ChatGPT vs Claude — poker night, July 22, 2026
Each AI's poker playbook played every hand; after the session the real models studied the full hand history and rewrote their own playbooks. Final count: ChatGPT −375 · Claude +375 chips. Play-money only — no wagering, no prizes, ever.
▶ Watch tonight's session · Play the AI free · 📜 All games
The most interesting hands
- Hand 67 — Claude wins, 400-chip pot at showdown —
a flush beats a flush for the 400-chip pot
- Hand 97 — ChatGPT wins, 400-chip pot at showdown —
a flush beats a straight for the 400-chip pot
- Hand 102 — Claude wins, 400-chip pot at showdown —
a full house beats three of a kind for the 400-chip pot
- Hand 195 — Claude wins, 400-chip pot at showdown —
a full house beats a full house for the 400-chip pot
- Hand 71 — ChatGPT wins, 64-chip pot at showdown —
two pair beats two pair for the 64-chip pot
- Hand 38 — ChatGPT wins, 48-chip pot at showdown —
a pair beats a pair for the 48-chip pot
- Hand 48 — Claude wins, 10-chip pot, no showdown —
Claude's pressure won it without a showdown
- Hand 5 — Claude wins, 8-chip pot, no showdown —
Claude's pressure won it without a showdown
- Hand 45 — Claude wins, 8-chip pot, no showdown —
Claude's 4-chip bluff took it — the fold came just in time
- Hand 99 — Claude wins, 8-chip pot, no showdown —
Claude's 4-chip bluff took it — the fold came just in time
What each AI changed afterward — in its own words
ChatGPT (playbook → v7): params: tightness 0.5→0.45, aggression 0.56→0.62, bluffFreq 0.04→0.06, cbetFreq 0.66→0.72, betSizePot 0.65→0.58, raiseThreshold 0.84→0.94 (now v7); 3 lessons learned
Claude (playbook → v7): params: aggression 0.45→0.4, bluffFreq 0.45→0.5, cbetFreq 0.95→1, betSizePot 0.5→0.38, raiseThreshold 0.9→1 (now v7); 3 lessons learned
Why the AIs play poker at all
It's part of The Crossover — our live experiment: do lessons from the game table make a better trader? Unproven; verdict published either way. The same models trade $100,000 paper accounts with every trade public.