agsalaska said:
2023NCAggies said:
https://www.foxnews.com/media/polling-guru-gives-donald-trump-66-chance-winning-presidential-election
Silver thinks it will be Trump
Thats a big change. Two weeks ago he was like 51-49 Biden.
I think what we are seeing is as we get closer to November the modeling starts relying more on current polling and less on historical data. I suspect if the polling doesn't change much his modeling will show Trump even more likely.
I can't remember exactly where Silver was in 2016, maybe 75-25 Clinton. He and Cook both knew that Trump had a real if narrow shot and even spelled out where. Funny thing is he got roasted for saying it by the Dems before the election and the Republicans for some stupid reason after.
He isn't with 538 anymore. But, same general model (aggregate polls, weight them by ranking, account for bias, factor fundamentals and Monte Carlo all the paths to >=270 and publish probabilities).
Individual States in '16 he was pretty close, but when tight wins are within the MOE, there's only so much you can do on a electoral forecast. He had Trump at 1/3 and HRC and 2/3, i.e., Hillary had more paths to 270, but Trump still had some paths to 270 in their models.
From Nate's post,
Quote:
This is less of a concern for him, though conditional on winning Wisconsin and Michigan and Pennsylvania, Biden wins the Electoral College about 97 percent of the time in our simulations.
Boils down to the rustbelt.