time stats are not real stats. They are participation ribbons.
The fans who predict player performance well are not the ones who memorize the most numbers. They are the ones who know which numbers matter and which ones just look important. Plus-minus looks important. Most of the time, it is not.
For another stat that gets misused constantly, read our explainer on true shooting percentage. TS% is actually useful when you know what it measures. The key is the same for every advanced stat: understand the formula before you trust the conclusion.
Common Questions
What does plus-minus mean in basketball?
Plus-minus measures the point differential while a specific player is on the court. If a player's team scores 102 points and allows 94 during the 34 minutes he plays, his plus-minus is +8. It is a team stat applied to an individual, with no adjustment for lineup quality, opponent strength, or the player's direct involvement in the scoring.
Is plus-minus a good stat for measuring player performance?
Raw plus-minus is a poor standalone measure of individual performance because it blends team success with individual contribution. A player can have a strong plus-minus simply by sharing the court with great teammates or weak opponents. Analysts prefer adjusted variants like Box Plus/Minus (BPM) or on/off splits that attempt to isolate individual impact.
What is the difference between plus-minus and Box Plus/Minus?
Plus-minus is a raw team point differential during a player's minutes. Box Plus/Minus (BPM) is an estimate of a player's contribution per 100 possessions, derived from box score stats and scaled to league averages. BPM attempts to credit the player who actually produced the stats rather than giving everyone on the court equal credit.
Why do NBA broadcasters use plus-minus so often?
Broadcasters use plus-minus because it is simple to calculate in real time and easy to explain in one sentence. It provides a single number that sounds definitive, even though the underlying context is complex. It is a production-friendly stat, not an analytically rigorous one.
What is the best advanced stat for evaluating basketball players?
There is no single best stat, but serious analysts rely on a blend. Box Plus/Minus and Player Efficiency Rating (PER) are common box-score estimates. On/off splits show team performance with and without the player. Regularized Adjusted Plus-Minus (RAPM) uses multi-season data and regression to isolate individual impact from lineup noise. Each measures something slightly different.
Plus-minus promises a clean answer to a messy question. Real basketball impact does not fit in one number. The smarter you get about stats, the more you learn to distrust the easy ones. If you want to test your reads against other fans who care about the same thing, GAGE is built for that. One game. Same stats for everyone. No house. Just your knowledge against theirs.