Learning from the (un)dead
on Halloween

As Halloween arrives, it has been revealed that researchers at Aalto University are investigating how a zombie plague would spread through Finland. 

It’s a light-hearted project, but it offers serious insights into global challenges, such as containing a pandemic – or even coping with disinformation. 

The research expands on traditional epidemiological models by adding a simulation of people and zombies moving around in Finland, within and between cities. The use of individual-based simulation (instead of a population-scale model) with a spatial component enables the research team to model things like quarantining an infected region or the difference between a zombie plague starting in the densely populated capital region or much smaller cities elsewhere.  

Estimating some of the model’s parameters was a particular challenge. ‘What’s the right probability for a human winning an encounter with a zombie? The problem is that we’re walking blind here, because real data on such questions is severely limited,’ says Professor Lauri Viitasaari of Uppsala University, who is in charge of the mathematical model. 

‘The large number of human-zombie interactions that have to be simulated makes this model computationally intensive,’ says Natalia Vesselinova, the postdoctoral researcher who heads the computational side of the project. To cope with this, the team came up with a simpler but still realistic model of the interactions. ‘That lets us run the simulation with less computational power or simulate larger scenarios with similar computational resources.’ 

One of their main findings so far has been that the plague would spread extremely quickly, so interventions to block it would have to happen very fast. Starting with a single zombie in Helsinki in their basic model, there would be just seven hours to completely quarantine the capital (or kill the zombies). Otherwise, zombies would inevitably overrun the country. 

‘I shouldn’t have found it surprising, but I was surprised at how quickly we have to react to keep our population alive. It made me think about moral issues like the rights of individuals versus the rights of a population,’ says Pauliina Ilmonen, the mathematics professor leading the study.

These difficult moral questions are familiar from COVID-19. The zombie plague simulation offers a way to explore the effects of different interventions and consider them in the context of diseases with different features, such as how quickly they spread or how severe they are. Because it simulates individuals’ actions, it can also be used to test how disinformation would affect the spread of an epidemic (eg, by having some ‘zombie deniers’ ignore warnings). 

The model can be adapted to simulate other countries or regions, and it can also be used to investigate other things that spread like a disease. For example, it could be used to model the spread of rumors or disinformation and test the effect of strategies to control them.