Nate is basically a Bayes-theorem practitioner, which is all about probability and nothing about certainty.
The flaw in his Bayesian approach is that he starts with polls that IMHO some of which are pure fabrications, and if it’s mae-up, all bets are off.