an educated political analysis is far more useful than anything this guy will ever come up with. and, that's coming from somebody with advanced degrees in mathematics.
listen to the political pundits, not the statisticians. politics is not physics, and it's just stupid to pretend that it is. this is just not a math problem.
you want to look at snap polls very close to the election, not models months away from it. the role of a model in an election needs to be to properly distribute data from snap polling, not to smooth it out. so, we need to find some way to take polling a few days before the election and apply it properly to the swing states. that's what a good model can do.
trump's numbers seem to rely more on reactions to terrorist attacks than anything else. if the next few months are relatively calm, he has absolutely no chance - and that is a rhetorical term, not a mathematical one. but, you want rhetoric here, not math. don't be fooled by scientific sounding quacks. but, if we see a spate of attacks, he could win on a wave of anti-muslim xenophobia. the clinton campaign has likely calculated the opposite.
put another way: he has to find a way to trick people into making a rash and poorly thought out decision. the data is pretty clear that he doesn't have a chance, otherwise.
but, please do keep in mind that the whole basis of our economy is to trick people into making irrational decisions.
i'm not going to make numbers up out of thin air and then cite arbitrary procedures that you don't understand in order to justify them.
http://fivethirtyeight.com/features/election-update-why-our-model-is-bullish-on-trump-for-now/
i think he got a bump from orlando, and then another bump from munich. munich just happened to coincide with the convention.
we haven't yet seen a factor that would give clinton a bump. but, i think that's reflective of her failure to frame the narrative. for better or worse, trump appears to be in control of the narrative. that is, the fact that it's august and we're not sure what issues might give her a bump is pretty pathetic.
there's still time for clinton to take control of the narrative, but it seems to me that this election is going to be measured in terms of reactions to global terrorism - and he owns her on that issue, whether her campaign likes it or not.
that doesn't mean he's going to win, mind you. the conservatives controlled the narrative in canada in the last election, but that didn't help them win because they were so unpopular.
she should be running on the economy.