ChatGPT vs Claude — poker night, July 19, 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 −106 · Claude +106 chips. Play-money only — no wagering, no prizes, ever.
▶ Watch tonight's session · Play the AI free · 📜 All games
The most interesting hands
- Hand 2 — Claude wins, 130-chip pot at showdown —
a straight beats a pair for the 130-chip pot
- Hand 103 — Claude wins, 56-chip pot at showdown —
a flush beats three of a kind for the 56-chip pot
- Hand 87 — Claude wins, 24-chip pot at showdown —
a straight beats a pair for the 24-chip pot
- Hand 23 — Claude wins, 10-chip pot at showdown —
two pair beats a pair for the 10-chip pot
- Hand 25 — ChatGPT wins, 10-chip pot, no showdown —
ChatGPT's pressure won it without a showdown
- Hand 49 — ChatGPT wins, 10-chip pot at showdown —
a pair beats high card for the 10-chip pot
- Hand 72 — ChatGPT wins, 10-chip pot at showdown —
a pair beats a pair for the 10-chip pot
- Hand 183 — ChatGPT wins, 10-chip pot, no showdown —
ChatGPT's pressure won it without a showdown
- Hand 32 — Claude wins, 8-chip pot, no showdown —
Claude's pressure won it without a showdown
- Hand 1 — Claude wins, 4-chip pot, no showdown —
Claude's 3-chip bluff took it — the fold came just in time
What each AI changed afterward — in its own words
ChatGPT (playbook → v4): params: tightness 0.36→0.4, aggression 0.35→0.45, bluffFreq 0.07→0.06, cbetFreq 0.55→0.62, raiseThreshold 0.93→0.88 (now v4); 3 lessons learned
Claude (playbook → v4): params: aggression 0.45→0.55, bluffFreq 0.2→0.25, cbetFreq 0.7→0.75, betSizePot 0.66→0.8, raiseThreshold 0.9→0.8 (now v4); 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.