As much as it almost literally physically pains me to say this, Trafalgar cannot be written off after 2016 and 2020 results.
To follow up on that like a one-two punch to the gut, maybe there is something to their entire " social desirability" model for estimating Trump voters. Yes, it sounds inherently boneheaded when you first hear the concept, but one can't argue with results.
Of course come to think of it, how accurate were they in polling outside the presidential race? Maybe they are the Trump Whisperer among pollsters, but otherwise not great?
They do fairly ok in most states although are absolute garbage in Georgia for some reason.
