Monopoly game engine with mathematically rigorous AI — Markov chains, property valuation, and strategic trading.
Repo: github.com/chrisaacson69/monopoly
The Vault’s gaming research pages contain the abstract theory. The Monopoly repo contains the applied implementation. Findings from the project validate or challenge the theory, which feeds back into the Vault.
| Vault (theory — research/gaming/monopoly/) | Monopoly Repo (applied) | Status |
|---|---|---|
| Nash Bargaining Problem | research/positional-value-analysis.md — 3-player Orange coalition, Shapley value |
Vault has abstract theory; repo has Monopoly-specific worked example |
| Multiplayer Coalition Problem | research/positional-value-analysis.md — coalition formation dynamics |
Vault has abstract theory; repo applies to Orange split |
| Bilateral Trade Valuation | research/valuation-notes.md — denial value, time-to-impact, knockout probability |
Vault has trajectory framework; repo adds concepts (now captured in Frontier Trade Theory) |
| Subgraph Investment Optimization | ai/property-valuator.js — EPT and ROI calculations |
Vault has abstract framework; repo has implementation |
| Frontier Trade Theory | research/simulation/GROWTH-AI-SUMMARY.md — tournament results |
Theory in vault; empirical validation in repo (results summarized in frontier page) |
| Subgraph Trade Engine Spec | research/simulation/subgraph-*.js — implementation of the spec |
Spec in vault; code in repo |
| Praxis: Agent Teams | research/SESSION_SUMMARY.md — Markov engine, technical architecture |
Vault has economic framework; repo has Monopoly application |
The rule: All theory lives in the Vault (research/gaming/monopoly/). Implementation lives in the repo. Cross-link, don’t duplicate. Repo research files (positional-value-analysis.md, valuation-notes.md) are earlier working notes now superseded by the vault’s theory pages.
Full Monopoly implementation in JavaScript with a deep AI stack. The engine models the complete game (properties, auctions, trades, jail, chance/community chest) and the AI uses Markov-chain positional analysis to compute Earnings Per Turn (EPT) for every property, driving trade evaluation and strategic decisions.
simulation/ for comparative testing