Forecasting Wuhan Coronavirus Cases (Performance Update)
Checking in on reality
A week ago, I put out a primary long-term forecast for the Wuhan Coronavirus (nCoV-2019). This was actually a day after I presented a short term, very simple exponential forecast, which was decent but not at all capable of forecasting in the long term due to the nature of the exponential function.
I wanted to do a quick check to see how that primary long-term forecast has held up in the week since:
When it comes to forecasts like this, transparency and honesty are incredibly important. We should expect to be wrong when trying to predict an unknowable future and, when we are, figure out why and determine how we can improve.
In this case, I wish I could say that I had some crystal ball or magic insight into disease modeling - but really, I’ve been really lucky so far. My simple forecast has held up pretty well.
The first couple of days were uncanny, off by only a few infections. Then, it seemed like the virus was taking a slower trajectory than the forecast was predicting, which would be amazing news.
Unfortunately, the last two days have seen a serious acceleration in total infections and not only caught up to the original forecast but exceeded it. I am hoping that this is just a temporary bump and things will slow once again.
I’ve made many mentions in previous posts about forecasting as a useful benchmark. So far, while this has been an accurate forecast, that still means we could end up with a total of 89,000 infections by the end of all this.
If reality continues to outpace this forecast, given the exponential nature of viral spreading, the final infection count could be tens if not hundreds of thousands of cases higher. That doesn’t even take into account a pandemic scenario where this outbreak is truly uncontainable.
This isn’t meant to fear monger but more to communicate that this forecast is already a benchmark representing a truly terrible outcome. We want to see the actual data come in below the predicted amounts, as many lives are at stake.
So, let’s continue to hope reality undershoots this forecast. I will continue to monitor new data releases give updates as material information comes through.
Enjoy the article? If you haven’t already, subscribe for free below and never miss a post. No spam. Cancel whenever!