A small limestone slab housed in a Dutch museum may represent a previously unknown ancient Roman board game, according to new research using artificial intelligence to analyze wear patterns on the artifact. The 8-by-6-inch oval-shaped stone, found in the city of Heerlen—once the Roman settlement of Coriovallum—had been largely overlooked as a fragment of architectural debris until recent examination revealed signs of extensive play.

Led by archaeologist Walter Crist of Leiden University, the study scanned the slab using high-resolution 3-D imaging to map microscopic grooves and abrasions formed by repeated use. By feeding this surface data into Ludii, an AI platform trained on roughly 100 Roman-era games, researchers simulated thousands of rule variations to identify those that best matched the physical wear. The approach narrowed the possibilities to nine plausible rule sets for a two-player blocking game, one in which opponents maneuver pieces to trap each other in the board’s corners.

The newly proposed game, dubbed ludus coriovalli or the Coriovallum game, features asymmetrical gameplay involving sliding small pieces—potentially made of glass, bone, or earthenware—along the etched lines. This strategic blocking format challenges previous assumptions that such complex game mechanics emerged only in the medieval period. Crist described the stone as evidence of a lost recreational tradition on the Roman frontier, distinct from the more elaborate gaming objects found in urban centers.

However, not all experts are convinced by the interpretation. Ulrich Schadler, an archaeologist at the University of Fribourg and director of the Swiss Museum of Games, expressed skepticism about the artifact’s authenticity as a genuine Roman game board. Schadler noted that the slab’s clean, straight lines differ from typical Roman game boards, which are generally more crudely inscribed, and suggested the pattern might be a random or modern projection rather than historical evidence of gameplay.

Veronique Dasen, also from the University of Fribourg but not involved in the study, acknowledged the significance of the research in addressing the broader challenge of inferring rules from scant archaeological remains. While the conclusions remain unsettled, the study highlights the potential of combining digital technology with traditional archaeological methods to revisit overlooked artifacts.

Discovered in the late 19th or early 20th century but only recently reexamined through this new lens, the limestone slab may represent a rare glimpse into everyday leisure during the Roman Empire. As Crist observed, the absence of contemporary written records about such games does not preclude their existence; rather, it underscores the importance of innovative approaches to uncovering lost aspects of ancient culture.