So, I know last week there was this huge wave of hype that Trump is so screwed and that Harris is gonna win and blah blah blah. This mostly came out of a huge round of polls released last week that seemed to dramatically change the race, shifting many states 1+ points toward Harris. However, as I write this, more polls have come in and many of those gains are being mitigated.
One cannot go by individual polls, and expect to extrapolate some sort of conclusion about the elections. Even a good wave might not mean much in the grand scheme of election season. Before that wave, I was actually starting to wonder if Harris was beginning to backslide. because every day I'd see a poll make Harris's odds a little worse, a little worse, a little worse. She's at 58%, no wait, 58%, no wait, 52%, 50%, 48%, now 51%, now 60%, etc. That's how it's been going. Like, here's the chart that represents the overall trend of the polls based on the tipping point state (which is what I define a candidate's odds by):
As we can see, since around August 10th or so, we've basically been in toss up mode. And we've never really gotten out of there. I consider races a tossup if the confidence I have in either candidate winning is basically 60% or less. harris peaked at 60% last week, and is down to 56%. Keep in mind, I gave Hillary the same 56% on election day in 2016 and look at how that turned out.
You cant be certain AT ALL at that rate. You literally might as well flip a coin. 56% is basically saying "4 in 7", meaning take 7 random outcomes, Trump would win 3, Harris 4. It's not quite 50-50, but it's very close to it. So no, Trump isn't screwed, it isn't over for him, he's still a very dangeous threat to democracy with a high chance of winning, even if he is the underdog by this point, and yeah, don't count trump out. He can win, he can come back, no, I'm not saying this because I'm a Trump supporter huffing copium, this is a statistical fact. The same statistical fact got him into office in the first place anyway.
As I see it, here's how I grade probabiilities:
50-60% Tossup/tilt (=<1% polling average)- Means that one candidate might be slightly favored to win but the race is basically a coin flip.
61-84% Lean (1.1-4% polling average)- the race leans a certain way, and one candidate is statistically favored to win, but keep in mind, 40% is 2 out of 5, and 16% is 1 out of 6. So an upset can happen.
84-97.7% Likely (4.1-8% polling average)- race is heavily leaning in a certain direction, but perhaps not to the degree that it would pass the rigors of statistical significance (95% confidence in a two tailed test).
97.7%+ Safe (>8% polling average)- race leans so far in such a direction that an upset is statistically insignificant and should happen 1 in 20 times or less (1 in 40 in a one tailed test like here).
Since august 10th, the race has functionally been a tossup. And while I would give Harris a slight advantages and if i were forced to call the race either way it would be for harris, I'm really not sure. Again, take 7 random outcomes, Harris wins 4, Trump wins 3. Harris is favored but not by much. I dont care if a single new poll has Harris up 5 or 6 in PA, especially when the next day two trump +3s will come out to flip it back the other way. We dont know. Let's face it, no one who is honest fricking knows. And I dont take anyone seriously who claims to know (allen lichtman included). Buckle up, the next 6 weeks are gonna be WILD.
No comments:
Post a Comment