Skip to Content

Simulating Medieval Warfare on Supercomputers

The Medieval Warfare on the Grid project is akin to applying the simulation game Sid Meier’s Civilization to real questions of history.

Historians don’t know how many soldiers were at the battle that marked the beginning of the end of the Byzantine empire, or how long it took those soldiers to get there, or how their leaders even acquired the resources to keep them fed. But all of these questions are now being answered, or at least addressed, by a project called Medieval Warfare on the Grid.

Agent-based simulations like Medieval Warfare on the Grid consist of agents and the environment in which they interact

Historical accounts put the size of the army led by Emperor Romanos IV Diogenes at the battle of Manzikert at up to 100,000 troops, but could an army that large really have been raised and supported in AD 1071? To find out, a group of computer scientists, archaeologists and historians teamed up to create what’s known as an ‘agent-based simulation’ of the Byzantine Army as it marched from Constantinople into what is now modern Turkey. (Here’s a paper on the research and a short video.)

Agent-based simulations model complex things like societies by simulating the simple ways that individuals in that population interact. At the head of those agents is an Emperor choosing, for example, the route the army will take, but the actual progress of its march consists of relatively straightforward things like setting up and breaking down camp, acquiring food, disseminating orders and the like. All of which this simulation includes.

This kind of simulation works because, as complicated as the behavior of individual humans might seem, in aggregate, crowds of humans aren’t so different from termites or birds or any other animal. Group behaviors like flocking and nest building can be achieved with just a few basic rules describing the inclinations and interactions of the individuals in a group.

While models of this kind can never definitively answer questions about the historical accuracy of various claims, they can eliminate some possibilities, while also helping to address matters on which historians are mostly silent. These include things like how the economies of medieval societies were structured (think of taxes, transportation and food production) in order to support large standing armies.

The agent-based simulator used in this case doesn’t simulate battles, because the decision-making that takes place in combat is too complicated to model at present. Just simulating how thousands of individual agents march from one location to another over a period of days requires significant parallel processing.

Sometimes setting up camp is the last thing a Byzantine soldier ever does.

That doesn’t mean the agents in this simulation are immortal, however: the soldiers at right aren’t exactly napping. Run out of food and, well: An army, as they say, runs on its stomach.

Follow Mims on Twitter or contact him via email.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.