Results
NBA results Monday 5 October 2026: AI predictions review
NBA AI predictions review for Monday 5 October 2026: the models went 1 of 3, with two confident calls falling flat. Full results and model tallies inside.
Monday's three-game slate produced a 1 of 3 record for the ten-model consensus, a return that sits awkwardly against the confidence on display before tip-off. Two of the three calls were backed by near-unanimous majorities, and both of those majorities were beaten.
The biggest surprise came in Philadelphia, where 90% of the models expected the visiting New York Knicks to edge a tight one and instead watched the 76ers run away with a 120-97 win. The consensus had the game as a one-point margin; the actual margin was 23.

Monday's slate ended 1 of 3 for the ten-model consensus.
Where the AI got it right
The one clean hit came in Detroit, where the Pistons held off the Phoenix Suns 109-107. The consensus leaned Detroit's way with 80% of the ten models, and while the most common predicted scoreline of 112-106 overshot the final total, the winner was called correctly. Average confidence was 37%, so this was a pick made with modest conviction rather than certainty.
Detroit Pistons vs Phoenix Suns prediction
That is the full extent of the correct calls on the day. The other two fixtures on the board both went against the models, leaving the consensus with a single win from three attempts.
Where it got it wrong
Atlanta's visit from Memphis was the most emphatic miss of the night in terms of unanimity. All ten models picked the Hawks, at 100%, with a most common scoreline of 115-108 and an average confidence of 37%. The Grizzlies won 132-123 instead, a nine-point road victory that no model in the panel saw coming. A unanimous pick that fails is the hardest kind of result to explain away.
Atlanta Hawks vs Memphis Grizzlies prediction
Philadelphia was nearly as one-sided in the other direction. Nine of the ten models, or 90%, expected the Knicks to win, with a most common scoreline of 108-109 and the same average confidence of 37%. The 76ers won 120-97, a 23-point blowout that turned a projected one-point game into a rout. The models were not merely wrong on the winner; they were wrong about the shape of the contest entirely.
Philadelphia 76ers vs New York Knicks prediction
It is worth noting that the average confidence across all three fixtures was identical at 37%. The panel was not loudly proclaiming certainty on any of them, yet the distribution of picks was lopsided in two cases. Low confidence and high agreement turned out to be a poor combination on Monday.
Models of the day
Claude Sonnet 5 was the only model to finish above .500, going 2 of 3 and separating itself from a crowded middle. Six other models — GPT-5.6 Luna, DeepSeek V4 Pro, Qwen 3.8 Max, MiMo V2.5 Pro, Grok 4.6, Kimi K3 and GLM 5.3 — each finished 1 of 3.
At the bottom, Mistral Medium 3.5 and Gemini 3.8 Flash both went 0 of 3, missing every fixture on the slate. With only three games to work from, the spread between the best and worst model was two correct calls.
Full results
- NBA: Atlanta Hawks 123-132 Memphis Grizzlies — consensus Atlanta Hawks win (100%) WRONG
- NBA: Detroit Pistons 109-107 Phoenix Suns — consensus Detroit Pistons win (80%) RIGHT
- NBA: Philadelphia 76ers 120-97 New York Knicks — consensus New York Knicks win (90%) WRONG
A 1 of 3 day is a reminder that unanimous agreement among models is not the same as a safe pick. Tuesday's slate is already loaded, with fresh predictions from all ten models available on the today's NBA predictions page.
Next up: today's nba predictions from the same panel, and the results archive with every graded call. Forecasts are informational only.