AI Figures Out the Rules of a Mysterious 2,000-Year-Old Board Game from Ancient Rome

Image by Wal­ter Crist

As far as enthu­si­asm for board games goes, no con­ti­nent has yet out­done Europe. Its advan­tage could lie in the high­ly devel­oped cul­ture of low-cost leisure evi­dent in quite a few of its soci­eties; it could also owe to the fact that board games seem to have been played there con­tin­u­ous­ly since antiq­ui­ty. We’ve long had evi­dence of exam­ples like the “Roman mill game,” bet­ter known today as nine men’s mor­ris, which Ovid appears to men­tion in his Ars Ama­to­ria of the very ear­ly first cen­tu­ry. Not that mod­ern knowl­edge of Roman table­top gam­ing is com­plete. In one puz­zling case, the stone board above was unearthed in a for­mer Roman town in the Nether­lands, but how a game was played on it remained a mys­tery — until machine learn­ing came along.

“To exam­ine whether the object may have been used as a game board, we per­formed use-wear analy­sis to iden­ti­fy evi­dence for game­play and we sim­u­lat­ed play using arti­fi­cial intel­li­gence (AI),” write the team of researchers who recent­ly pub­lished a paper on the sub­ject in the jour­nal Antiq­ui­ty. They used a sys­tem called Ludii, engi­neered to ana­lyze board-game rules.

“This soft­ware allows for AI-dri­ven play­out sim­u­la­tion, where two AI agents play a game against one anoth­er, which can gen­er­ate quan­ti­ta­tive data on game­play. In this instance, we explored whether the rules of a game would pro­duce the wear pat­tern seen on the stone.” The idea, in oth­er words, was to let the com­put­er play against itself using dif­fer­ent rules until it came upon a game that would con­tin­ue to abrade away the sur­face of the board in the same fash­ion as it already was.

This process nar­rowed it down to games “in which the goal is to block the oppo­nent from mov­ing, and those in which the goal is to place three pieces in a row.” These have a fair­ly long doc­u­ment­ed his­to­ry, from Scan­di­navi­a’s hare­tavl, to Italy’s gio­co dell’orso to Spain’s liebre persegui­da, to Greece’s kiné­gi tou lagoú. You can down­load what the research sug­gests is the most plau­si­ble rule set for this par­tic­u­lar Roman board game here, board design includ­ed. One play­er takes the side of the “hunter,” with four pieces, and the oth­er the side of the “prey,” with two. The for­mer tries to trap the lat­ter’s pieces, mov­ing only along the board­’s lines; in the next round, the roles reverse. The hunter who does the job in the fewest moves wins. Why not invite friends over to spend an evening play­ing like a Roman? For a thor­ough­ly ancient good time, first recon­struct as best you can the ambi­ence of the ther­mopoli­um at home.

Relat­ed con­tent:

Watch a Playthrough of the Old­est Board Game in the World, the Sumer­ian Roy­al Game of Ur, Cir­ca 2500 BC

Behold the First Amer­i­can Board Game, Trav­ellers’ Tour Through the Unit­ed States (1822)

Monop­oly: How the Orig­i­nal Game Was Made to Con­demn Monop­o­lies & the Abus­es of Cap­i­tal­ism

Kurt Vonnegut’s Lost Board Game Is Final­ly for Sale

The Fiendish­ly Com­pli­cat­ed Board Game That Takes 1,500 Hours to Play: Dis­cov­er The Cam­paign for North Africa

The Fear and Loathing in Las Vegas Board Game, Inspired by Hunter S. Thompson’s Rol­lick­ing Nov­el

Based in Seoul, Col­in Marshall writes and broad­casts on cities, lan­guage, and cul­ture. He’s the author of the newslet­ter Books on Cities as well as the books 한국 요약 금지 (No Sum­ma­riz­ing Korea) and Kore­an Newtro. Fol­low him on the social net­work for­mer­ly known as Twit­ter at @colinmarshall.


by | Permalink | Comments (0) |

Sup­port Open Cul­ture

We’re hop­ing to rely on our loy­al read­ers rather than errat­ic ads. To sup­port Open Cul­ture’s edu­ca­tion­al mis­sion, please con­sid­er mak­ing a dona­tion. We accept Pay­Pal, Ven­mo (@openculture), Patre­on and Cryp­to! Please find all options here. We thank you!


Leave a Reply

Quantcast