AI Generated Summary (I’ve been expirimentign with it):
- Kamala Harris had a tough day in the forecast despite gains in national polls.
- She leads by 3.8 points nationally but has a 47.3% chance of winning the Electoral College.
- The model adjusts for convention bounce, assuming her polls are inflated.
- Harris’s numbers may improve if she maintains her current standing.
- A concern is the lack of polls showing her ahead in Pennsylvania, a key state.
- Recent polls show Pennsylvania as a tie or slightly favoring Trump.
- Harris has a 17% chance of winning the popular vote but losing the Electoral College.
- RFK’s dropout and endorsement of Trump may impact her in Rust Belt states.
- Tim Walz has had a strong rollout as Harris’s VP, but there’s speculation about Josh Shapiro.
Confused. It says she has less than a 50/50 shot at winning the electoral college, but then “a 17% chance of winning the popular vote but losing the Electoral College”
I’m not a statistician, but what fucking math is showing any reasonable chance that Harris loses the popular vote? Because that’s the only way that math checks out.
Electoral win: 0.473
Electoral loss: 1 - 0.473 = 0.527
Electoral Loss + Popular Win = 0.17
I can’t lookup the formula right now, but this indicates that the chance of Harris losing the popular vote is significant
Wouldn’t this basically mean
Electoral loss = 0.527
Then within the electoral loss odds, there’s a 0.17 chance she wins the popular vote so it would be around
Electoral Loss + Popular loss= 0.357
Electoral Loss + popular win = 0.17
So she has 35.7% odds of losing the popular vote and thus is 64.3% likely to win the popular vote (assuming there is no data that shows her losing popular vote but winning election)
RCP has Harris at a 1.8% favorite in its polling average. At the beginning of this month Trump was winning in these polls (granted partially because of Biden’s degraded mental state). So opinions could legitimately change.
Additionally, Kennedy is still at 4.5%, and generally 3rd parties support degrades the closer to the election it gets. It’s difficult to say where that support will degrade to. If Trump’s support falls to 1% and Trump takes 3% of that support, he’ll be ahead in the polls.
Finally, all the national polls have a margin of error between 2-3%. It’s just as possible that Harris is up by 4.8% as it is that she’s down by 3.8%.
So the chance of Harris losing the popular vote is significant.
Wasn’t 2016 enough to show this fucking degenerate doesn’t actually know what the fuck he’s talking about?
He’s too busy jerking off people like Peter Thiel and Marc Andreesen to be a serious commentator.
Not disputing the he’s an idiot, but 28% likelihood events happen about once out of every three times on average.
The image below is not direct an indication that he’s an idiot. That stuff that he writes in his book and that pours out of his mouth, yes, that’s another matter.
Yeah, I think people confuse win probabilities with polling percentages.
If a pollster predicted that ten races each had a Democratic win probability of 70%, and the Democrats won all ten of those races, that prediction would be wrong—as wrong as if the Republicans won six of the races.
He (and the rest of 538, where he was at the time) were criticized at the time for giving Trump much better odds than most. They were still wrong, but less so than the rest.
This sounds like conservatives who said Fauci was an idiot because he first said not to use masks, then to use masks all the time, then to use them if you are indoors or have a compromised immune system.
Making mistakes and correcting them with new data is a sign of intelligence not stupidity.