If you believe that humans have certain rights then it matters who is a human. I don’t think Trump thinks much about what rights humans possess.
I feel like this time I need to defend our democracy and it matters that we all vote given that the last election had high turnout on both sides. It maybe the same this time again.
Without hearing more of what he’s saying I’m inclined to think it’s a combo of sloppy mid-sentence re-phrasing by Trump and selective editing rather than an actual statement that women aren’t human.
I feel kind of dirty defending Trump, but that clip isn’t all that damning to me.
Plenty of other reasons to hate Trump, of course.
Many people have died earning the right for us all to vote in the US. Small sacrifice to exercise that right, impactful or not.
I found a few more futures markets (Kalshi & IBKR)…these are “who will win the POTUS” markets…I’ll break down the EVs eventually if I get a round tuit.
Polymarket: 55R - 45D
PredictIt (party): 49-53% R, 47-54% D
arbitrage
actually, there’s a bit of arbitrage if buy the “No” contracts right now:
That’s a sweet, sweet, 4% return.
PredictIt (Harris): 47-54% chance she wins.
PredictIt (woman president): 48-53% chance
PredictIt (POTUS by name): 49-52% Trump, 47-54% Harris
Kalshi: 48-53% Trump, 48-53% Harris
All of the polls & markets I’m tracking have finally converged…
Polymarket, IBKR, Kalshi, 538, 270, & RCP:
D | R | Δ |
---|---|---|
251 | 287 | 36 |
They don’t, however, agree on which states are likely to flip. It’s some combination of WI (10) & MI (15) might go R and NV (6), PA (19), and maybe GA (16) & NC (16) might go D. Harris, of course, would need at least three of those…and these things are highly correlated.
Predictit has a 10% fee on gains so the 2 contracts would cost 96 and return 95.
Never let the facts get in the way of a good arbitrage opportunity.
Selzer is saying in interviews that it isn’t that Trump’s support in IA has declined, but rather that the previous undecideds are breaking hard for Harris. In PA, the Harris internal numbers allegedly say that voters who have decided in the last week are going for Harris by double digits. I think that she might hold MA, WI, PA, maybe AZ, and pick up NC.
Nate Silver has added this chart to his model’s outputs:
(That’s actually just the first page of 9, showing the odds of all combinations of outcomes in those states…)
cdf of outcomes:
…although those will presumably change slightly as the final polls come into the model today.
I am worried by some of the comments made by Michigan rep debbie Dingel. I trust her judgment and she is claiming that the race will be very tight in Michigan.
It’s also possible that Iowa is leaning towards Harris because of the tariffs.
This is not good analysis because he most likely assumes independence of state level outcomes which frankly is dubious at best. For instance, if Harris wins NC then there is a very high chance that she wins everything else, no? So he is clearly underestimating the scenario where Harris wins every swing state.
He doesn’t assume independence of states. That is why he had Trump at 30% in 2016 while everyone else had Trump at < 1%: Silver was literally the only poll aggregator who wasn’t assuming independence between states.
In the discussion of his model, he’s written that there’s a correlation metric in play.
I probably shouldn’t have described the results as being “odds” – they’re just the percentages of simulations in each run that produced those particular results.
(EDIT: I think you can see the correlation mechanic manifest by virtue that the all-red and all-blue scenarios are the ones that most often occur in that run of his model.)
Ah yes, you are right, that would be impossible under the independence assumption. So does he explain how he got the correlations?
The top chart certainly illustrates the importance of the PA result in his model.
I’ve always assumed that he does a Bayesian mixed binomial model.
This is very interesting analysis, actually. Seems quite plausible if you analyze it carefully. Although I feel bad that my prediction has only 3.2% chance!
Update from the morning run of Nate Silver’s model:
He’s now got Harris with winning the Electoral College in 49.2% of 40k simulations.
results by state:
I suspect there will be another release tonight, and possibly one final release tomorrow.
I assume there is some general underlying movement (“Swift Boat,” Sarah Palin says something stupid, Trump “dances” for 45 minutes, etc.) that sways all the swing states’ swing voters to varying degrees.
Like, when I decide not to buy something, a million others decide not to buy it. (No fixing the ten million that do buy it, whatever “it” might be, be it an idea or a product or a service.)
Exactly how much for each occurrence? Trump says and does something stupid 24/7. How could anyone keep up with the reactions of voters?