Creating a Superhuman AI for Contract Bridge

This project aims to develop a foundational AI model capable of analyzing and bidding contract bridge deals at a superhuman level. The AI is rooted in advanced analytics and strives to enhance strategic decision-making in the complex world of bridge. The target of the foundational model is BBO's GIB bidding system

Target Audience

This project is designed for software developers with an interest in artificial intelligence, machine learning, and bridge bidding systems. It has no end-user components.

Prerequisites

General Prerequisites

• Python 3.12 or later
• Polars: A fast DataFrame library for data processing.

Bridge-Specific Prerequisites

• Endplay: A Python library for analyzing bridge hands.

Description of Files

• Pipelines to Superhuman Bridge Analysis and Bidding.md: Describes the methodology for analyzing and bidding bridge hands.
• bbo_bidding_vocabulary.txt: Contains a glossary of bridge-related terms (e.g., "Rebiddable," "HCP," "Solid").
• bbo_bidding_expressions.txt: Lists expressions used to evaluate bidding criteria (e.g., "Rebiddable_C," "HCP >= 15").
• requirements.txt: Specifies project dependencies. Use the following command to install them:

  pip install -U -r requirements.txt
  

References

• GIB System Notes: Documentation for BBO's GIB system.
• GIB Descriptions: Additional details on GIB bidding logic.
• Ben GitHub Repository: A related open-source project for bridge analysis.
• IJCAI Paper on AI and Bridge: Academic research on AI strategies for bridge.