Nate Silver

Nate Silver @NateSilver538

@politicalmath Yeah I think that's right. — PolitiTweet.org

Nate Silver @NateSilver538

In fact, I found the result noteworthy precisely because the "too fasts" have usually beaten the "too slows" when questions like these are asked. There's some suggestion here of a shift in public opinion over re-openings. — PolitiTweet.org

Nate Silver @NateSilver538

This is kind of a weird characterization of the poll. It found 14% of Nevadans saying "too fast", 36% saying "too slow" and and 46% "about right". So while the plurality is "about right", there are more "too slows" than "too fasts". — PolitiTweet.org

Nate Silver @NateSilver538

It's worth keeping in mind that polls were unbiased *on the whole* in 2018, though there were some biases in different states/regions that counteracted one another a bit. https://t.co/juXo1os1ZW https://t.co/Yo5sTgPcyF — PolitiTweet.org

Nate Silver @NateSilver538

Of the four key Midwest/Rust Belt battlegrounds, we're getting a pretty clear hierarchy where WI and PA are more competitive than MI and MN. You wouldn't go so far as to say Biden doesn't need to worry about MI/MN, but they've become less likely tipping points. https://t.co/iAGM4p2JS8 — PolitiTweet.org

Nate Silver @NateSilver538

So Biden's losing ground in 3 of 4, but the one where he's gaining (MN) is arguably the most important of the four. Why? WI has been polled a ton, so no one poll will influence our model that much. The other 3 haven't been polled much, but MN has more electoral votes than NH+NV. — PolitiTweet.org

Nate Silver @NateSilver538

Biden's lead in our polling averages before/after these polls: MN +6.4 -> +7.4 NH: +8.4 -> +6.9 NV: +6.5 -> +5.9 WI: +6.9 -> +6.5 — PolitiTweet.org

Nate Silver @NateSilver538

Incidentally, you may notice that a LOT of the prominent Senate challengers this year have little or no experience in elected office. That doesn't seem to be as much of a disadvantage as it once was. You can run a perfectly decent campaign as a "generic" Democrat or Republican. — PolitiTweet.org

Nate Silver @NateSilver538

We find that there still *is* an advantage from casting more votes across party lines, i.e. moderates and "mavericks" continue to overperform other factors held equal, but there is so much party-line voting these days that few members of Congress fall into that bucket anymore. — PolitiTweet.org

Nate Silver @NateSilver538

The candidate quality measures we use in our model are experience (highest elected office held), fundraising and the presence/absence of scandals. All of those are becoming LESS predictive. In general, the incumbency advantage is decreasing, also. — PolitiTweet.org

Nate Silver @NateSilver538

Not hugely surprising, but we find—in running regressions on every House race since 1998 and every Senate race since 1990—that the importance of candidate-quality measures is decreasing, while the importance of state or district partisanship is increasing. — PolitiTweet.org

Nate Silver @NateSilver538

RT @Nate_Cohn: Tomorrow AM: NYT/Siena in Wisconsin, Minnesota, Nevada and New Hampshire — PolitiTweet.org

Nate Silver @NateSilver538

@thehowie Totally fair. Plus Spain's data has always been a mess and that's maybe the most interesting/important case right now. — PolitiTweet.org

Nate Silver @NateSilver538

I can never keep track of who's saying that Europe's covid response is something for the US to emulate and who's saying it's something to avoid. — PolitiTweet.org

Nate Silver @NateSilver538

So we finally got an Oregon poll and it's Biden +12. Actually not such a spectacular result for him given a +7 or +8 lead nationally, although if that national lead falls apart Biden has far bigger problems than Oregon. https://t.co/XHnxcVePJa — PolitiTweet.org

Nate Silver @NateSilver538

So 3 of the past 11 elections (2008; 1996 and 1976) featured late polling swings bigger than Biden's current lead in the *popular* vote and a couple of others were close. — PolitiTweet.org

Nate Silver @NateSilver538

@BenjySarlin @NickRiccardi I also think people's patience with closures will continue to erode over time. So that's another reason to frontload measures even if you have to put an explicit time frame on them. — PolitiTweet.org

Nate Silver @NateSilver538

@NickRiccardi @BenjySarlin I think it's important to distinguish between a vaccine being *approved* and one being widely *available*. So I don't object to that. The stuff though where it's like "even if a vaccine reduces severe illness by 75%, we'll still have to TKTK"...I don't know how rational that is. — PolitiTweet.org

Nate Silver @NateSilver538

@BenjySarlin Actually you see a lot of the opposite conversation, i.e. "we'll still have to be careful even after a vaccine", which I think is both unhelpful psychologically and possibly wrong policywise (depending on details but at some point there becomes an acceptable level of risk IMO). — PolitiTweet.org

Nate Silver @NateSilver538

In 2016, the Trump campaign did a much better job of allocating campaign resources to the right states. https://t.co/HxlrxrABmz Their strategy this year is bizarre, conversely, spending to defend states like GA and IA while Biden outspends them in the tipping-point states. — PolitiTweet.org

Nate Silver @NateSilver538

Then again, if Biden is winning the sorts of voters you need to win ME-2, he's probably also winning the Midwest + PA, in which case the scenarios where it matters probably won't come into play. — PolitiTweet.org

Nate Silver @NateSilver538

Our model's prior in ME-2 is pretty pro-Trump, but the polls are starting to move it to more of a toss-up. If Biden is improving relative to Clinton with older working-class whites, it's pretty winnable for him. https://t.co/QFFPODAQ3u — PolitiTweet.org

Nate Silver @NateSilver538

@mattyglesias idk, I think there's been a fair bit of magical thinking in the notion that this trade-off is easy to avoid, at least until we get rapid, at-home testing. — PolitiTweet.org

Nate Silver @NateSilver538

So much of US covid policy seems to be driven by trying to fulfill two somewhat incompatible mandates ("we can't let people get covid!" / "we can't keep society shut down!") rather than any acknowledgment of the trade-offs, etc. — PolitiTweet.org

Nate Silver @NateSilver538

RT @dandrezner: Shy Biden voter alert! https://t.co/XvBCv3zTJP https://t.co/ihY8YTjpKl — PolitiTweet.org

Nate Silver @NateSilver538

We have a good idea of where things stand. Maybe Trump's still in a mini convention bounce and Biden's lead grows to 8/9? Maybe voters give Trump credit for the improving economy and it shrinks to 6/7? But mostly we're waiting for the debates and any October Surprises. — PolitiTweet.org

Nate Silver @NateSilver538

The fact is the presidental race is kind of boring from a polling standpoint right now. (Tho please don't hesitate to refresh my website 400 times a day.) — PolitiTweet.org

Nate Silver @NateSilver538

@NickRiccardi Yeah, we'd been including oversamples as part of the sample size, but we're going to change to the weighted sample size as that should give a better estimate of the true MOE. — PolitiTweet.org

Nate Silver @NateSilver538

It's a bit of a weird mix. Generally a point or two worse than the polling averages for Biden in the core tipping-point states. He's ahead so the bar is high. But some of his better results come in states (GA, CO, ME) that haven't been polled much lately. https://t.co/DGTCfza88d — PolitiTweet.org

Nate Silver @NateSilver538

@thecity2 There's no grade because these groups (actually a collaboration between various well-known pollsters) are new, but they're assigned a default rating based on their methodological characteristics, which should be equivalent to around a B grade in this case. — PolitiTweet.org