Adapted by:
Javier Pastor
Senior WriterComputer scientist turned tech journalist. I've written about almost everything related to technology, but I specialize in hardware, operating systems and cryptocurrencies. I like writing about tech so much that I do it both for Xataka and Incognitosis, my personal blog.
Alba Mora
WriterAn established tech journalist, I entered the world of consumer tech by chance in 2018. In my writing and translating career, I've also covered a diverse range of topics, including entertainment, travel, science, and the economy.
A few years ago, archaeologists made a unique discovery at the Shahr-i Sokhta site in Southeastern Iran: an ancient board game. This game is around 4,500 years old and experts believe it to be the oldest complete board game ever found. However, the challenge was that no one knew how to play it... until AI entered the scene.
Persian games. The Shahr-i Sokhta site was part of the Helmand civilization during the Bronze Age, located east of Iran and south of Afghanistan. Among other findings, archaeologists uncovered a game board featuring 20 circular spaces, four dice, and 27 geometric pieces. But how did players play the game?
No instructions included. Traditionally, ancient board games rarely included written rules. Instead, instructions were passed down orally. This likely resulted in modifications to the rules among different player groups over decades or even centuries. However, this lack of definitive rules also makes it easier for experts to deduce how the game might have been played. According to New Scientist, the goal isn’t to find exact rules but rather to identify those that make the most sense.
AI’s key role. The AI systems used to understand how to play ancient games utilize different techniques. One example is Monte Carlo tree search, which was also employed in the development of DeepMind’s AlphaGo.
These systems can simulate possible rules based on the game’s board structure and available pieces. They evaluate several potential rule variations to determine which ones aren’t only logical but also lead to enjoyable gameplay.
Beyond chess. Among these ancient games are real treasures, like Senet, which was found in the tomb of Tutankhamun. The image below depicts the Egyptian queen Nefertari playing Senet, highlighting its popularity in that culture. Another example is the Royal Game of Ur from ancient Mesopotamia. Its secrets were unraveled after researchers deciphered a cuneiform tablet in 2007.
First studies. One of the earliest applications of AI in analyzing ancient board games is the game Ludus Latrunculorum, known as “the game of brigands.” Romans used to play this game, although experts believe it had existed in Greece in an earlier form. It’s one of the best-known ancient games. Writings from that era allowed researchers to reconstruct its rules with a high degree of reliability.
The Dutch study. Cameron Browne from the University of Maastricht in the Netherlands led the Digital Ludeme Project. The research work investigated around 1,000 traditional games spanning 6,000 years of human history, including Ludus Latrunculorum.
Simulations. Studying the Ludus Latrunculorum game wasn’t straightforward. Variants of it appeared on boards of differing sizes and formats throughout history. However, AI played a crucial role in discovering possible rules. Simulations were conducted to analyze which board configurations made the most sense given the most plausible rules.
The study revealed a strong connection among three still-active games: Kharebga, Seega, and Tablut. In the end, researchers concluded that smaller boards were likely used in the Roman version of Ludus Latrunculorum, while the larger boards may have been intended for other unidentified games.
How to play Shahr-i Sokhta. Similarly, AI assisted in reconstructing the probable rules of the ancient game Shahr-i Sokhta. Developer Sam Jelveh and archaeologist Hossein Morad created an online playable version of the game, complete with comprehensive information on how it was intended to be played.
Endless research. More than 200 computer scientists, archaeologists, and historians are collaborating on the GameTable project. The goal is to develop even more advanced AI tools aimed at uncovering the rules of ancient board games. This research also seeks to explain why some ancient games have transformed into modern counterparts–such as the Royal Game of Ur eventually evolving into backgammon–while others have faded into obscurity.
Images | Amirkhan33 (Wikimedia Commons, CC BY-SAC 4.0) | Nina de Garis Davies (Wikimedia Commons/MET) | Persian Wonders
Related | The 33 Best Board Games: From ‘Catan’ to ‘Codenames’
Log in to leave a comment