Dr. Walter Crist

The image shows a rectangular Roman dice tower (turricula) made of four copper alloy (bronze) plates with punched cut-out Latin letters and cut-out decorative patterns. At the bottom front is a stepped exit chute with small bronze bells attached to the opening. The tower is a Roman anti-cheating device. It has an open top and is hollow inside except for three staggered, downward-sloping plates, designed to randomize dice as they fall, ensuring unpredictable dice rolls. When the dice rolled out of the exit chute they rang the bells! There is a decorative dolphin either side of the stepped exit chute. The top of the front plate has two decorative pine cone finials. Height 25 cm. There is a single die shown next to the stepped base to illustrate how it was used. The front inscription reads: PICTOS VICTOS HOSTIS DELETA LVDITE SECVRI Translated as: ‘The Picts defeated, the enemy has been destroyed, play in safety’. Around the top of the three remaining sides, a second inscription made with cut out letters reads: ‘UTERI/FELIX/VIVAS’ translated as ‘Use happily; may you live well’. Found at a Roman villa at Froitzheim in Germany in 1985.

Roman anti-cheating gaming accessory!

This Roman ‘turricula’ (dice tower) was used to ensure a fair roll of the dice! 🎲🎲🎲

Dice dropped into the top, tumbled over sloping internal levels, and appeared randomly below.

From Froitzheim, Germany, AD 300-400

📷 LVR-Landesmuseum Bonn

#Archaeology

The British Museum photo shows a six-sided, cube-shaped die carved from rock crystal on a grey surface. The six sides are numbered one to six. The numbers are represented by circular incised markings, which comprise of a dark outer ring with a dark dot at the centre. Length: 9 - 13 millimetres, width: 9 - 13 millimetres. Display lighting casts a shadow beneath the die.

Timeless design!

Roman rock crystal gaming die marked one to six like modern dice 🎲 1st-2nd century AD.

📷 British Museum www.britishmuseum.org/collection/o...

#Archaeology

Thanks for the shoutout!!

How researchers learned the rules of an ancient Roman board game A team of researchers in the Netherlands set out to decipher the rules of an ancient Roman board game, with an assist from artificial intelligence.

Some analog #archaeogaming on NPR yesterday: finding the rules of an ancient Roman board game with @waltros.bsky.social and others: www.npr.org/2026/02/23/n...
#classics

Above: piece of stone engraved with lines with the appearance of a game board. Below: illustration of the lines with circular black and white game pieces on them, showing how the pieces could move.

Researchers have used artificial intelligence to model possible rulesets for a Roman board game, concluding it was a kind of game previously unknown in Europe until the Middle Ages, pushing back evidence of their play by several centuries.

A playful #AntiquityThread 1/13 🧵

🏺 #Archaeology

THANK YOU 💯💯💯

The second union of archaeologists in Greece taking a stance on the EAA saga. They represent the contract archaeologists, the largest sector; they call for a boycott: the majority of the local archaeological community in Greece does not support the Athens EAA meeting, under the current leadership.

Photo by Het Romeins Museum

photo of a Roman stone game looking like a wheel of cheese in a still life

Ok everyone jk, it’s not a game it’s actually cheese

AI has helped scientists from Heerlen, Maastricht and Leiden confirm a mysterious Roman stone is a real board game—and even reconstruct its rules nearly 2,000 years later. The discovery pushes strategy games further back in time. Curious to play it yourself? 🎲 Read more: edu.nl/u6cgy

#AI

The limestone object, about 21 centimetres wide and shown here with stone counters, caught archaeoludologist Walter Crist’s eye in the Het Romeins Museum in Heerlen. To crack its rules, Crist and his colleagues applied an AI-powered game system that contains the mechanics of thousands of games, past and present, from around the globe.

Forbidden chocolate chip cookie (Roman board game made out of a rock)

Ludii Portal Home of the Ludii General Game System.

Fancy trying out the game yourself? You can play with the ruleset online here: ludii.games/details.php?...

NEW AI simulation helps calculate the rules of a previously unknown Roman board game, pushing evidence for the playing of blocking games in Europe back centuries!

🆓 doi.org/10.15184/aqy...

