Yes, it’s alright to be scared or frustrated or pessimistic. But we shouldn’t allow that to bias our thinking. Harris has run an excellent campaign. She can win, even if it turns out she doesn’t.
Because we believe in science? Seriously, being cocky is only possible by ignoring the data or lying to yourself about it. I also don’t think it helps deliver a positive outcome.
I remain optimistic. I have I think very rational reasons for thinking that the polling error will significantly favor Harris this time. But being cocky is not justifiable on the basis of those reasons alone.
Yes, I meant to be facetious, but we ought not be cocky to irrational degree, or despairing to an irrational degree.
If polling is science, then so is astrology.
Polling is as much a science as meteorology is. That some choose to misunderstand what it can and cannot say is on them.
There is no non-delusional or lying to yourself and others to look at the data we have and be cocky about victory, any more than there is reason to be despairing that all is lost.
Well… how much do people receive good and accurate advice on understanding polling from the traditional media? Not a lot, right?
You are a kind of epistemological knight when it comes to polling, and I appreciate it. But the kind of resource you are is not known to the public. At least, not very much.
Tucker is weird. His entire life, since mommy left him, he has had a grudge. Mommy was a hippie and left Tucker and his brother to dad. Ever since then he has hated liberals. He has these punishment fantasies.
It’s only reliable about a day or two in advance, and even then it’s wrong half the time?
Or as a journalist friend of mine once said, “The only poll that matters is the one on Election Day. Until then, they’re just filler material at the bottom of Page 2.”
Harris has done very well. I think she has done about as well as can be expected and she has a real shot at winning. But there are so many bigots in this country she has a steep hill to climb in order to win.
That was more or less true of meteorology until the early 1990s. Since then, thanks to computer modeling of past parameters and greater timely input of current parameters, it’s more like “reliable about five or six days in advance, and even then it’s wrong about 15 or 20% of the time.” Remarkable, really.
The relationship of polling to elections has some parallels to how weather reports work (especially forecast models that use polls as their main input), but polling has other built-in sources of unreliability, so it will never be quite as “good” as weather forecasting. But it has gotten somewhat better (useful for some purposes, not all) over the decades, despite new challenges like cell phones; and these “problems” are of course magnified when polling and/or elections are very close (only because we want to know who will win, so the difference between 49 and 51 means more to us than the difference between 64 and 66).
But you knew this.
If Trump had not been shot, I think he would have little chance of winning. Yes, it is impossible to prove. But if I am right, that does not have anything to do with the number of bigots.
Here in Maricopa county the mail-in ballots are identical to the ballots handed out on election day except for the creases made from folding and putting them into the envelopes. Like those, they are counted by machine and are counted as they are received up to the Sunday before election day, but any results are not released until the polls close. The ballots received after Sunday are held and counted after the polls close.
If you have enabled comms with the registrar you will receive, by email and/or SMS when your ballot is being prepared, in the mail, received, and the signature (on the envelope) verified. After a fiasco for the primary election in 2016 when the number polling places was reduced by a third leading to people waiting in line for hours to vote, the permanent early voting list (PEVL) was established. You need only to ask to get a mail-in ballot once and will be on the list until you opt out. Roughly 90% of the votes for both the primary and general election are mail-in.
I don’t think Trump being shot at (he got hit by a ricochet at most) moved his popularity one way or the other. Fox News, Trump, and some MAGA talking heads have tried to make hay out of the assassination attempt, but your rank & file MAGA followers don’t seem to care that much about it.
Agree. The D’s need a prominent figure who has the panache to do a Jimmy Johnson 1993.
Imagine if Obama or someone similar came right out and said it: “We will win the election. And you can put it in three-inch headlines.”
I wish meteorology was that good.
I like to ride my bike to work but don’t like riding in rain. I’d love to know with confidence if it is going to be raining at the time I’ll be riding home. What do I get? A 40% chance of rain that hour. 60% chance of not rain. Do I ride? Are they wrong when it is raining on my way home?
Hurricane Milton wasn’t predicted to be as powerful as it became until a few days in advance, and exact landfall locations were only able to approximated. The tornados across the state were not well warned about.
Meteorology, with huge sets of data, can do an increasingly better job with probabilistic forecasts. And they remain probabilistic. The huge snowstorm sometimes doesn’t show. Especially that first one of the season that the media hypes with hourly storm tracker warnings and ends up as a dusting.
I agree, though IMHO it doesn’t matter how well the Dem candidate does this time around, in terms of predicting the outcome. I think a big problem is that social media* allows networks of various cohorts to be connected and fed, so we don’t really know what’s going on out there behind the various headlines and MSM providers. I believe that this is so significant that only professional intelligence agencies would have any chance of even skimming the surface of what’s out there.
*Note: I’m not saying that social media is the root of all evil or anything of that nature.
DesertDog is correct here, except that the PEVL is now the Active Early Voting List (AEVL). In 2022, the Republicans changed the law to make that if you don’t vote early at least once in two two-year election cycles, you are dropped from the AEVL (unless you respond to a letter stating you are about to be dropped).
[quoting link]
In an election which is going to garner at least 150 million votes cast, I simply don’t believe that any sampling of 2,500 people can provide a reliable prognostication. (To use the hurricane forecast analogy, this is like trying to predict a hurricane by examining one square mile of the Atlantic Ocean for weather patterns).
These polls are junk science. That’s why they get “weighted” and “balanced”; it’s taking random numbers and then playing with them to find a basis for armchair analysis. Then, when they are inevitably “off” from the actual tabulated results, they become fodder for even more “analysis” of the deviations and variations in results.
It certainly gives political wonks something to chew on. But I just don’t believe they are reliable.
The weather forecast is based on observation points, not everything happening in a square (or cubic) mile. The nearest two observation points could be dry with rain in between.
I will give you this. When the forecast says 40 percent chance of rain, I am truly confident that the real chance of rain is between 35 and 45 percent. When Nate Silver says that the chance of Trump winning is, say, 52.3 percent, I do not think the real probability, if we could know it, is as close as with the weather percent.