I’ve been watching elections professionally for going on two decades.
From a fair amount of experience I’ll say one thing: if your theory is based on some sort of systematic error on the part of most pollsters, you’re almost certainly in for a really long election night.