If you went back to October to look up a prediction a stranger on a message board made you lost too.
3 Likes
Best case scenario: We’re going to have an Avignon Trumpacy based out of Mar-a-Lago supported by 40% of the country. Every other plausible outcome is worse.
2 Likes
Anyone remember when a republican led senate committee published a 1000 page report saying that Russia interfered with the 2016 election to assist Trump…a week ago. Or when the Supreme Court ordered him to turn over his tax returns a few weeks ago?
Institutions may not have bathed themselves in glory, but to adopt Trump’s perspective that he’s “winning” is simply false. Like, do people believe Covid is over and done with just because Larry Kudlow says it is?
1 Like
A week ago and a few weeks ago is 3 months before the election lol
Both parties trying to outdo each other on the Israel and police side lol
https://twitter.com/sahilkapur/status/1298738089330970635?s=21
And as a result, what has happened? The 1,000 page report was barely discussed or noticed by 99% of the public, and Trump still hasn’t turned over his tax returns.
Yes, like 30-40% of people in this country do.
Unrelated to this thread and just sort of a personal musing, but I deeply respect people who admit when they are wrong rather than just instinctively digging in.
2 Likes
Cuse, a half dozen times.
1 Like
I have predicted him being more likely than not to keep power, not win. I’m on the fence on that.
I think Biden has a 99% chance of winning among votes cast.
I think it’s around a flip among votes that count.
I think Trump is a favorite to maintain power.
1 Like
This is both falsifiable and based on a pretty loose recollection, but my impression is that there is a relative paucity of polling these days. 538’s model has larger error bars when there has been lots of polling compared to when there has been little, and that is correct.
Simplest explanation is that Nate upped his variance for this cycle
3 Likes
This might also be true, but if so, he hasn’t been terribly public about it or the methodology behind it.
https://twitter.com/Nate_Cohn/status/1298805722621370370?s=19
https://twitter.com/Nate_Cohn/status/1298809160340058112?s=19
Biggest problem with Nate’s modeling is that is isn’t factoring in cheating or Trump successfully throwing the entire election into chaos.
1 Like
On top of Nate baking a lot of uncertainty into the model, including COVID-specific uncertainty, baseline assumptions about Presidential partisanship have changed. Polling 5+ points ahead and the history of 25 years of straight Dem victories in Presidential elections in those states told models in 2016 to give Hillary a high chance of winning Michigan and Wisconsin. But now her loss is a datapoint that cuts against Biden’s polling lead.
1 Like
