I really don’t trust the polls these days. There are just too many people who don’t answer the phone. I’m certain there’s a huge bias. And i have no clue which way that bias tilts.
I do think trends in polls are likely meaningful. But not the absolute value.
This didn’t happen. We’ve been over this.
This much I can agree with. No reason from polling to be optimistic either.
To stay a bit within the thread subject, polls don’t need to be skewed to have insufficient resolving power. She could in fact be either significantly ahead or significantly behind with the current state of polling. If WI has the same error in the same direction as 2020 then the current aggregated result on 538 portends a loss by over 6%, for example. Not as bad if it is off by as much as it was in 2016, then she loses the state by about 2%. Or it could go the other way!! No way to know. If everything else was accurate (spoiler, it isn’t) either of the first two mean she loses the election.
Results nationally are better than individual states, typically plus minus something like 3% is reasonable.
Probably though the least likely possibility? That they are exactly right or even within two in all the states within three or four points either way. And if not all mostly on the money then the odds are more likely they are mostly wrong in the same direction than randomly wrong.
If the polls today are exactly predictive of Election Day results Harris wins; if they are systematically off by just a mere 1% in Trump’s favor he wins the EC.
If you want to call that being ahead you are free to do so. To me this is the same as at best a one point basketball game with five minutes to play and both teams fighting for a rebound.
Except I went thru all the key states.
I am leery myself.
I am guardedly optimistic.
I’m not counting Democrats that stick their fingers in their ears saying, “nuh-uh”. How many times did she visit Wisconsin? Zero times in the last 3.5 months. Why were Democrats in Michigan begging her to visit?
Polls are entirely entertainment, not news, not accurate and not anything. Especially anything involving the name Silver. He just lusts after older blonde women, IMO. If Kamala went blonde, “Silver Predicts 98% win for Kamala!” would be the headline— in probably his own news sites. Jimmy the Greek was always more accurate.
Excellent analogy.
EXACTLY.
She spent her and money informed by what state polling was saying would be likely tipping point states to win the EC. Polling had her consistently up by roughly 5 in WI. 538 was giving it an 85% win forecast. She instead concentrated on three then swing states that were polling closer that could have each given her alternate paths with 15 visits per: Florida, Ohio, and Pennsylvania.
That was a very EC path considered and polling informed strategy. It failed because those many visits and lots of money spent did not deliver any of those then polling very close states. And because state by state polling both had a systemic error in Trump’s favor and one that was especially large in Wisconsin.
So, do you think she’s gonna do that?
If so, when would be a good time?
Yes, that’s probably the best sports analogy I’ve heard this election cycle.
Why not just count campaign events in each state?
| Clinton Campaign - Campaign Events (All): | 151 | 100.00% |
|---|---|---|
| Florida | 36 | 23.84% |
| Pennsylvania | 26 | 17.22% |
| North Carolina | 24 | 15.89% |
| Ohio | 18 | 11.92% |
| Michigan | 8 | 5.30% |
| Nevada | 8 | 5.30% |
| Iowa | 7 | 4.64% |
| New Hampshire | 6 | 3.97% |
| Virginia | 5 | 3.31% |
| Wisconsin | 5 | 3.31% |
| Colorado | 3 | 1.99% |
| Arizona | 3 | 1.99% |
| Illinois* | 1 | 0.66% |
| Nebraska (CD-2) | 1 | 0.66% |
| Trump Campaign - Campaign Events (All): | 248 | 100.00% |
|---|---|---|
| Florida | 35 | 14.11% |
| North Carolina | 31 | 12.50% |
| Ohio | 30 | 12.10% |
| Pennsylvania | 28 | 11.29% |
| Virginia | 18 | 7.26% |
| Colorado | 16 | 6.45% |
| New Hampshire | 15 | 6.05% |
| Iowa | 14 | 5.65% |
| Michigan | 14 | 5.65% |
| Nevada | 9 | 3.63% |
| Wisconsin | 9 | 3.63% |
| Arizona | 7 | 2.82% |
| Georgia | 3 | 1.21% |
| Maine | 3 | 1.21% |
| New Mexico | 3 | 1.21% |
| Indiana | 2 | 0.81% |
| Minnesota | 2 | 0.81% |
| Missouri | 2 | 0.81% |
| California | 1 | 0.40% |
| Connecticut | 1 | 0.40% |
| Mississippi | 1 | 0.40% |
| Nebraska | 1 | 0.40% |
| Texas | 1 | 0.40% |
| Utah | 1 | 0.40% |
| Washington | 1 | 0.40% |
Both tables show campaign events after party conventions.
