Results
NBA results Friday 20 November 2026: AI predictions review
AI models called 3 of 4 NBA games right on Friday 20 November 2026. See where the consensus hit, where it missed, and the top models of the day.
Friday night in the NBA delivered three comfortable covers and one sharp reminder that unanimity is not immunity. The ten-model consensus, run through TuringStats before tip-off, finished the slate with a 3 of 4 record, but the lone miss was a unanimous call that went the other way.
The surprise came in Minneapolis, where all ten models backed the Timberwolves to beat the Knicks. New York had other ideas, leaving the consensus red-faced on an otherwise profitable evening.

The consensus went 3 of 4 on Friday night.
Where the AI got it right
Memphis Grizzlies vs Golden State Warriors prediction: The models were not unanimous here—only 70% of the ten picked Memphis—but they were emphatic about the direction. The Grizzlies won 126–98, covering the most common projected score of 116–111 with room to spare. Average confidence sat at 57%, the lowest of any correct call on the night, which makes the result a useful reminder that a split consensus can still land cleanly.
Milwaukee Bucks vs Portland Trail Blazers prediction: This was as close to a lock as the models get. All ten picked Portland, with an average confidence of 74% and a most common score of 118–119. The Trail Blazers delivered a 124–114 win, covering the projected margin and then some. The 100% consensus was justified.
San Antonio Spurs vs LA Clippers prediction: Another unanimous call, and another comfortable cover. The models projected 118–107; the Spurs won 124–112. Average confidence was 75%, the highest of the night, and the result never seriously threatened to go the other way. San Antonio's margin was slightly larger than expected, but the direction was never in doubt.
Where it got it wrong
Minnesota Timberwolves vs New York Knicks prediction: The only blemish on the card, and a painful one for the consensus. All ten models backed Minnesota, with an average confidence of 62% and a most common score of 115–108. The Knicks won 102–99, a low-scoring affair that bore little resemblance to the projected shootout. The models were not just wrong on the winner; they were wrong on the pace. It is the kind of miss that stings precisely because it was unanimous—there was no dissenting voice to point to afterwards.
Models of the day
Nine of the twelve models listed on the day finished with a 3 of 4 record: GPT-6.1 Sol, Claude Opus 5.5, DeepSeek V4 Pro, Qwen 3.8 Max, MiMo V2.6 Pro, Grok 4.7, Kimi K3, GLM 5.3 and Meta Muse Spark 1.3 all got the Timberwolves game wrong and everything else right. The bottom three—Gemini 4 Argon, NVIDIA Nemotron 3 Ultra and Tencent Hy4—each finished 2 of 4, meaning they missed one additional game beyond the consensus error. No model went perfect on the night.
Full results
- NBA: Memphis Grizzlies 126–98 Golden State Warriors — consensus Memphis Grizzlies win (70%) RIGHT
- NBA: Milwaukee Bucks 114–124 Portland Trail Blazers — consensus Portland Trail Blazers win (100%) RIGHT
- NBA: Minnesota Timberwolves 99–102 New York Knicks — consensus Minnesota Timberwolves win (100%) WRONG
- NBA: San Antonio Spurs 124–112 LA Clippers — consensus San Antonio Spurs win (100%) RIGHT
For today's slate, including projections and model breakdowns, visit the 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.