Abstraction for Solving Large Incomplete-Information Games
Authors: Tuomas Sandholm
AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | There has been a significant amount of experimental work on abstraction in games over the last dozen years, and experience suggests that in practice in large games such as heads-up Texas Hold em poker, finer-grained abstractions yield programs that play better against other programs. Furthermore, it has been shown experimentally in heads-up limit Texas Hold em poker that finer-grained abstractions yield programs that have lower exploitability that is, they play better against their worstcase opponent (Johanson et al. 2011). Experiments show that it finds a lossless abstraction when one is available and lossy abstractions when smaller abstractions are desired. More recently, we conducted exploitability experiments (nemesis computations) on the clairvoyance game (Chen and Ankenman 2006), Kuhn poker (Kuhn 1950), and Leduc Hold em (Waugh et al. 2009a); it exhibited less exploitability than prior action mappings (Ganzfried and Sandholm 2013). |
| Researcher Affiliation | Academia | Tuomas Sandholm Computer Science Department Carnegie Mellon University |
| Pseudocode | No | The paper describes algorithmic approaches but does not contain structured pseudocode or algorithm blocks (clearly labeled algorithm sections or code-like formatted procedures). |
| Open Source Code | No | The paper does not provide concrete access to source code for the methodology described in this paper, nor does it include a specific repository link or explicit code release statement. |
| Open Datasets | Yes | More recently, we conducted exploitability experiments (nemesis computations) on the clairvoyance game (Chen and Ankenman 2006), Kuhn poker (Kuhn 1950), and Leduc Hold em (Waugh et al. 2009a); it exhibited less exploitability than prior action mappings (Ganzfried and Sandholm 2013). These games are well-known or defined by the cited works. |
| Dataset Splits | No | The paper does not provide specific dataset split information (exact percentages, sample counts, citations to predefined splits, or detailed splitting methodology) needed to reproduce the data partitioning. It discusses game environments and experiments within them but not data splits. |
| Hardware Specification | No | The paper mentions support from "XSEDE computing resources provided by the Pittsburgh Supercomputing Center" and "Intel Corporation and IBM for their machine gifts" but does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments. |
| Software Dependencies | No | The paper mentions algorithms and techniques (e.g., CFR algorithm, integer programming) but does not provide specific ancillary software details, such as library or solver names with version numbers, needed to replicate the experiment. |
| Experiment Setup | No | The paper discusses aspects of the game setup (e.g., "stacks 200 big blinds deep as in the Annual Computer Poker Competition (ACPC)") but does not provide specific experimental setup details such as concrete hyperparameter values, training configurations, or system-level settings for the algorithms implemented. |