Michigan was the 5th most visited state for the Clinton campaign and 8th most visited for the Trump campaign.
Wisconsin was 9th most visited by the Clinton campaign and 10th most visited by the Trump campaign.
Both campaigns focused on swing states.
Your earlier statement…
… is absurd when you look at where the campaigns actually focused their energy.
Source: 2016 Presidential Candidate General Election Events Tracker (maintained by FairVote) - Google Sheets
Jay Kuo (Status Kuo Substack) and Pennsylvania political commentator Joshua Smithley help shine a floodlight on recent Republican poll zone-flooding in Pennsylvania.
(my emphasis)
I can’t verify the bolded for myself – his model is a proprietary black box, and just a raw count-up of the glut of poll results doesn’t yield his current forecast.
Meanwhile, since July there has been a direct financial line from Nate Silver to J.D. Vance backer and conservative mega-donor Peter Thiel. “Propagandist” Hubbell pointed this out in his column last Friday and further noted the same poll zone-flooding mentioned by Kuo and Smithley in my post above (my emphases):
“Pay no attention to the man behind the curtain.”
I received a slew of emails today forwarding articles or posts about Nate Silver’s updated election forecast for the presidential race. Sadly, Nate Silver has become a partisan in the fight for democracy—and appears to have his thumb on the scale against Kamala Harris. Silver currently works for a firm owned by Peter Thiel—the Republican megadonor and weird Elon Musk wannabe.
At a high level, you should know that Silver has decided to “factor out” about 2% of Kamala Harris’s support in polls because he believes it is a temporary “post-convention bounce.” As I wrote previously, if Silver believes that the sea-change that occurred when Joe Biden withdrew is like anything that happened in 2020 or 2016, he should take up another profession. Moreover, Silver appears to call “tied” races in favor of Trump and includes dozens of low quality polls created by Republican operatives to “flood the zone.” According to Silver’s critics, he fails to discount those low quality polls when calculating the “average” of polls.
All that to say just because it’s “polls!” doesn’t mean it should accepted at face value.
Meh. The connection to Theil as an investor in a company to Silver is a very convoluted wink nod thing but whatever on that. My question really is what sort of polling results would I want the public to be falsely believing if I wanted to have my thumb on the scale against Harris?
As a partisan I think the public believing it is a close race with Harris even as a slight underdog is the preferred perception. Complacency of marginal maybe Harris voters is the greatest danger.
I’m taking it that you’re speaking hypothetically. In the event that you’re not:
Perception should have nothing to do with published polling analysis. I understand that removing all human bias is logically impossible … but polling should be as close to sacrosanct as feasible. Polling (or analysis thereof) shouldn’t be intentionally used to extract a lean from a given cohort, or to put up a false front obscuring the true state of a given race.
No, and I hope- never.
The NYT is obsolete
Yep. Silver sold his company and is now selling out. .
More rhetorically. You accuse Silver of putting his thumb on the scale against Harris in service of his alleged corporate master. Of course my take is that that is absurd, all the more so because making a not close election seem closer than it is would IMHO serve the Harris side well.
Sure, if the polls were showing Harris a 20% favorite. But the polls are showing Harris beating trump by 2-4%- and amount that could change overnite.
Silver sold out.
Do you think she’d have better or worse odds of winning if she did an interview with them? Or do you think it wouldn’t change the odds at all?
Hard to say. I mean the NYT subs are so solidly blue, I can not see how it would move the needle. I guess it could help a tiny bit, but is it worth taking electioneering time off to do it?