Or so he claims.
As stated already, Nate’s details are hidden. But in this article he put in a chart (about 2/3 of the way down under “how big is the Electoral College bias”) that maps the national polling mark to his win probabilities, that I found really helpful - Harris polling nationally between +3 and +4 maps to an ~85% win chance, polling between +4 and +5 hits a nearly 96% win chance. His 50/50 win chance mark is around Harris +2 nationally.
For those that like to follow FiveThirtyEight’s forecast: Harris increased markedly today. She had been stuck around 53-54 for a while, crept up to 55 yesterday, and today is at 59 (more specifically, 588 out of 1,000 trial runs).
Also per FiveThirtyEight, same link: The seven major battleground states (Blue Wall + NC, GA, AZ, NV) are ALL at 50-50 or better for Harris:
WI - 64%
MI - 60%
PA - 55% (this one is improving a lot)
NC - 51%
GA - 50% (also improving recently)
AZ - 50%
NV - 55%
Yes … it’s a bit of an up-and-down ride checking in on this stuff every day. Let’s see if Harris can maintain – or even build upon – this level of support.
53% to 59% is not a marked increase. It’s barely more than noise. I’d be looking for 75-80% before I consider it a “marked increase.”
He’s never gotten 48%, 2016 he won with 46.1% and 2020 he lost with 46.8%.
Yeah I’m adding a point of buffer to his traditional 47% peak. Anyway this cycle on 538 he’s hit lows in lower half of 43s and now is 45.7. The point we can agree on is that his range is pretty narrow and he is not far from his ceiling.
Harris is currently running 48.2 there and her ceiling is yet to be tested. Could she get over 50? And stay there?
Yeah, it is - it is 6 points. That is a marked increase.
If you go here-every pol after the 11th has Harris leading. Not much in the way of post debate swing state polling so far.
6 points is a major swing of you’re talking about the percentage of people who say they are voting for Harris. It is barely more than noise if you are talking about the percentage chance of her winning. It really means nothing.
I am personally going to resolve to not make this same point to @DrDeth after this one … last … time.
Imagine @DrDeth that you have a heart condition that you need surgery for. Do you feel significantly different if the doctor gives you a survival estimate of 59 vs 53%? Much less scared as you are put under? Much less worried about whether or not you will wake up on the other side? “Oh that’s SO much better!” Do you think the doctor parses those numbers as significantly different?
I think a 6% spike is a marked increase. Is it temporary, an outlier? Maybe. Does it guarantee a victory? Of course not. But a 6% jump on a 100 point scale amounts to…give me a second…carry the one…6%!
Seriously, I think the automatic “this means nothing” response to the slightest optimism or good news is becoming fascinating. Seriously, we won’t jinx her if we smile over positive stuff.
FWIW: Up to 60 at 538.
I do not think medical prognosis statistics are good enough for this to be meaningful. A big problem for prognosis (and reason for optimism if I have a disease but don’t feel too sick to be optimistic) is that treatment for many conditions has recently changed and improved. And historical survival statistics cannot take that fully into account.
Political polling might go the other way. As polling response rates plummet, it becomes harder to poll. If the response rate trendline was part of a model, it would push results closer to 50 percent.
I would be interested to know how response rates this year compare with 2020.
Apples and cauliflowers / Okay, would you feel not so good if Harris was 6% behind trump?
And note- just because someone disagrees with you,m it does not mean they dont understand you. This is a nice trend. It is not the end of the struggle.
Remember when Joe Biden was running 2% behind trump, and everyone was going “Get rid of the old geezer loser!”?
Yep
Mind you- it is not something to celebrate with fireworks and a brass band. It is just a nice trend.
I don’t remember anyone writing the quoted insult, or even a close equivalent.
And I remember some posters saying the opposite — that Biden was doing great except for one bad debate, and that if he dropped out, the Democrats would go into circular firing squad attack mode.
The problem wasn’t just his bring behind, but how static the race was. Nothing Biden did bounced the polls. That was more dispiriting to me than the absolute numbers. The trend line tells the truth.
If Biden had just 2% behind Trump in the forecast then no one would have been panicking. Even 6% behind. That’s clearly tossup. If Harris was 6% up in polling I would be very optimistic … and the forecast would likely be 80 to 90% probability or higher. Her not winning then would require a major systemic polling error.
Gawd I have no self control to stop myself banging my head.
In the polls.
Amazing that my memory is better than yours.
My “comment on polling” is that the margin of error concept often cited makes little to no sense.
First, it is based on the idea that there could be up to a five percent chance an individual poll is outside the margin. The number five is arbitrary. If a post of mine, or any of us, had less than a fifteen percent chance of being wrong, IMHO that would be excellent by SDMB standards.
And most discussion incudes multiple similar or averaged polls. Add the sample sizes together, and the traditional margin of error gets really small.
ON THE OTHER HAND, there are bigger reasons polls can be wrong other than sheer sampling bad luck. In the U.S., the response rate is as low as one percent. This creates big to impossible possibilities for error compared to door to door surveys in nations where people are very friendly, commonly home, and most cooperate (Example: Bengladesh).
Polls are great for showing trends. Harris is doing better than Biden did. Comparisons, like that Harris is doing better in Pennsylvania than Florida – probably, but less sure. Who will win? Less sure yet. And pegging a firm number to those levels of uncertainty is impossible.
.
.
The margin of error is not an arbitrary number. It is based on the number of respondents and other factors.
The calculation of the 95% confidence range for a given poll is not arbitrary, but deciding that 95% (and not say 90% or 99%) is the threshold for “the” margin of error is an arbitrary historical standard - everyone’s got their own threshold for comfort that varies with the subject matter and stakes of the outcome.
To oversimplify a bit: The confidence objective is inversely related to the confidence interval (the MoE). A sample size may prove to be inadequate relative to the confidence required, depending upon the results. In which case, you have to either sample more, or accept a lower confidence level or a wider MoE.
And the MoE is purely a mathematical exercise, given the sample size, the sample results, and the confidence objective. There’s nothing arbitrary about it, unless deciding on the level of confidence needed is considered so.