1 Comment

The "voting intention" models used by pollsters are nothing but the most amateurish possible forecasts.

They try to build a forecast from a single poll (red flag number one) and that forecast does nothing more than allocate "don't knows" proportionally to decideds.

This is a pseudoscientific approach that discounts the evidence that undecideds lean conservative - and conveniently ignores this is precisely what *just* happened in the Mayoral Elections, and why their "voter intention" forecasts failed miserably just a few weeks ago.

It's a little sad they demand to be taken seriously with a track record this poor

Expand full comment