I wnder if this is the ad that he is so upset about.
Eliza Doolittle (conservative icon):
“Hurricanes Hardly Happen”.
Not to put too fine a point on it but Wang has been right the whole time. This is a low volatility race. His baseline meta-margin was about Clinton +4 and it was pretty much expected to travel up and down 3 from that transiently driven by news cycles (like the Khan attacks one way and Clinton’s illness the other) but center of gravity that it would return to was near that +4. So here we are … 538’s PollsOnly calling it +4.6, NowCast as +5.2 and Wang’s slower to move meta-margin at +3.1.
It aint over. Odd things can still happen. But the panic some have cycled through here based on news cycle noise has been impressive.
Wow - never even heard anyone say that he was. Pretty realistic image, though, in the Snopes article.
I have to wonder if this will depress turnout among his base. They’re not exactly conversant with the real world, and likely consider Snopes to be a Far-Left Plot to Destroy America.
This is extremely common among people who claim that other people get offended too easily: They want to offend others, but once you offend them, well, that’s just too much! That’s terrible! You need to be brought to heel, or else.
They’re the ones who are accustomed to having an “or else”, after all; they’re the ones the world (or at least our culture) used to coddle and protect from anything seriously offensive, while allowing them to offend others with impunity. Now that their behavior has consequences, and the others can offend them without being slapped down instantly, well, the world offends them, and they need a safe space.
Everyone needs a safe space sometimes. What annoys is when people are hypocritical about it, and especially when they try to deny to others things they jealously guard for themselves.
Being an asshole from California who is not effected by this, I hope Hurricane Matthew doesn’t have any effect on the election.
Don the Con seems to be very angry about people playing back what he said.
I’m getting this image of Don as Dr. Strangelove. While tweeting his left hand is trying to keep his right hand from typing. When he hits send, he stands in front of the mirror and his right hand tries to keep his left hand from slapping himself.
Then he stumbles back to his phone and cries out “Mein Putin, I can Tweet!”
The problem with Wang is that his aggregation was out of touch with nearly everyone else. I don’t mind that the prediction doesn’t move–it shouldn’t, and that’s why I see 538 as flawed. The best predictions stay as steady as they can, which means you should have things move slowly.
But the aggregator should respond quickly. His doesn’t. His lower volatility is because his entire system is slower moving. Maybe that’s a superior measure. But I hate how he then uses that to shit on others. He even pulled a Trumpian “sad” about Nate Silver in the comments, just because someone quoted his tweet about how the VP debate might play on SNL.
Trump is touting a Reuters poll on Twitter saying he’s ahead in Arizona, Colorado, Iowa and Georgia.
He also claims leads in Pennsylvania
and Virginia
The actual Reuters site puts Clinton up by 40 to 33% over Trump nationally.
You notice how his graphic people tilt Clinton’s name slightly? I’m surprised they don’t embed a sad trombone when you mouse over her name.
Why?
Do you think that within 538 NowCast is intrinsically superior to PollsPlus because it moves quickly and bigger and PollsPlus moves slowly and less?
Wang’s method is based on only state polls and can only move as state polls come in. Yes, as a result it moves less. Is that reduction of noise or reduction of signal?
His approach also does not come up with an expected value but uses a polynomial method to come up with a median value, which apparently also decreases the volatility.
Mr. Trump just lost a significant number of votes here in Nevada, i think:[
](http://hosted.ap.org/dynamic/stories/U/US_CAMPAIGN_2016_THE_LATEST?SITE=AP&SECTION=HOME&TEMPLATE=DEFAULT&CTIME=2016-10-05-21-27-57)People everywhere take names seriously. What Mr. Trump did will piss a lot of people off.
How do you think he’d react if someone called him Donald Drumpf and then told him “no, you’re wrong; your name is Drumpf”? :dubious:
I can’t wait for Language Log to get a hold of this, if only so I can understand exactly how Trump’s mispronouncing a very simple, straightforward name.
More seriously: This is pathological. This is a sign of some underlying pathology. It makes no fucking sense, it’s hurting him, and yet he’s still doing it. It’s the political equivalent of fecal finger-painting, or hoarding. He can’t be wrong, and he will demand he’s not wrong well past the point of those demands being consistent with a fully-formed adult personality.
I think Wang’s explanation of why other sites don’t calculate the probability in that way is rather unfair:
So basically, Wang thinks his competitors either lack “expertise”, or are intentionally sacrificing accuracy to give themselves more opportunities for “commentary”.
This omits a fourth reason, which is that Wang’s polynomial approach treats the probabilities of winning each state as independent, and some of his competitors (notably Nate Silver) don’t consider this a valid approach.
Pardon, but did you just now tune in? It’s Trump.
Yeah, everything you ask is correct.
I only ask because up until now I counted you among those who paid attention.
The fact remains, however, that Wang’s methods are accurate when it comes to predicting outcomes:
Ultimately, disagreement about these things is reasonable. It’s good, in fact; it gives everyone a chance to sharpen their tools and it gives us spectators a chance to see how close things are by comparing a range of rational people all analyzing the same data. It’s even reasonable to differ in the face of both methods performing well in an objective sense. Wang might be playing up the disagreement in this context, but in every other respect his website is drama-free and math-heavy, and it’s a constant wonder to observe such a simple model get it right time after time. I mean, it’s almost like looking at the evidence is a good way to predict the future…
Hey, I get a sense of wonder every time I look at a rainbow, too. Some things don’t change.