So the standard type of mathematical model that's used to describe early stage epidemics (such as this one) basically assume exponential growth. It's hard to g

So the standard type of mathematical model that's used to describe early stage epidemics (such as this one) basically assume exponential growth. It's hard to get the math to stop doing that when a disease reaches a non-exponential stage. That's why you got some bonkers predictions for Ebola - if you let an exponential model run for a year, it's always going to be bad.