Our server costs ~$56 per month to run. Please consider donating or becoming a Patron to help keep the site running. Help us gain new members by following us on Twitter and liking our page on Facebook!
(November 8, 2016 at 8:48 am)TheRealJoeFish Wrote: Well, here's how I'd say it: there are 182 EVs that it would be absolutely shocking, pundits-and-pollsters-need-to-start-from-square-one type stuff if Hillary lost. These are: CA, CT, DC, HA, IL, MA, MD, NJ, NY, OR, RI, VT, and WA.
There are another 41 that, the way polls have been going this election, are almost certainly Clinton's: ME (3/4), MN, NM, VA and WI.
And then there are 45 where Trump has never led, that really *should* be Clinton's, although there's a little more uncertainty. These are CO, MI and PA. This takes Hillary up to 268, which is her "firewall" everyone's been talking about.
The way polling's been going, though, she has a better-than-even chance in FL, NC, NH, and NV. This gets her the 322 everyone's been talking about. She could also pick up Maine's more conservative district (Maine gives 2 to the state winner and 1 to the winner of each of its 2 districts), Nebraska's most liberal district (similarly, Nebraska gives 2 to the state winner and 1 to the winner of each of its 3 districts), and (on the outer edge of possibility, in decreasing order), AZ's 11, IA's 6, OH's 18, GA's 16, or - based on recent polling, and this would be the real stunner - AK's 3. Her real "ceiling" is 378; anything beyond that would require flipping Utah, Missouri, South Carolina, or Texas, all of which are very solidly in Trump's camp at present after flirting with maybe looking just a tiny bit like they almost could be swing states if everything went right for Clinton (p.s. it hasn't).
It's the interplay between the "should be Clinton's" and "more likely than not to be Clinton's" categories that's interesting. It would be dramatic if, for instance, Clinton won FL or NC but lost MI or PA. The demographics say that should be possible, although the actual polling's not really saying it's likely. And this is what 538 does so well: they correlate errors across states. They don't say, "well, there's a 20% chance we're off in PA, and a 20% chance we're off in MI, so there's only a 4% chance we're off in both"; rather, they say "there may be a 20% chance we're off in PA, but if we are, it's probably due to factors that will greatly increase us the odds of us being off in MI too." That's part of the reason why they currently give Clinton a 72% chance, whereas a lot of other reputable sites are giving her 95%-99% chances (the betting markets are around 75% to 80% I think).
I'm on record as predicting a close one - 278 for Clinton, 260 for Trump. But that's just sort of my way of expecting what I perceive to be the worst of the "likely" outcomes. I think.
Nope, Nate has her above 300, me personally, since I started paying attention to polls in 08, but especially in 12, I really do not see this as being as close as the right wants everyone to believe. I have meddled with the interactive maps and Trump can win FLA and NC and Ohio and still lose. As of this morning Hillary is leading in FLA. Hillary will be over 300 and the election will be called between 10pm and midnight.
Well, I more or less agree with you. If I had to pick now, I'd say she's over 300. But when I make a prediction I stick with it dammit! All I'm really getting at with an official prediction of "278-260" is really what you're saying: reminding myself that "Trump can win FLA and NC and Ohio and still lose." I just need to keep repeating that to myself. All of the stat sites say the most likely outcome is 322-256, and I'm certainly a believer in statistics.
How will we know, when the morning comes, we are still human? - 2D
Don't worry, my friend. If this be the end, then so shall it be.
All I'm really getting at with an official prediction of "278-260" is really what you're saying: reminding myself that "Trump can win FLA and NC and Ohio and still lose." I just need to keep repeating that to myself. All of the stat sites say the most likely outcome is 322-256, and I'm certainly a believer in statistics.
