Who will be the next US President?

[Old Greg]

I think Nate may have got his Bayesian priors a bit off. We'll see.

I make a profit if Hillary wins, a slightly smaller one if Trump triumphs. Which will be a small comfort as we kiss goodbye to a functioning ecosphere...

Except we won't really see. Nate Silver's view is that there is a lot of uncertainty so he give a wide range of results. He also shows a higher probability of a HRC landslide than other models.

[pjclarke]

Fair point.

Actually, in the event of a Trump landslide, there'll be money to be made buying cheap US shares and funds on the inevitable slump and equally inevitable rebound.

[Old Greg]

Fair point.

Actually, in the event of a Trump landslide, there'll be money to be made buying cheap US shares and funds on the inevitable slump and equally inevitable rebound.

If you look at the fivethirtyeight figures of:


Hillary Clinton
71.9%
Donald Trump
28.0%

You can see that there is a tiny gap between their combined vote and 100%. That gap is the 13.2% probability of McMullin winning Utah and no candidate reaching 270 and the House breaking the deadlock in favour of McMullin (and also a 0.3% chance of Johnson winning an electoral college vote etc.).

[pjclarke]

Yes, I heard that. McMullin between 800-1000 at Betfair.

Clinton by a modest margin. But I won't be staying up.

[missinggreenfields]

Current oddschecker:

Clinton 2/9
Trump 9/2
Sanders, Biden, Ryan, Kaine 250/1
Kasich, Pence 500/1
Johnson 2500/1
Stein 5000/1

[SimonMac]

Any one watch Rich Hall's Presidential Grudge Match last night?

Really interesting on the whole process and history

[Old Greg]

If Trump wins, will they update the Doomsday Clock?

[jamesbrown]

I think Nate may have got his Bayesian priors a bit off. We'll see.

This election is largely about turnout and the assumptions made about the split among undeclared/undecided voters, which is unusually high. For me, Nate's probabilities pass the smell test more than others, but they aren't testable, of course. Actually, there are several factors in this election that make it exceptionally difficult to model.

[Old Greg]

This election is largely about turnout and the assumptions made about the split among undeclared/undecided voters, which is unusually high. For me, Nate's probabilities pass the smell test more than others, but they aren't testable, of course. Actually, there are several factors in this election that make it exceptionally difficult to model.

Yes, it seems a pragmatic approach.

[LondonManc]

Current oddschecker:

Clinton 2/9
Trump 9/2
Sanders, Biden, Ryan, Kaine 250/1
Kasich, Pence 500/1
Johnson 2500/1
Stein 5000/1

I've had a tenner on Trump; seems exceptional value at that price. I don't necessarily think he'll win but anything longer than 7/2 is worth a punt.