Every NBA game, projected before tip-off.
Win probability, a spread lean, and a projected total for every game on the nightly slate, each shown next to the market price with a confidence tier. The model ran last season and returns for 2026-27 when the schedule opens in late October.
Win probability, a spread lean, and a projected total
- Win probability per side: each team's modeled chance to win outright, shown next to the market's moneyline price.
- Spread lean: the side the model prefers against the number, since winning the game and covering the margin are two different questions.
- Projected total: an over or under lean built from expected pace and matchups, framed by a projected final-score range.
- Schedule & rest flags: back-to-backs, road trips, and rest days are read as inputs, so a tired team's number reflects the spot it is in.
- Confidence tiers: an Elite, Strong, Good, or Moderate label on every lean, so the slate sorts by conviction rather than raw percentage.
Three boards for the player markets
The game model sets the table. The player boards below take it down to the box score, from the marquee points model to a full browsing board and a numbers-first projections view.
The marquee scoring board: every player's projected points read against the line, framed by minutes, pace, and role.
Preview →Props BoardSoonThe whole night's player markets in one place: points, rebounds, assists, threes, and PRA, each line beside the model's lean.
Preview →Prop ProjectionsSoonThe numbers view: a modeled figure for each player market, set against the line and updated for rest and rotation news.
Preview →Pace, matchups, and the schedule
An NBA game projection starts with pace, the number of possessions a matchup is likely to produce. Two fast teams create more trips down the floor, and more trips lift both the projected total and the range of scores each side can post. A slower, half-court matchup pulls the total the other way before a single shot goes up.
Then the matchups on both ends. How a team's offense stacks up against the defense across from it, who guards whom, and where the mismatches sit. Those shape both the win probability and how comfortable the model is with a spread lean, described here at a high level, the same way our methodology page frames the MLB model.
Finally the schedule. The NBA plays a dense calendar, and rest is a real edge. A back-to-back, a fourth game in six nights, or a long road trip can quietly move a number, so the model reads those spots rather than treating every team as fully rested. Every lean still carries a confidence tier, read exactly the way it is on the baseball boards.
NBA projections, answered
When do the NBA projections come back for 2026-27?
The NBA model ran last season and returns for the 2026-27 campaign, which tips in late October. This page is the preview. The live game and player boards appear on the site as the season opens, then refresh nightly through the schedule.
What is an NBA game projection?
It is the model's read on a single matchup: a win probability for each side, a lean on the spread, and a projected total that drives an over or under call. Each number sits next to the market price, so the gap we call edge is easy to see.
How do back-to-backs and rest change a game projection?
A team on the second night of a back-to-back, or deep into a road trip, tends to play with heavier legs than a rested opponent, which can dull both its scoring and its defense. The model treats schedule and rest as plain inputs, so a tired team's number reflects the situation rather than its full-strength form.
What is the difference between a spread lean and a win-probability lean?
Win probability is only about who wins the game outright, at whatever price each side carries. The spread adds a margin the favorite has to cover, so a team can win the game and still fail to cover. The board shows both, since the model can favor a team to win while preferring the other side against the number.
What do the confidence tiers mean on an NBA game?
Every lean carries an Elite, Strong, Good, or Moderate label that describes how wide a gap the model sees between its number and the market's. A higher tier reflects a wider gap, not a promise, and the same taxonomy is used across every board on the site.
Be ready for opening night
Start with the live MLB boards today, and the NBA projections will be waiting the moment the season tips off for 2026-27.