Well, you can't expect this guy to "believe" math when he doesn't understand it:
Clinton thus jumped to a 14-point lead over Trump in the NBC/Wall Street Journal poll on Monday, only to see that lead significantly drop to nine points in the same poll just one day later. That's a 36-percent swing in 24 hours.
The Rasmussen poll had Clinton ahead by seven points on Monday, but Trump took the lead in that poll by Thursday morning by two points, a swing of 128 percent!
No, that's a 5 percent and 9 percent swing, respectively. This is the same as when a food gives you a "200% increase" in the odds of you getting some obscure cancer, when the study measured a 0.001% chance in on cohort and 0.003% in the other. It's just wrong, and it's one of the many ways media unintentionally mislead the public--reporters aren't generally well versed in statistics.
I'm of a mind with Silver. I think it would be far more likely than not for someone to under report voting for Clinton. To get truthy myself, I think that Trump supporters are macho, and friends of Trump supports feel the need to keep up. When your buddy starts making fun of you for not drinking Miller Lites at the same rate as him, you have an incentive to lie about your beer chugging skills. My guess is that there's a lot of people who are closet Clinton supporters, but who feel the need to keep up appearances. I wouldn't bank on it, however. Actually, that's not true. I gave Clinton 429 electoral votes in our office pool, by far the most of anyone who's playing. It's unlikely she'll get there, but I think the chances are well beyond non-zero.