|
Originally Posted by SanJuanStar
If you say so but if all your predictions are way off and always on the side of Democrats and on top of that you get paid by bias liberal corporations then if it walks like a duck , quacks like a duck, looks like a duck, it must bu a duck. He gives them what they want to hear. Great for donations for the party, he gets a cut.
I'm still waiting to see his secret formula that got him so way off in 2014, 2016, 2020 that they had to change it's formula after the chopping block in 2023 that ABC hired Elliot Morris from the Economist (another liberal rag) to head the site as editorial director of data analytics which Silver didn't like but then again ABC/Disney owns it and pays the bills.
Let's resume Nate paid by ABC/Disney
2014 Senate 36 seat race: "An equally important theme is the high degree of uncertainty around that outcome. A large number of states remain competitive, and Democrats could easily retain the Senate". Republicans won 24 to 12 races and flipped 9 seats. It wasn't even close.
2016 Democratic primary. FiveThirtyEight argued that Senator Bernie Sanders could "lose everywhere else after Iowa and New Hampshire". Result Sanders wins 21 more times for a total of 23. Including 538 getting Michigan wrong.
2016 General election: Gives Hillary 85% chance of winning in August and by election time 71%.Final Election Update: There’s A Wide Range Of Outcomes, And Most Of Them Come Up Clinton. Not only he got wrong the blue wall. He predicted Hillary would win Florida and North Carolina. He was off by 100 electoral votes
2018: ABC reshuffles and sends him from ESPN to ABC.
2020 General Election: Here We go again, He gives Biden 89% of winning including winning Florida, North Carolina and the 2nd Maine Congressional district, all 3 wrong.
Results: the election was decided by 40k votes in 3 battleground states by less than 1% and both won 25 states each.
So he was way off. It wasn't 89% to 10%. It was more like 50.1 vs 49.9. He could have gone 60-40 and still be off but respectable but 89 to 10? LOL
Then his model predicts that Pelosi and the Democrats were adding seats in 2020, She loses 10 seats.
Then he predicts that U.S. Senate seats in Maine and North Carolina will flip blue. Loses both. He wasn't even close in Maine because it was 8.6% the difference.
So We don't know what's his formula but it seems part of it is having his big fat thumb on the scale while getting paid by ABC/Disney.
..........and in 2022, 538 predicted a larger GOP majority than what they actually won. No model is perfect, to expect it to be is unrealistic.
https://projects.fivethirtyeight.com...orecast/house/
I would do some more thinking on what probability really means. To predict the results correctly, but "Man, it was close!" is still getting it right. All probability measures indicate is the assumed likelihood of a particular outcome, those probability numbers include "tail" results that are narrower than what the model projects as the most likely outcome, but still indicate "victory" for one candidate or the other.
I bet $$$ on Biden winning WI, MI, and PA in 2020. I was sweating bullets, but guess what? He won the states, I won the money.
The progression of vote modeling over the last 20 years is really interesting. What Nate does well is now duplicated by other people, and AI type machines and can process precinct level data over several elections cycles in seconds. Nate tried to branch out into sports, betting, pop culture, but it was not successful. I don't think the man is interesting or a good writer, but he is a good statistician.
The issue with that is..........people don't understand stats...........so lots of Dems were in literal tears over Trump's win because "Fivethirtyeight said Hillary was going to win!!!" After going 50/50 in 2012, the expectation was that Silver would be the perfect crystal ball each time, but probability doesn't work that way. Being right 89% of the time means being wrong 11%. (责任编辑：)
|
