AI helps researchers identify 2,000-year-old roman board game stone
3D scans and artificial intelligence reveal how the ancient game may have been played
Researchers in Europe have discovered what seems to be an ancient Roman board game stone which dates back 2,000 years after they used artificial intelligence and 3D scanning technology to study its markings.
The discovery was found in Heerlen, located in the Netherlands. The findings appeared this week in the academic journal Antiquity. Experts from Maastricht University studied the etched lines and wear patterns together with the stone structure to establish its historical value and function.
The stone features deep lines carved into its surface. Restoration Studio Restaura Luk van Goor said detailed 3D scans showed some lines were deeper than others and that the edges were neatly finished, suggesting a completed object rather than unfinished work.
Leiden University Archaeologist Walter Crist explained that visible wear appears along the carved lines. He noted that the marks are consistent with pieces being repeatedly slid across the surface, strengthening the argument that the stone was used for gameplay.
After mapping the wear patterns, researchers used artificial intelligence to simulate different Roman-era board game rules. The AI models tested how pieces might move under various rule sets.
One configuration, likely involving blocking an opponent’s pieces, closely matched the wear marks found on the stone.
Karen Jeneson, curator of the Roman Museum in Heerlen, said the rules identified by researchers align with games from similar cultural periods.
-
Grandparents are turning kids' photos into AI art
-
Google faces €403 million GDPR fine over location data
-
WhatsApp adds starred music feature to make status songs easier to find
-
Facebook and Instagram back online after sudden US outage shocks thousands
-
Does China agrees with US AI safety fears?
-
WhatsApp adds 8 emoji categories on Apple Watch
-
Meta fights Ofcom over new WhatsApp and Instagram online safety requirements
-
ZCode tried uploading your files 564 times, researcher finds
