I know no one needs to see it but.... Yeah, I'm seriously bored and seriously procrastinating. :mrgreen:
Mashmont explained earlier in this thread that what he's doing is merely dividing the new value by the old value, and then multiplying the result to get a "percentage increase." To make his "predictions," it looks like what he's doing is multiplying a 4-day old value by an arbitrarily reduced "percentage."
Needless to say, this model produces some... issues.
Here's his 3/25 "prediction," and presumably why he thought there would be no more than 130,000 cases:
And his "fixed" "prediction:"
The most obvious flaw of this method is that after a few days, it suggests that the total number of cases will actually
drop. (This is indicated in bold.) Obviously, that's not possible.
The next issue is that after a few days, his predictions become ludicrously inaccurate. To wit: So far, cases are doubling roughly every 2.5 - 3 days. If that keeps up (which is likely,) then his prediction will be off by 330,000 cases. As a result, and I'm sure this will surprise no one, he will "fix" his predictions day after day, thus trying to hide the massive long-term inaccuracies.
It's really fascinating what happens when a total lack of math skills collides with wishful thinking, yes?