Season record

Every graded NBA AI prediction of the 2026–27 season so far, for points, rebounds and assists. Every figure shows its count; no date or player is left out.

All graded rows of the season, both measures side by side

Lean hit rate

48.5% 554 of 1,143 graded

Prediction accuracy · within 10 %

15.5% 177 of 1,145 scored

  • 1,143 graded points predictions
  • 11 dates · Oct 28 – Nov 9, 2026
  • 20 void (did not play or postponed) not counted

By position

most accurate first

Loading the season's graded rows…

Position = the model group that predicted him: guard, forward or center, or the all-positions model. Ordered by prediction accuracy (within 10 %).

By starter / bench

most accurate first

Loading the season's graded rows…

Starter or bench = his lineup slot that night. Ordered by prediction accuracy (within 10 %).

Where it is most accurate

bold = the figure the rule stands out on

Loading the season's graded rows…

Each rule is one of the built-in presets of the slate's advanced filters, measured on every graded row of the season for this stat; only rules with 30 or more graded rows are listed, highest lean hit rate first. A rule is a filter, not a recommendation.

Most accurate players

top 5 by within 10 %

Loading the season's graded rows…

Over the last 10 graded dates (Oct 30 – Nov 9); a player needs at least 3 scored games in that window (– qualify).

By date

a date opens that slate
DateGraded Lean hit Within 10 %
Oct 281450.0%7 of 147.1%1 of 14
Oct 3011851.7%61 of 11817.8%21 of 118
Oct 318355.4%46 of 8310.8%9 of 83
Nov 15244.2%23 of 529.6%5 of 52
Nov 220645.1%93 of 20615.4%32 of 208
Nov 419949.2%98 of 19914.6%29 of 199
Nov 52857.1%16 of 2817.9%5 of 28
Nov 613351.1%68 of 13318.8%25 of 133
Nov 79638.5%37 of 9617.7%17 of 96
Nov 810144.6%45 of 10115.8%16 of 101
Nov 911353.1%60 of 11315.0%17 of 113
Season1,14348.5%554 of 1,14315.5%177 of 1,145

How the record is counted