How Archaeologists Identify Ancient Games

A grid scratched into a paving stone is one of the quietest objects an excavation can produce. No inscription, no glory, no king’s name, just a few lines and a scatter of pebbles. Yet to an archaeologist, these are among the most human traces a site can offer: the marks of people passing the time. Recognizing a game in the ground, and then reconstructing how it was played, is a craft in itself.

Reading the board

The first task is simply to tell a gaming surface from idle scratching or decoration. Grids alone are suggestive but not proof; a tic-tac-toe of lines could be a mason’s measurement. What persuades is context and wear. A Roman ludus latrunculorum board, for instance, takes the form of a regular grid, often something between 7x8 and 12x8 squares, and turns up carved into floor tiles, scratched into the paving of forts, and incised on the steps where soldiers waited. Find such a grid beside gaming tokens, in a place where people gathered with time to kill, and the case grows strong.

Wear tells the rest. Examined under magnification, the incised lines of a heavily used board show grooves worn unevenly, with some routes rubbed deeper than others by pieces dragged across them again and again. This is the signature of real play, and it distinguishes a board that was used from one that was merely drawn.

The tokens and the dice

Boards rarely survive alone. The smaller evidence, including counters of bone, glass, or stone, knucklebones, and dice, is often what confirms a find. Dice are especially telling. The four tetrahedral dice of the Royal Game of Ur, read by counting marked corners, are unlike any modern die and instantly identify the family of game they served. Marked throwing-sticks point toward Senet. When the board itself has rotted away, a handful of pieces in a grave can still name the game that was buried there.

Recovering the rules

The hardest problem is not finding a game but learning how it was played, since ancient peoples saw no need to write down what everyone already knew. Archaeologists work like detectives. Sometimes a lucky text surfaces: the cuneiform tablet that let Irving Finkel reconstruct the Royal Game of Ur is the celebrated example. Where no rules survive, researchers treat the worn stone almost as a fossil of play, testing candidate rule-sets in computer simulation and checking which produces wear patterns matching the real board. Recently a Roman board from the Netherlands had its likely rules proposed with the help of artificial intelligence doing exactly this.

The reconstruction is always provisional, always honest about its uncertainty. That honesty is the point. A faithful replica should reflect what the ancients built and how confidently we know it, and where the scholarly debate still lives.

Every board in our collection is grounded in this evidence: the surviving artifacts, the texts, and the careful inference. See how the record becomes an object you can hold at our games.