Data Biased Robust Counter Strategies

bubbagump · **Posts:** 61 **Favourite Bot:** your

Data Biased Robust Counter Strategies

Name of Authors: Michael Johanson, Michael Bowling
Department of Computing Science, University of Alberta
March 2009

Abstract
The problem of exploiting information about the environment while still being robust to inaccurate or incomplete information arises in many domains. Competitive imperfect information games where the goal is to maximally exploit an unknown opponent's weaknesses are an example of this problem. Agents for these games must balance two objectives. First, they should aim to exploit data from past interactions with the opponent, seeking a best-response counter strategy. Second, they should aim to minimize losses since the limited data may be misleading or the opponent's strategy may have changed, suggesting an opponent-agnostic Nash equilibrium strategy. In this paper, we show how to partially satisfy both of these objectives at the same time, producing strategies with favourable tradeoffs between the ability to exploit an opponent and the capacity to be exploited. Like a recently published technique, our approach involves solving a modified game; however the result is more generally applicable and even performs well in situations with very limited data. We evaluate our technique in the game of two-player, Limit Texas Hold'em.

Download: http://webdocs.cs.ualberta.ca/~games/poker/publications/AISTATS09.pdf

indiana · **Posts:** 7731 **Favourite Bot:** V12

Related thread: http://pokerai.org/pf3/viewtopic.php?p=20456#p20456

iamnobody · **Posted:** Mon Nov 30, 2009 6:53 pm

Link in the first post is broken.

Here is a working link : http://games.cs.ualberta.ca/~bowling/pa ... istats.pdf

c2008 · **Posts:** 1115 **Favourite Bot:** Johnny #5

Question about the opponent model used in this paper... Am I mistaken, or are they just using opponent statistics as their model? For example, at any given decision node they assume the opponent plays with their action distribution, regardless of the hand he holds? If each observation is a full game, where the opp cards are revealed, how do they calculate with what probability the opp folds, and with what cards?

Quote:

3.3 OPPONENT BELIEFS
A belief about the opponent’s current strategy can simply
be encoded as a strategy itself. Even a posterior belief derived
from a complicated prior and many observations still
can be summarized as a single function mapping an information
set to a distribution over actions, the expected
posterior strategy1. In this work, we will mainly take a
frequentist approach to observations of the opponent’s actions
(although we discuss a Bayesian interpretation to our
approach in Section 7). Each observation is one full information
game of poker: both players’ cards are revealed.
The model of our opponent will consider all of the information
sets in which we have observed the opponent acting.

spears · **Posts:** 1239 **Favourite Bot:** my bot

I think they mean they were using hand histories where both players cards were known even if they folded.

Data Biased Robust Counter Strategies

Who is online