Nate Silver

Nate Silver @NateSilver538

It's hard to tell whether GOP leaders are being strategic and cynical (e.g. hoping to keep their base excited) when indulging claims of election-rigging, or are just wholly disconnected from reality. But in some ways I wonder if it's even possible to distinguish between the two. — PolitiTweet.org

Nate Silver @NateSilver538

@DavidLauter Yeah, for sure. — PolitiTweet.org

Nate Silver @NateSilver538

The error in non-swing state polls was often bigger than in the swing states (e.g Trump was way underestimated in states such as OK and WV) although those states didn't tend to attract a lot of high-quality polling. This was also an issue in 2016, FWIW. — PolitiTweet.org

Nate Silver @NateSilver538

This averages out to Trump +3.3. Probably pretty similar in national polls, where we'll most likely wind up with an error in the 3.5- or 4-point range I think. — PolitiTweet.org

Nate Silver @NateSilver538

Estimates of the polling error in swing states, making educated guesses about what the final numbers will look like in states that aren't done counting yet. WI Trump+8 IA T+7 FL T+6 MI T+5 TX T+5 OH T+5 NH T+4 ME T+3 PA T+3 NC T+2 NV T+2 MN T+2 VA T+2 AZ T+2 NM T+1 GA T+1 CO B+1 — PolitiTweet.org

Nate Silver @NateSilver538

Georgia, which has a Republican Secretary of State, has probably been the most transparent state in the country with respect to its vote counting process, with precise updates on exactly how many ballots were left to count in which counties. There is a lot of doublespeak here. — PolitiTweet.org

Nate Silver @NateSilver538

Good example of how the numbers of uncounted ballots are often pretty squishy (and gives Democrats a slightly higher chance of mounting a comeback in Alaska once all ballots are counted). — PolitiTweet.org

Nate Silver @NateSilver538

Yeah. Said this many times before, but I've always thought stuff like "there's never any going back to normal" was counterproductive public health messaging. There's a much better case for undertaking strong/costly interventions if they have a time frame with a payoff at the end. — PolitiTweet.org

Nate Silver @NateSilver538

Does not sound like the final batches of AZ ballots will necessarily be that favorable for Trump, and he needs them to be quite favorable in order to come back. — PolitiTweet.org

Nate Silver @NateSilver538

RT @poniewozik: There's a Black Lodge backwards-talk version of the election where the same results happen but PA counts its votes first an… — PolitiTweet.org

Nate Silver @NateSilver538

Especially given that those states are in different parts of the country (South/Midwest/West/Northeast). We've seen in the past few elections largely regionally-driven polling errors. Polls badly underestimated Trump in WI, but only modestly in AZ and PA and not at all in GA. — PolitiTweet.org

Nate Silver @NateSilver538

One slightly unrelated point, but the fact that Biden could lose any 2 states from among the GA/WI/AZ/PA group and still win the election makes his win a bit more robust (and a bit less "close") than if he had to sweep those exact states. — PolitiTweet.org

Nate Silver @NateSilver538

However, we've used long-run rather than short-run data to calibrate the model precisely because we didn't think the uncannily good performance of polls over 2004-2012 or thereabouts was necessarily sustainable given declining response rates to polls, etc. — PolitiTweet.org

Nate Silver @NateSilver538

Yeah, this is interesting. Polls have generally been more accurate over the period that 538 has made forecasts (2008-2018) than they have been over the long run. So there have been somewhat fewer upsets than the model expected based on the long-term data it was trained on. — PolitiTweet.org

Nate Silver @NateSilver538

So I'm admittedly kind of prickly about criticism after elections that both our model and our broader reporting handled quite robustly, which only happened because we put so much work into them. Especially given that the topline outcome + 48/50 states were "called" correctly. — PolitiTweet.org

Nate Silver @NateSilver538

There is a LOT of work that goes into thinking about how to model out poll error and uncertainty. Thousands of hours of painstaking, detailed work. It's hard. Also a LOT of work that goes into how to describe it visually, and verbally, which is equally painstaking. — PolitiTweet.org

Nate Silver @NateSilver538

We have years and years of well-calibrated forecasts. (If anything, underdogs have won a bit less than they're supposed to, although not to a degree that's statistically significant.) We know what we're doing with these probabilities. https://t.co/8oqu1OmvAz — PolitiTweet.org

Nate Silver @NateSilver538

The reason Biden's win probability was ~90% is precisely because he could withstand a fairly large polling error when Clinton couldn't, which is exactly what happened. Indeed, our model assumes polls are fairly error-prone. https://t.co/F47OcVHqNK — PolitiTweet.org

Nate Silver @NateSilver538

I've promised myself not to do too much dunking but I can't believe how many dumb takes there was like this based on preliminary results. Especially because Biden always looked to be in pretty decent shape if you were actually following the data. So for posterity's sake. https://t.co/HmjnNJsd8T — PolitiTweet.org

Nate Silver @NateSilver538

This claim, apart from being sort of irrelevant, is only true because there are still votes left to count. Biden should eventually win the non-California popular vote, too, I'd guess by ~1.5m votes or thereabouts. https://t.co/p7rsIYecey — PolitiTweet.org

Nate Silver @NateSilver538

I'd add to this: the political circumstances may have worked out fairly fortuitously, here. Trump/Pence are claiming credit for the vaccine. But Biden/Harris are taking over. Both sides are going to have some trusted voices promoting the vaccine as being a Good Thing. — PolitiTweet.org

Nate Silver @NateSilver538

Thread from me from a while ago on how to interpret polls about vaccine uptake. — PolitiTweet.org

Nate Silver @NateSilver538

Yeah this seems important! The Pfizer vaccine is a bear to distributed (two doses, requires cold storage). But it probably also implies good things for other vaccines, too. — PolitiTweet.org

Nate Silver @NateSilver538

In 2018, California had only counted ~2/3 of its vote by midday on the Monday after the election, so that would tend to make me tend to doubt the Edison estimate that 89% of votes have already been counted there. — PolitiTweet.org

Nate Silver @NateSilver538

California says it has ~3.6m votes left to count, whereas the Edison Research estimates used by the NYT and other imply ~1.8 million instead. Anyone know who's right there? Will makes a difference to the popular vote and maybe to some house races there. https://t.co/asjmHKt7A3 — PolitiTweet.org

Nate Silver @NateSilver538

RT @newschambers: Shares in Zoom $zm are dropping on foot of the vaccine news. https://t.co/bhrWMmpda5 — PolitiTweet.org

Nate Silver @NateSilver538

Really outstanding news if this holds up — PolitiTweet.org

Nate Silver @NateSilver538

Nevada's a good example of why you'll want to wait for all votes to be counted before coming to too many conclusions about polls. Biden lead there now up to +2.6 as more Clark Co. ballots are counted and will probably rise further. Poll average was +5.3. https://t.co/W8fMXBHBOr — PolitiTweet.org

Nate Silver @NateSilver538

Not to imply that any of this is strategic in the first place. — PolitiTweet.org

Nate Silver @NateSilver538

To the extent some voters like ticket-balancing and thus the president's party usually faces a disadvantage in races for Congress, probably not a bright idea for Republicans to be playing up ambiguity about who will be POTUS on Jan. 20 when those Georgia runoffs are on Jan. 5. — PolitiTweet.org