Deleted tweet detection is currently running at reduced
capacity due to changes to the Twitter API. Some tweets that have been
deleted by the tweet author may not be labeled as deleted in the PolitiTweet
interface.
Showing page 201 of 910.
Nate Silver @NateSilver538
If you're operating on the theory that this is noise and swings are uniform, then maybe Biden improves to +6/+7 once the final PA polls are in and declines to +7/+8 in WI/MI. But, IDK, we've been waiting a long time for PA/MI/WI to converge and they haven't yet. — PolitiTweet.org
Nate Silver @NateSilver538
On the positive side for Biden, GA and WI/MI really stand out here. On the negative side, his numbers haven't really improved at all in PA, FL or AZ. — PolitiTweet.org
Nate Silver @NateSilver538
How our polling averages have changed since the before first debate... basically over the past month, in other words. Generally speaking, Biden has gained ~1 point or so over the past 30 days, but hard to find consistent regional patterns. https://t.co/5xSfLSENhE — PolitiTweet.org
Nate Silver @NateSilver538
10. Adding those (and marking past Trafalgar polls as R partisan) turned out to make no difference. Biden’s chances were 89 percent before and 89 percent afterward, while Democrats’ Senate chances were essentially unchanged (77.5 percent vs 77.1). Have a good Friday night, all. — PolitiTweet.org
Nate Silver @NateSilver538
9. Having resolved this, we've also added a backlog of Trafalgar polls that were conducted over the past few days to our averages. — PolitiTweet.org
Nate Silver @NateSilver538
8. One of these cases came to our attention in a (friendly) conversation with Robert Cahaly, Trafalgar's CEO. They have been helpful with some things and we know they're very busy right now, so we appreciate their time. — PolitiTweet.org
Nate Silver @NateSilver538
7. But to repeat: what we know now is a handful of Trafalgar polls were released without any initial indication of a partisan sponsor, when there had been one. — PolitiTweet.org
Nate Silver @NateSilver538
6. We're also having to be a bit expedient here because it's four days to the election. There are some further threads—some related to Trafalgar but mostly on more general themes—that we'd like to have time to more fully think thru and report out. — PolitiTweet.org
Nate Silver @NateSilver538
5. We’ve updated our policy to make this more explicit. But we’ll plan to revisit things after the election. We’ve been pretty relaxed in the past (i.e. not classifying polls as partisan when most others would) and we may need to get stricter. https://t.co/37N8bDqsnB — PolitiTweet.org
Nate Silver @NateSilver538
4. Therefore, in cases where a firm is doing a substantial amount of polling for partisan sponsors but has not been transparent about which polls are done for which clients, we'll have to go ahead and label all of their polls as partisan. — PolitiTweet.org
Nate Silver @NateSilver538
3. It's important to know the sponsor/client for many reasons, including that our averages handle partisan/internal polls slightly differently: https://t.co/FxbzezloWI — PolitiTweet.org
Nate Silver @NateSilver538
2. This doesn't neatly fit into any of our current policies, although it goes against the transparency that we generally ask of pollsters. — PolitiTweet.org
Nate Silver @NateSilver538
Quick thread. Warning: not super exciting. 1. In the course of our reporting on Trafalgar Group—part of the due diligence we often do while entering polls—we've learned that some of their polling was done for partisan clients that weren't clearly disclosed. — PolitiTweet.org
Nate Silver @NateSilver538
People seem to be conflating "they won't count (mail) ballots that arrive after Election Day"—true in some states, not others, and currently the subject of much litigation—with "they won't count ballots after Election Day, period", which would be orders of magnitude more radical. — PolitiTweet.org
Nate Silver @NateSilver538
This was a fairly weird day for polling with a lot of volume but relatively few high-prestige polls ... but it's hard to find anything much with a favorable trendline for Trump. https://t.co/9AeTKX5o0O — PolitiTweet.org
Nate Silver @NateSilver538
I spend 90% of my time thinking about the thing that will happen 10% of the time. — PolitiTweet.org
Nate Silver @NateSilver538
@jtlevy ~90% (I'm not sure how the rounding rules work) — PolitiTweet.org
Nate Silver @NateSilver538
Particularly if Ds with outstanding mail ballots are hearing messages from the Biden campaign etc. and are also now more likely to use dropboxes, or to decide to vote in person instead, it's not clear what the partisan composition of late-arriving mail ballots will look like. — PolitiTweet.org
Nate Silver @NateSilver538
In Florida, for example, mail ballots received by the morning of Oct. 14 were +21 D by party registration. But ballots received in the past 48 hours were just +6 D. — PolitiTweet.org
Nate Silver @NateSilver538
Gonna keep repeating this point, because it has a variety of implications for vote counting, what happens if some ballots come in after the deadline, etc.: There is lots of evidence that Democrats sent their ballots in sooner than Republicans. — PolitiTweet.org
Nate Silver @NateSilver538
RT @JustinSandefur: The pressure to shut down the PA vote count is going to be insane. Even if the 538 forecast for PA proves accurate (Bi… — PolitiTweet.org
Nate Silver @NateSilver538
RT @ezraklein: "So the Electoral College edge makes a big difference and is why there’s been this very bifurcated, binary kind of world whe… — PolitiTweet.org
Nate Silver @NateSilver538
When can you expect election results in each state? And which types of ballots will be counted first? You're going to want to bookmark this link. https://t.co/FuKt5gYZxU — PolitiTweet.org
Nate Silver @NateSilver538
Currently this has a fairly big effect on our turnout forecast in Texas, Montana and Hawaii, and a smaller one in Georgia, North Carolina, New Mexico and Washington. — PolitiTweet.org
Nate Silver @NateSilver538
RT @baseballot: Excited to present FiveThirtyEight's Ultimate Guide To Election Night™—including when to expect results in all 50 states! h… — PolitiTweet.org
Nate Silver @NateSilver538
We've now updated our presidential turnout estimates to account for early voting data. Here's how it works (it's pretty straightforward): in states where a high percentage of people have already voted relative to our model's initial estimate, we revise the estimate upward. — PolitiTweet.org
Nate Silver @NateSilver538
See, this is why you can't infer much about how early voting from party alone. In Florida, Ds currently have only a 2-point edge (D 40/R 38/I 22) in early + absentee voting. But if the partisan splits are as below, it would translate to Biden being ahead 56-39 with those voters. — PolitiTweet.org
Robert Griffin @rp_griffin
Early vote statistics have been plentiful this cycle. As always, there's some uncertainty in interpretation. Some t… https://t.co/QmKlRKw0LY
Nate Silver @NateSilver538
I say "obvious-ish" because if it weren't for 2016, we'd probably be spending a lot more time reminding people that (i) the tipping point state is quite a bit closer than national polls and (ii) systematic polling errors can occur. But everybody knows that stuff because of 2016. — PolitiTweet.org
Nate Silver @NateSilver538
In certain ways, I don't think the presidential election models are really telling you very much this year. Biden's way ahead in the polls; that should be obvious. He's also not so far ahead that he can be completely comfortable, however, which should also be obvious-ish. — PolitiTweet.org
Nate Silver @NateSilver538
RT @ezraklein: Learned a lot from this conversation with @NateSilver538. One thing I admire about the way he does data journalism: the mode… — PolitiTweet.org