🏺 #Archaeology

@waltros.bsky.social @eric-piette.bsky.social @Dennissoemers.bsky.social

Cover of the February 2026 issue of Antiquity, featuring a person in a high-vis jacket excavating a wooden trackway in peatland.

The first issue of the year is out now! Featuring great #archaeology such as:

🎲 Calculating the rules of an unknown Roman board game
🕳️ The purpose of Peru's iconic 'Band of Holes'
🍷 Glass trade from the Roman-Islamic period in Jordan

& more!
www.cambridge.org/core/journal...

🏺

Other explanations don’t fit what we know (too sloppy for masonry guidelines, nonsensical for an architectural plan, decoration doesn’t explain wear, incomplete opus sectile is inconsistent with manufacturing process), a game board is the interpretation that best fits what we know about the stone.

We then analyzed the results of the games the AI agents played to identify those which showed disproportionate piece movement in the same place as the wear on the stone. Nine rulesets, all a type of blocking game like haretavl, produced such results.

We modeled candidate rulesets from these documented games on the different possible board configurations. @ericpiette.bsky.social and @dennissoemers.bsky.social implemented the rules in the Ludii software and designed AI simulated play to track piece movement over 1000 trials of each ruleset.

One such game is haretavl, doucmented in Denmark in the 18-19 century and with archaeological evidence going back to the medieval period, from Ireland to Latvia. One player with three pieces attempts to block their opponent who has one piece.

But, what game could produce this disproportionate wear? We looked to the Ludii games database, with >1000 ancient and traditional games and identified games on boards with a similar number of playing spaces (the intersections of the lines).

ludii.games/library.php

We examined the surface further with a microscope. Near the line with visible wear, the microtopography of the stone is flattened. Away from the lines, the surface is lumpy, and identical to the underside of the stone.

We then 3d scanned the stone to look for invisible signs of wear. We noticed that such wear, shown in pink, was concentrated along the lines where pieces would be moved.

Up close, I could notice damage to the top surface consistent with abrasion damage caused by sliding Roman glass or stone game pieces along the stone. You can see the smooth area along the bottom right diagonal.

Back in 2020, I visited the Thermenmuseum (now Het Romeins Museum) in Heerlen, 🇳🇱, and saw this object on display as a game. I did not recognize the pattern as a game documented anywhere else, so I was skeptical. I asked the curators if I could have a closer look.

Ludus Coriovalli: using artificial intelligence-driven simulations to identify rules for an ancient board game | Antiquity | Cambridge Core Ludus Coriovalli: using artificial intelligence-driven simulations to identify rules for an ancient board game - Volume 100 Issue 409

Really proud of this work, where we combined traditional archaeological methods and AI simulation to say that yes, this object was more likely than not a game board.

How did we do it? I’ll explain 👇👇

doi.org/10.15184/aqy...

Is this carved rock an ancient Roman board game? The lines worn into an engraved limestone object from the Netherlands are consistent with the idea that it was a Roman game board, according to an AI analysis

The lines worn into an engraved limestone object from the Netherlands are consistent with the idea that it was a Roman game board, according to an AI analysis

It was great chatting with you too! Thanks for writing about our research!

Is this carved rock an ancient Roman board game? The lines worn into an engraved limestone object from the Netherlands are consistent with the idea that it was a Roman game board, according to an AI analysis

This carved rock may have been a board game, according to a study in which AIs played out thousands of games with slightly different rules against each other to reproduce which moves might have led to the wear and tear seen on the board. 🧪 🤖 🎲

www.newscientist.com/article/2514...

Researchers crack the rules of unknown board game from the Roman period Researchers have used AI to reconstruct the rules of a board game carved into a stone found in the Dutch city of Heerlen. The team concludes that this type of game was played several centuries earlier...

www.universiteitleiden.nl/en/news/2026...
Researchers discovered that the wear on the stone is most likely linked to the playing of so-called blocking games: board games in which the goal is to prevent the opponent from moving.

The research article at Antiquity will be up soon ☺️

My work with @eric-piette.bsky.social @dennissoemers.bsky.social et al. using AI-simulated play to identify rules that produce piece movement that matches use-wear patterns on an object from Roman Coriovallum is getting picked up in the popular press 😁