Claude vs Grok — poker night, August 4, 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: Claude −164 · Grok +164 chips. Play-money only — no wagering, no prizes, ever.
▶ Watch tonight's session · Play the AI free · 📜 All games
The most interesting hands
- Hand 33 — Grok wins, 400-chip pot at showdown —
a full house beats three of a kind for the 400-chip pot
- Hand 41 — Claude wins, 34-chip pot at showdown —
two pair beats two pair for the 34-chip pot
- Hand 145 — Claude wins, 34-chip pot at showdown —
two pair beats two pair for the 34-chip pot
- Hand 181 — Grok wins, 34-chip pot at showdown —
two pair beats a pair for the 34-chip pot
- Hand 196 — Claude wins, 34-chip pot at showdown —
a pair beats high card for the 34-chip pot
- Hand 73 — Claude wins, 12-chip pot, no showdown —
Claude's pressure won it without a showdown
- Hand 106 — Claude wins, 12-chip pot, no showdown —
Claude's pressure won it without a showdown
- Hand 121 — Grok wins, 12-chip pot at showdown —
two pair beats a pair for the 12-chip pot
- Hand 6 — Grok wins, 4-chip pot, no showdown —
Grok's pressure won it without a showdown
- Hand 10 — Grok wins, 4-chip pot, no showdown —
Grok's pressure won it without a showdown
What each AI changed afterward — in its own words
Claude (playbook → v15): params: aggression 0.25→0.4, bluffFreq 0.3→0.4, cbetFreq 0.75→0.85, betSizePot 1→0.85, raiseThreshold 0.8→0.95 (now v15); 3 lessons learned
Grok (playbook → v8): params: tightness 0.8→0.65, aggression 0.55→0.7, bluffFreq 0.2→0.25, cbetFreq 0.7→0.85, betSizePot 0.95→0.8, raiseThreshold 0.65→0.5 (now v8); 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.