On Optimal Strategies for Wordle and General Guessing Games
Authors: Michael Cunanan, Michael Thielscher
IJCAI 2023 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We develop a general method for finding optimal strategies for guessing games while avoiding an exhaustive search. This work is developed to apply to any guessing game, but we use Wordle as an example to present concrete results. We specifically demonstrate using these theorems to show the Wordle strategy found by our framework is optimal. As shown by Table 1, using combined valuations does offer an improvement over using any single valuation alone. Table 2: Using APMINTOTAL to find a good strategy for Wordle. Table 3: Using APMINTOTAL to find a good strategy for other popular guessing games. Table 4: Using V1, . . . , V5 to filter potential starting Wordle guesses, starting with 12972 possible guesses. |
| Researcher Affiliation | Academia | Michael Cunanan and Michael Thielscher School of Computer Science and Engineering, University of New South Wales |
| Pseudocode | No | The paper provides definitions and theorems but does not include any structured pseudocode or algorithm blocks. |
| Open Source Code | Yes | Full source code for this experiment and all subsequent ones is available at https://github.com/cunananm2000/WordleBot. |
| Open Datasets | No | GW and SW are publicly known sets and can be found in Wordle s source code. Our research into Wordle strategies was initially conducted on Wordle s original sets of guesses and secrets (before 15th February 2022), and so we describe our work using these sets, with |GW | = 12972 and |SW | = 2315. The paper describes the data sources (Wordle's source code) and states they are 'publicly known', but it does not provide a direct URL, DOI, or a formal citation with author/year for accessing these datasets. |
| Dataset Splits | No | The paper focuses on finding optimal strategies for guessing games and evaluates them using a 'TOTAL' metric. It does not involve machine learning model training or provide specific details regarding training, validation, and test dataset splits. |
| Hardware Specification | No | The paper does not specify any hardware details (e.g., GPU/CPU models, memory specifications, or cloud computing resources) used for running the experiments. |
| Software Dependencies | No | The paper mentions that full source code is available but does not list specific software dependencies with version numbers (e.g., Python version, specific libraries, or frameworks). |
| Experiment Setup | Yes | To limit this exponential growth, we propose that instead of searching over all g G, only search over the best n guesses in G; we call n the search breadth. Table 2: Using APMINTOTAL to find a good strategy for Wordle. n APMINTOTAL(SW , n) Starter 1 7944 TRACE 5 7921 SALET 10 7920 SALET 20 7920 SALET |