An NBA total is a bet on the combined final score of both teams. The sportsbook posts a number (228.5 is a typical example), and you bet over if you think the two teams combine for 229 or more, or under if you think they combine for 228 or fewer. Both teams are on the same side of the bet. You're not picking a winner, just whether the game goes high-scoring or low-scoring.
This guide is for people new to NBA betting. I'll cover how totals actually work, the five inputs that move every line, what rest does to scoring, three common beginner mistakes, and a worked example. By the end you'll know more about totals than 80% of public bettors.
Standard pricing on a total is -110 on each side. That means you bet $110 to win $100, and the sportsbook keeps roughly 5% as their cut (the vig). To break even over time, you need to win 52.4% of your bets. If you can hit 55%, you're profitable. If you hit 53%, you're treading water. Below 52.4%, you're losing.
This is why "I went 6-4 last week" doesn't mean you're winning. 6-4 is 60%, which sounds great. Over 100 bets, you need 53+ wins. The book's edge is built into every line you take, so picking which side is right isn't enough; you have to pick sides that beat the implied probability the line creates.
Sportsbooks set NBA totals using a calculation that boils down to five inputs. Most casual bettors think they're picking based on "team X is hot." The book is already pricing the matchup using these inputs and nothing else.
Pace is the number of possessions a team gets per game. A fast-pace team like the Indiana Pacers averages around 105 possessions; a slow-pace team like the Memphis Grizzlies averages closer to 96. More possessions equals more chances to score. The book multiplies both teams' pace numbers (with some adjustment for each defense's ability to slow the game) to arrive at the expected possession count for the matchup. This is roughly 60% of the total.
How many points per 100 possessions each team scores. Top NBA offenses are around 118-120 points per 100 possessions; weak offenses are 108-112. Combined with the possession projection, this produces the raw scoring estimate.
How many points per 100 possessions each team allows. Elite defenses hold opponents to 108-110; weak defenses allow 118+. Books blend each team's offensive efficiency with the opponent's defensive efficiency to land on a realistic per-possession scoring rate.
A team on the second night of a back-to-back shoots roughly 1.5 percentage points worse on threes and gives up about 2 extra points per game on defense. A team coming off two days of rest plays sharper. The book bakes this in but doesn't always price it correctly when both teams are on a back-to-back, which is where the under has an edge.
Books regress to the season-long average more slowly than they should. If a team has shot 32% from three over the last 6 games (well below their 36% season average), it's about half noise and half adjustment. The model knows this and weights recent form lightly; casual bettors overweight it and get burned.
The fatigue effect in the NBA is bigger than people think. Specifically:
A team on the second night of a back-to-back shoots 1.0-1.5 percentage points worse from three. That sounds small. Across 35 three-point attempts (typical NBA team), it's roughly 1 fewer made three, which is 3 fewer points of offense. Combined with sharper defense from the rested side, the swing is 4-6 points per game from fatigue alone.
If BOTH teams are on a back-to-back, the under is the cleanest play. Books often set the line at the season-average total because they're worried about looking foolish if the game goes wild. The actual scoring usually lands 3-5 points below the season average. The Bookie Bullies model flags these games every morning and prices the projected total 2-4 points below the book's number.
Travel matters less than rest, but it matters. A West Coast team flying in for a 7 PM Eastern game effectively plays with a 3-hour body clock disadvantage. The first quarter usually shows it in shooting percentages. Books don't price this aggressively; it's a small but real edge over a sample of 30-40 games.
The Suns vs Warriors at 8 PM Saturday isn't more likely to go over because it's the marquee game. If anything, marquee games attract over money because the public loves overs on big games, which moves the line up. The fair total might be 232; the marquee bump pushes it to 234.5. Now the over has to clear a wider gap. Most casual bettors are paying tax on the entertainment value of the matchup.
"The Nuggets just dropped 138 yesterday, they're hot." NBA shooting variance is enormous. A 138-point game has nothing to do with whether the next game goes high-scoring. Books regress to season-average and so should you. If you're betting based on the last 1-2 game results, you're betting on noise.
This is the single largest mispriced factor in NBA totals. Most bettors check whether a star player is available but not whether the team is on a back-to-back. Even with the star playing, a fatigued team gives up real points and scores real fewer points. If you only do one thing differently after reading this, check the rest schedule before every NBA total bet.
Say tonight's game is Pacers (105 pace, second-best offensive efficiency, average defense) vs Magic (98 pace, average offense, top-5 defense). Pacers are home off two days of rest; Magic just finished a road back-to-back.
Naive math: Pacers' season scoring against average teams: 118 points. Magic's season scoring: 108. Combined naive total: 226.
Book's total: 225.5. The book applied a small upward adjustment for the Pacers' home advantage and a small downward adjustment for the Magic's recent low-scoring stretch. Roughly market-correct.
Now add the fatigue adjustment. Magic on the second night of a back-to-back: subtract 2 points from their offense (worse shooting), subtract 1 point from their defense (they give up more). Net: total moves down 1 point, to 224.5.
If the book's posted line is 225.5 and the fatigue-adjusted projection is 224.5, the under has roughly 1 point of edge. That's about a 53% true probability against the book's 50%. Profitable over time, not a lock on any one game.