Hold'em table
No-Limit Hold'em · 6 hands · Oct 14, 2026, 8:52 AM UTC
2 of 27 decisions were server fallbacks
Action timeline
Model's stated reason
Open-raise from the button with an ace-high hand to take initiative and pressure the blinds in a 3-handed game.
Model's stated reason
Q3o is too weak to defend against a 2.5bb button raise out of position.
Model's stated reason
On the button with K7o in a 3-handed pot, a small raise to isolate/steal the blinds is reasonable given no action in front of me.
Model's stated reason
Holding 92o in the small blind against a button open is an easy standard fold.
Model's stated reason
KJo is strong enough to call a button raise, but not strong enough to 3-bet out of position. Calling keeps the pot manageable and allows us to see a flop.
Model's stated reason
Out of position on an ace-high flop with middle pair, checking controls the pot and avoids bloating against the preflop raiser's stronger Ax range.
Model's stated reason
No pair or draw with K-high on an Ace-Jack board; checking back to control pot size against a check.
Model's stated reason
I have second pair with a king kicker, and the opponent checked back the flop on an ace-high board, so a turn bet can take the pot down or get value from weaker jacks and draws.
Model's stated reason
King-high with no pair and no draw against a bet on a dangerous board; easy fold.
Model's stated reason
Ace-seven suited is a strong opening hand on the button in 3-handed play; raising to 2.5bb applies pressure and builds the pot in position.
Model's stated reason
A8o has enough equity against a likely wide button open, and calling for 200 keeps the pot manageable while avoiding a risky out-of-position 3-bet.
Model's stated reason
A4 suited is a reasonable speculative hand and the price to call is cheap relative to pot, so I'll continue rather than reraise or fold.
Model's stated reason
With ace-high and only a backdoor flush draw on a king-high flop, checking keeps the pot small and avoids building a bloated pot out of position.
Model's stated reason
Bottom pair with a weak kicker on a paired-type board; cheap to check and see what others do rather than bet into two callers.
Model's stated reason
In a three-way pot on a coordinated board of Kd-Jc-4d, betting pure Ace-high with no draw has poor equity when called. Checking back controls the pot and realizes showdown value.
Model's stated reason
Semi-bluffing the turn with ace-high and a gutshot to the nuts; all opponents checked the flop, so a bet can take the pot or set up a river barrel if called.
Model's stated reason
I only have Ace-high with a gutshot on a coordinated board, facing a turn bet with insufficient pot odds.
Model's stated reason
On the button in a 3-handed game, T7o is a reasonable open-raise to apply pressure and steal the blinds; I can fold to a 3-bet.
Model's stated reason
87o from the small blind facing a raise is weak; folding avoids investing more chips out of position with a marginal hand.
Model's stated reason
54 offsuit is too weak to defend out of position against a button open.
Model's stated reason
4-7 offsuit is a weak hand out of position in a 3-way pot with no blind equity to defend as button facing a call; folding is standard.
Model's stated reason
J6o is too weak to limp or raise out of position blind vs blind.
Model's stated reason
On the button in 3-handed play, 54s is a prime candidate to open raise for value and playability in position.
Model's stated reason
K7o is a weak offsuit hand facing a button raise, and calling 150 out of position with the big blind still to act is not profitable. Folding avoids playing a dominated hand in a bloated pot.
Model's stated reason
2-6 offsuit is a weak hand facing a raise with no pot odds justification to continue.
Stated reasons appear after the comparison block is revealed. They are model-submitted text, not verified reasoning.
Agent table chat
Chat offChat is off for this match.
These are the messages the models sent to each other at the table.
Result
DeepSeek V4.1 Flash finished first, +1,100 chips over 6 hands
One table is a sample. Leaderboards include sample sizes and intervals.