Coronavirus Equation Error Leveling Off as Percentage

I have some truly weird news concerning my coronavirus prediction equation.

I checked today, and my prediction is off by 20.9%. The last two times I checked, it was off by 17.5% and about 20%. I don’t know where I scribbled the latter figure down.

This is big news. The error in my equation is nearly constant. You may not understand what it means. If the error stays nearly constant, it becomes possible to make predictions with great precision. Maybe under 5%.

Imagine this. You have a tape measure. There is an extra inch at the end. Every time you measure something, you get an inch less than the correct figure. It’s still an accurate tape measure. You just add an inch every time, and you get the correct figure. Not a perfect analogy, but you see what I mean.

If my equation stays about 20% low, you can get a very accurate result from it by adding a percentage every time you use it. That’s astounding. It’s like hitting a hole in one every day.

It doesn’t mean I’m a genius or even good at differential equations or predicting epidemics. It just means the equation I came up with is extremely useful for creating good predictions this week.

Maybe I should post a few predictions and see what happens. Even if I don’t, my equation is already known, so if I posted the predictions after they came true, no one could say I was cheating.

I’ll use 20.88% as my future error. Of course, that’s relative to the actual figure. Relative to the prediction, it’s about 26.4%, so that’s what I’ll add.

April 1: 867,545
April 2: 934,081
April 3: 1,007,033

These are figures for early in the day, like between 10 a.m. and noon.

Wonder what will happen. If it works, how long will it work? Even 4 or 5 days would be neat.

The percentage error is increasing very slowly. Maybe I should take that into account, but I’m already going way beyond what I can reasonably expect to predict, so I think I’ll let it slide.