This thread is to build a ranklist of the best poker papers.

Every member can vote for up to 3 papers: which will contribute one point for the respective papers.

Just reply to this thread with your 3 paper selections. I will maintain top20 ranklist (with reference to the respective threads) in this post. You can change your vote too (as long as you don't do it too often, just repost and I will recalculate).

Current ranklist
-----------------------------------------------------------------
6- Algorithms and Assessment in Computer Poker, Darse Billings, 2006.
6- Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment, 2002, Aaron Davidson
5- Robust Strategies and Counter-Strategies: Building a Champion Level Computer Poker Player, Michael Johanson, M.Sc. thesis, October 2007.
3- Regret Minimization in Games with Incomplete Information, 2008, Zinkevich, Johanson, Bowling and Piccione, U of A
3- Opponent Modelling and Search in Poker, Terence Schauenberg, M.Sc. thesis, 2006.
1- Strategy Evaluation in Extensive Games with Importance Sampling, 2008, Michael Bowling, Michael Johanson, Neil Burch, Duane Szafron
1- Approximating Game-Theoretic Optimal Strategies for Full-scale Poker, 2003, Billings et all.
1- Computing Robust Counter strategies (Johanson and Zinkevich, U of A)
1- Finding Equilibiria in Large Sequential games of Imperfect Information (Gilpin and Sandholm, CMU)
1- Pseudo-Optimal Strategies in No-limit poker - Rickard Andersson.pdf
1- An Artificial Intelligence Agent for Texas Hold'em Poker
1- Using Probabilistic Knowledge and Simulation to Play Poker
1- Identifying Game Players with Mouse Biometrics
1- Algorithms for abstracting and solving imperfect information games.

-----------------------------------------------------------------
Last update: 2009/11/25

classics:

1. Algorithms and Assessment in Computer Poker.
2. Approximating Game-Theoretic Optimal Strategies for Full-scale Poker
3. Robust Strategies and Counter-Strategies: Building a Champion Level Computer Poker Player

My personal voting:

1. davidson.msc.pdf (2002)
2. johanson.msc.pdf (2007)
3. schauenberg.msc.pdf (2006)

My choices are based on which papers had the best theoretical contribution to Poker AI, and are not necessarily the best placed to start learning. I have 4 papers, but the final two are tied

There's only one paper I think has a major advance (sparse updating of regret to avoid traversing the entire game tree at each iteration):

1. Regret Minimization in Games with incomplete information (Zinkevich, Johanson, Bowling and Piccione, U of A)

Second place is the idea of restricted nash response (a simple way of incorporating a restriction on opponent's strategy but which still give solvable games):

2. Constructing Robust Counter strategies (Johanson and Zinkevich, U of A)

My ties for third place are two elegant algorithms which turn out not to be very useful in practice: GameShrink, and Expectimax.

3. Finding Equilibiria in Large Sequential games of Imperfect Information (Gilpin and Sandholm, CMU)
4. Algorithms and Assement in Computer Poker (Billings, U of A)

Marv

1. For a good read, Regret Minimization, and Restricted Nash Equilibrium:
johanson.msc.pdf (2007)

2. For a good overview of important earlier work.
Algorithms and Assessment in Computer Poker - Billings

3. For a simple explanation of how to formulate the sequence form LP solution:
Pseudo-Optimal Strategies in No-limit poker - Rickard Andersson.pdf

Hi,

My votes :

1) Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment Good paper - and you have to respect someone who devotes the whole of page 2 in their thesis to "all your base are belong to us"

2) Strategy Evaluation in Extensive Games with Importance Sampling

3) Regret Minimization in Games with Incomplete Information

- PeppaPig

My choices :

1) Algorithms and Assement in Computer Poker (Billings, phd thesis)
2) Robust Strategies and Counter-Strategies: Building a Champion Level Computer Poker Player (Michael Bradley Johanson, msc thesis)
3) Regret Minimization in Games with Incomplete Information (Martin Zinkevich, Michael Johanson, Michael Bowling, Carmelo Piccione )


Yes, I'm a big fan of Alberta, of course

My personal vote is for Aaron Davidsons MSc thesis: Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment.

For my other two picks I'll have to think on it more (probably both will be UoAlberta papers, but which ones...).

1. Algorithms and Assessment in Computer Poker, Darse Billings, 2006.
2. Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment, 2002, Aaron Davidson
3. Opponent Modelling and Search in Poker, Terence Schauenberg, M.Sc. thesis, 2006.

My votes:

1. Opponent Modelling and Search in Poker, Terence Schauenberg, M.Sc. thesis, 2006.
2. Algorithms and Assessment in Computer Poker, Darse Billings, 2006.
3. Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment, 2002, Aaron Davidson

I think, opponent modelling is key in the end, while tree search could be a viable option in the future, even for NL Holdem.

Thats why my first vote goes to:

1. Using Probabilistic Knowledge and Simulation to Play Poker
2. Opponent Modeling in Poker: Learning and Acting in a Hostile and Uncertain Environment
3. Identifying Game Players with Mouse Biometrics (just for the fun of it, has nothing to do with poker)

I also started reading the papers using bayes networks, but I'm just not fit enough with statistics to get anything out of them. I'm aware of the importance of the bayes theorem and everything connected to it, still got a lot to learn.

1. Robust Strategies and Counter-Strategies: Building a Champion Level Computer Poker Player
2. Algorithms for abstracting and solving imperfect information games.
3. An Artificial Intelligence Agent for Texas Hold'em Poker