Empirical vs. Theoretical Probability in Human Binary Choices from Go Games

21,212 professional Go matches among 311 masters and 586 games from 293 amateur players were analyzed to investigate discrepancies between empirical and theoretical probability in binary choices. The data shows that higher-ranked players had a significantly higher chance of selecting black stones, challenging the assumption of equal odds. The dataset was authored by Yong-Hwan Kim and last updated on 2026-04-30.

Use Cases

• Modeling strategic interaction in binary choice games based on the described stone selection process.
• Analyzing the relationship between player skill level and choice probability based on the reported rank-based discrepancies.
• Studying deviations from theoretical probability in human decision-making based on the large-scale analysis of Go matches.

Strengths

• Includes data from 21,212 professional matches and 586 amateur games, providing a substantial sample for analysis.
• Explicitly compares outcomes for players of different skill levels, as described in the analysis.
• Released under a permissive CC-BY-4.0 license, facilitating reuse.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is very small (16.7 KB), indicating limited scope or highly aggregated data.

Provenance

Source

figshare, author Yong-Hwan Kim.

Collection Method

Analysis of professional Go match records and in situ games with amateur players.

Freshness

Last updated 2026-04-30 05:24:40; freshness should be verified.