Simbot

Simbot is a derivative of Poki. Instead of using Poki's formula-based betting strategy to make decisions, it runs simulations of the hand to completion, and determines the most profitable action to make based on the outcomes.

Configuration Options

Pre-flop Settings:

There are five tightness levels to choose from (tight, moderate, loose, loose-passive, and small stakes advisor). See Preflop Limit Tightness Settings for detailed information on these settings.

The pre-flop deception level lets you choose how much randomization it will use in pre-flop actions. If you set the slider towards honest it will raise with strong hands, call with moderate strength hands, and fold everything else. With high deception (tricky) it will have a higher chance of calling strong hands and even some very weak hands, and will raise some moderate hands.

Post-flop Settings:

For post-flop play there are two sliders. The first slider controls how much it will use its opponent modeling information in assigning cards for the simulation. This sets the style to either 'play the man' or 'play the cards'.

The second slider controls what types of actions the simulated opponents will do during the trails. If set to Call, the simulation becomes a basic all-in equity simulation. If the slider is moved towards Predict then it will use its opponent models to try and predict what each opponent would do in each simulated situation.

References

Improved Opponent Modeling in Poker.
Aaron Davidson, Darse Billings, Jonathan Schaeffer, and Duane Szafron,
Proceedings of the 2000 International Conference on Artificial Intelligence (ICAI'2000), Las Vegas, Nevada, pp 1467-1473, 2000.
http://www.cs.ualberta.ca/~darse/Papers/ICAI00.html

Using Probabilistic Knowledge and Simulation to play Poker.
Darse Billings, Lourdes Peņa, Jonathan Schaeffer, and Duane Szafron,
Proceedings of the Sixteenth National Conference of the American Association for Artificial Intelligence (AAAI-99), 1999.
http://www.cs.ualberta.ca/~darse/Papers/AAAI99.html

Back to Opponents