Modeling the Zombie Apocalypse

A group of undergraduate students at the University of Leicester recent put together an analysis of how quickly a zombie apocalypse would overtake the planet. Using calculations based on zombie films and television series, they found that it would only take about a hundred days to reduce the current population of the world down to a mere few hundred.

In the new analysis, the University of Leicester undergraduates assumed that each zombie would have 90 percent success at finding and infecting one human a day, a rate that would make the zombie virus twice as contagious as the Black Death, the plague that devastated Europe in the 1300s. The researchers further estimated that each zombie could live 20 days without braaaaaains.

Assuming a starting population of 7.5 billion people, approximately the world's population today, the students calculated that it would take 20 days for a single zombie to start an epidemic of noticeable proportions. At that point, the pandemic has begun. Assuming no geographic isolation, in fact, the human population would drop to 181 by day 100, with 190 million zombies roaming around.

With some geographical isolation, the situation would be a tiny bit better for humans. Assuming the zombie virus had to spread through contiguous regions and that zombies were somewhat limited in their ability to travel (not leaving their current region until there were 100,000 zombies roaming there), human survivors would number 273 by day 100.

A more realistic model might assume that each zombie could find fewer human victims over time, the students wrote, because there would simply be fewer humans to find. "We have also not included the possibility for the humans to kill the zombies," they wrote.

There are actually a couple of reasons this model is not that realistic besides those mentioned, and of course besides the fact that the kind of zombies being discussed here were only invented by George Romero in the late 1960's. If you cut the likelihood of infection in half to match the black death, my guess is that humans still get mostly wiped out, just in a longer time frame. As we know, that didn't happen. The black death killed about a third of Europe's population, which is a lot of people, but nowhere near enough to practically wipe out humanity.


Popular epidemiology models usually miss the fact that some percentage of the human population is resistant to every disease that exists. People talk about the virus from the 1919 Spanish flu pandemic as though it would be the most potent bio-weapon ever if it could be synthesized and released, but even the deadliest flu in known history still only had a mortality rate of 10-20%. Even Ebola, one of the most awful diseases known, has a mortality rate of about 70-75%.

My point is that if you assume a magical 100% mortality/infection rate, you can get numbers like these - but they don't correspond to characteristics of real-world diseases. There also is a difficult-to-quantify factor for members of the population that are entirely immune, which also exists for the vast majority of known diseases and can also reduce the spread considerably. The mortality rates up there are for infected individuals, not the entire population.

I do realize that breaking down this analysis is kind of silly, given that we're talking about zombies - entirely fictional entities that don't even exist in this form in folklore, and which were created for a film. But as with a number of my other posts, it's all about being realistic when estimating risks. If we ever want to get people to do it correctly, we need to start somewhere - and maybe looking at something as outlandish as a zombie epidemic is a good place to begin.