Monday, May 20, 2019

How Do Expert Predictions Fare?

Eh, not all that well.

That doesn't surprise me. Even in strictly scientific experiments, only very narrowly crafted models will be able to be replicated. The closer those models are to reality, the less predictable they are.

And, both economics and climate are complicated. Any model based on those disciplines will only work in that idealized framework. The minute you try to take those results into the Real World, the model's predictability falls apart.

Economic is messy because the results can be disrupted by people - their laws, their impulses, their tendency to follow the herd.

Climate is messy because we are no ways near a model that is lifelike AND predictable. Small changes - sunspots, volcanic action, politically imposed solutions - all are disruptive mechanisms.

No comments: