A projected rebound count for every player on the slate, from his last eight games, and the probability of clearing each half-point line around it. There is no book line on this board — read the note below before you use the number. Graded 86-44 (66.2%) at the play tier 2026-27 season — against the model's own line, not a book's.
Graded record, Oct 20, 2026 – Nov 11, 2026. Full NBA performance page — every market, every pick counted. † Rebounds is graded against a line the model sets from the player's own recent average (no book carries one), so its hit rate is not comparable to a market price.
Every player on tonight's slate with the over/under probability against the model's own line (no book carries one), the last ten games, updated three times a day. See plans, or read how the projections are built below.
See plans →Our odds feed carries no rebound lines, so this board does not compare a projection with a book. For each player it takes his last eight games (20+ minutes), computes the mean and spread of his rebounds, and sets the line itself: the mean rounded to the nearest half, with a floor of 2.5. The probability on the card is the chance he clears that self-set line under a normal distribution, and the alt lines (members) are the same calculation one and two rebounds either side.
That makes it a projection, not a pick. The right way to use it is to put the projected number (and, for members, the alt-line probabilities) against the line your own book posts. If your book has a starting centre at 9.5 and the card projects 11.4 with 74% over 10.5, that is information; the card's own "over 11.5" call is not.
The performance page shows rebounds at 196-87 (69.3%) last season, and that number is real — every play was settled against the box score. But it is the model against its own line. Because the line floors at 2.5, most graded plays were "under 2.5" on bench players who average about two rebounds: 251 of the 283 plays were unders, and the overs went 16-16. Nobody is offered that bet at -110. We show the number with a dagger everywhere it appears for that reason, and we do not rank rebounds among the markets that beat a price.
A version of this board against real book lines, with minutes, matchup and lineup inputs, is queued. Until it exists, this page says what it is.
Each card shows the self-set line, the model's over/under split around it, the last-8 average and, under Details, a bar per game for the last ten (green cleared the line shown, red did not); members also see the alt lines. The bars are the useful part: they show a player's real range, which is what you need against a book line. Injured and doubtful players are hidden by default; the toggle shows them.
Props are one night at a time; for the season ahead, the NBA season predictions project the standings and each team's win total from a simulation of the full schedule.
Points, rebounds, assists, double-doubles and triple-doubles, plus live in-game trackers. MLB, NFL and NHL included.
See plans