The AI Breakthrough That Unraveled an Ancient Roman Puzzle

AI DiagnosticsAI TherapyAI & Machine Learning

Imagine a game that has been shrouded in mystery for centuries, only to be solved by a team of researchers and AI.

TL;DR
• Ancient Roman board game solved by AI after 40 years
• Researchers used 3D scans and AI to unravel the rules
• The game is a type of strategy game known as a blocking game

Summary
A limestone slab from the ancient Coriovallum settlement in the Netherlands, excavated in 1984, has long puzzled scholars with its intricate grooves and wear patterns that suggest a board game. The stone, measuring roughly 8.3 by 5.7 inches, shows evidence of sliding pieces, hinting at repeated play rather than ceremonial use.

An international consortium—Maastricht University, Leiden University, Flinders University, Université Catholique de Louvain, and the Roman Museum—employed 3D scanning and AI to decode the game’s rules. Using the Ludii platform, researchers tested over 100 rule sets from historic games, allowing two bots to play thousands of turns and compare resulting friction patterns with the stone’s wear.

The simulations converged on a blocking game model, where players trap opponents’ pieces instead of capturing them. This aligns with known medieval blocking games like haretavl and gioco dell’orso, yet represents one of the earliest examples of such mechanics in antiquity. The study demonstrates how machine learning can uncover hidden cultural practices, offering a new lens for interpreting artifacts that lack textual records.

Published in Antiquity, the research highlights AI’s potential to illuminate forgotten traditions and encourages broader application of digital tools in archaeology. By revealing the game’s structure, the project not only solves a long‑standing mystery but also showcases how modern technology can bridge gaps in historical knowledge, fostering collaborative discovery across disciplines.

This breakthrough highlights the potential of AI in uncovering hidden stories in ancient artifacts, and we can't wait to see what other secrets it will uncover.