Flipping a coin

The model behind the air quality forecasting system has an error rate of 50%.

In the words of our dear minister Vidal:

“The predictive model has to be fixed; if it fails, it has to be fixed,” and he added, “let’s be frank, a mistake was made here, and the reason for the mistake is a forecasting system with far too wide a margin, on the order of 50 percent. What Intendente Erazo has said is that we need a predictive model that is much more predictive and not almost random, because 50 percent is almost heads or tails.”

Well, mathematically a “perfect coin” has exactly that probability, 50%, and this time the minister surprises us with his command of the rudiments of probability theory.

If Vidal’s statement is correct, then flipping a coin is as good as applying the predictive model, and if the model was that bad, I wonder: are they only noticing now?

Maybe they’ve known for a while, but flipped a coin to decide whether to reveal the situation.