The Zombie Apocalypse vs COVID-19
An exploration of epidemiological modeling
We’re going to compare the progression of the potential Zombie Apocalypse to our current COVID-19 epidemic.
We'll use a model known as SIR (which stands for Susceptible Infected Recovered) to this. I’ve put the fun (or sobering) graphs up front showing how these outbreaks might look. In the process, I’ll explain what the SIR model is and how it works. You’ll get an understanding of the data on this site.
So, to start, just what are the difference between the Zombie Apocalypse and COVID-19? Let’s list them out.
The differences
Zombie Apocalypse.
|
COVID-19.
|
The models
Here are the simulated progressions of outbreaks on the Isle of Oahu with a population of 953,207. The solid lines represent the Zombie apocalypse. The dashed lines represent the Covid-19 pandemic. These are our SIR models, which we’ll explain more below.
The first graph is about numbers of people in those categories. Blue are the people without the disease (who are susceptible), red are those who are infected, and yellow is those who have gotten better or died.
The second graph shows the cumulative probability of getting infected (orange) and the probability of getting infected on any given day (blue)
What to note
In the first graph, the Zombie Apocalypse (solid lines) becomes a catastrophe about 20 days sooner than the coronavirus (dashed lines). Everyone eventually becomes a Zombie. Roughly 24,000 people on Oahu (~2.5%) never become infected with the coronavirus.
In the second graph, with the coronavirus, the probability of infection hits its peak at 10% but then drops again. In the Zombie Apocalypse the infection rate increases every day until it reaches 26% and never drops back down after.
SIR Model Explanation
As a reminder, in our SIR model, the S stands for susceptible, the I stands for infected, and the R stands for recovered. Recovered is a bit of a misnomer because it means either that you got better or you died. The important thing for the model is that the person is no longer infectious. People will be in one of these buckets at any given time.
Susceptible
Look back up at the first graph, and note the blue lines (solid and dashed). This is our susceptible people, who are not (yet) infected with Zombie or Covid-19.
At the start, everyone is in the susceptible category (except “patient 0”). Over the progression of the epidemic, the susceptible number drops towards zero as people get infected.
Infected
In order for someone to move from the susceptible bucket to the infected bucket, a susceptible person has to bump into an infected person. The number of newly infected on a given day depends on both the number of susceptible people and the number of infected people. So at the start of an outbreak because there are few infected, it is hard to become infected because there are a lot more susceptible healthy people to bump into than infected.
The other factor is the disease itself and how easily people become infected. This is represented by the doubling rate (aka the growth rate, or R0). The more infectious the disease the smaller the doubling rate. I assumed the same doubling rate of 3 days for both of these epidemics. i.e. the number of infected will double every 3 days.
This doubling is what makes it hard for us to comprehend the potential danger. Because once they hit a certain point, numbers will begin to rocket up.
For our Zombie apocalypse, the number of infected just keeps growing until it hits 100%. Zombies don’t ever move in the recovered category (this is not the 28 Days Later version of the zombie apocalypse) and can’t swim off the island. We also don’t take into account the notion that someone might be immune.
For the COVID-19 progression, people recover in roughly 2 weeks. The model assumes that people can’t be infected a second time.
Recovered
Zombies don’t recover in our model, but in the case of COVID-19 people do recover and are no longer in the infectious pool of people and can’t transmit the disease.
Chances of infection (the second graph)
The second graph shows the chance of infection. The blue lines (solid and dashed( show the chance of becoming a Zombie or getting infected with COVID-19 on any one day. The red lines show the cumulative chance of your becoming a Zombie or having had COVID-19.
By the end of the outbreaks you almost certainly will be a zombie but you have got only a 97.5% of having had COVID-19. Said another way in this model 2.5% of the population will never get COVID-19.
I found this last point mysterious at first (surely everyone would get COVID-19) but no that is not the case. At the tail end of the outbreak of a disease in this model the disease extinguishes itself. There are no longer enough susceptible people and infected people for the infection to sustain itself. So this leaves a small subset of the population who have never had the disease. In the case of Oahu that works out to about 24,000 people.
This is of course just a simple model. In real life, there might be a reservoir that continues to carry and contain the disease or the disease mutates enough that recovered people can get reinfected. Animal populations can serve as reservoirs which is why during the bird flu epidemics China slaughtered lots of chickens. The flu and the common cold stick around because they change from season too season.
Take Aways
Other than the obvious, to avoid zombie apocalypse’s, there is the sobering notion that most people will get Covid-19 at some point. The only thing we personally and collectively act on is to change how fast that happens. The faster the doubling rate, the more likely we will overwhelm our medical infrastructure.
If we can slow the doubling rate, we will save lives.
The numbers you see every day are not the doubling rate, and can be deceptive