Everybody’s just struggling to find a new angle on polling results that are becoming basically trendless.
Just for context, that set of polls would represent a 64 to 28 electoral college vote split in Clinton’s favour.
Cook Political Report’s map, allows Clinton to lose every toss up and still get passed 270.
Princeton Election Consortium’s map(Sam Wang’s site) only has one tossup but allows Clinton to lose that and all “lean Dem” and still get over 270.
Sabato’s Crystal Ball map has no toss ups but would allow Clinton to lose all “lean Dem” and still just win.
Probably the brightest map for Trump is NPR’s, (very similar to ABC and NBC) which would still be a Clinton win if she lost all the toss-ups but a Trump win if he also picks up a little “lean Dem”.
Hard for me to think this is close. Trump kind of needs an historic collapse of Clinton support, istm.
Has no one else in the thread mentioned the Millennials for Trump photo with the Trump children?
I got it wrong… it’s the one in the middle who looks like a grown-up Draco Malfoy with weight issues. That’s Jr., right?
Jr has dark hair. It’s Eric that looks like Malfoy.
Ignorance fought. Danke.
You seem to have skipped over the currect RCP map, which has it at 229 electoral votes for HRC, 154 fro DJT, and 155 tossups (of which Trump would have to win ~75% to win). Most of the maps you posted appear to be from several weeks ago (August 15th, 16th, and 18th). Since then, the race has gone from ~HRC +6 to +3 (looking at the RCP average). It’s impossible to know yet if that’s the beginnings of a “historic collapse” or just some normal fluctuation in the polling data, but I think your 3-week-old data is misrepresenting the current state of the race.
Jr is the one who would be in the community theater production of American Psycho.
@HurricaneDitka:
I missed it because I was just going through 270towin’s list but I think that the RCP map is the odd man out including Pennsylvania at +6 Clinton in the tossup category.
The same thought has crossed my mind. The numbers look good for Hillary in the Electoral College, but in this crazy year, a bad performance (or two) by her in the debates, and a good or at least not-awful performance by Trump, could make a big difference.
Nice.
Pennsylvania is one of those states that RCP has right on the edge of toss-up / lean HRC territory. If you look at their full list of rating changes, you’ll see that PA has switched on 4/24, 5/10, 8/4, and 9/2. One more good poll for HRC and it’ll be back to “lean Clinton”
Anyways, my point was that your list was taken in the afterglow of a Clinton high. In the 538 forecast, her best day was August 14th when she peaked at an 89.2% chance of winning and Trump slumped to just a 10.8% chance. Since then, she’s declined fairly steadily to 69.1% and Trump has risen fairly steadily to 30.8%. That’s what’s been happening in the weeks since your 270towin maps were last updated.
Looking it over that RCP is incredibly cautious. To get in the “solidly _____” category it looks like you need +20! So New York at Clinton +17, which hasn’t gone Republican since Reagan, is only “likely Democrat”.
I don’t think you’re correctly grasping what “close” means.
Obviously if the election were held and these polls were in fact accurate, Clinton would win. She is not winning by a huge margin, but you don’t have to win by a huge margin.
What makes it close is that she is not very far ahead and it’s just the first week of September, not Election Day. Roughly speaking her lead in popular vote is four points, assuming the polls are right. Four points is not a big lead. It’s close. If Trump picks up four and a half points, your electoral college maps will look very different.
Looking at the number of electoral votes “Assigned” to Clinton by a projection map and saying “gosh, she’s at 332, that’s 62 more than she needs, she’s way ahead” is, if I may be a little blunt, the wrong way to examine probabilities here. The various state results are not, for the most part, independent events, and you can’t treat them as being individual dice rolls, conclude Trump has to make 10 saving rolls and therefore conclude his odds are astronomically bad. The state polls are connected events. There is no realistic scenario in which Trump, say, stages a comeback unique to Virginia and wins it, but makes no headway in any other battleground state. (Conversely, there is no realistic scenario where Clinton performs well in South Carolina and wins it, but blows it everywhere else.) These are, rather, CONNECTED events. If Trump generally does better over the course of the rest of the campaign, his numbers in almost all states will rise.
His challenge is to make up that ground in popular vote across ALL states in play, and the nation as a whole. If he can make up four points he can win.
Polls rarely swing that much after Labor Day. It looks close to you because the numbers seem kind of close but history shows it isn’t. Presidential candidates leading polls at this point in the campaign have almost always won. Yes there will be some volatility; yes everyone should still get out and vote against Trump but I think barring some major event I am much more confident in a Clinton win than 538 thinks.
538 has the same problem the rest of the media has: To show “unbias” and to generate views, they have to cover the race as being close.
And if I may be blunt in return, I was giving Trump huge swathes of states, so I don’t see how you could acuse me of thinking each is an isolated event.
Trump may have determined that not releasing his taxes may cost him a million votes; but releasing his taxes may cost him 10 million votes.
He may be wrong in that assessment, but really, that he has come to such an assessment is the only reason I can think of why he’s not releasing his taxes.
Oh my God, that’s horrifying.
Nate Silver has responded to that criticism many times with the point that it’s much more important to him to be right than it is for most media. He wrote an entire book about how pundits don’t care about being correct. To suggest that 538 is just another media source full of talking heads and bullshit is to ignore his entire ethos and body of work.
And it would be very easy to show this bias if it existed. He’s been predicting elections for over a decade now. If his predictions have been biased toward showing an election as “close”, it would be easy to show. Remember, he doesn’t just predict the overall race. He predicts the outcome in each state and the results for each congressional district. There’s lots of data to to look through for that bias.
But no one has produced any analysis that shows such a bias.
So, I’m inclined to believe that this claim is bullshit.