Nate SilverVerified account

@NateSilver538

Editor-in-Chief, . Author, The Signal and the Noise (). Sports/politics/food geek.

New York
fivethirtyeight.com
Joined August 2008
1,610 Photos and videos Photos and videos

Tweets

You blocked @NateSilver538

Are you sure you want to view these Tweets? Viewing Tweets won't unblock @NateSilver538

  1. 6 hours ago

    You're making this way too complicated. My critique is that you're offering false precision. There's just not that much we can say about the Electoral College right now beyond a couple of fairly loose priors (e.g. its more likely to help than hurt Trump).

    --the data covers many modes: online, live interview RBS+result, online+live interview RBS, in-person, and the hard results. --there is extensive analysis, including implied by your own pre-election analysis, that 2018 electorate ~ = RVs/2020
    Show this thread
    58 replies 34 retweets 305 likes
    Undo
  2. 6 hours ago

    There's also a fairly large volume of polling that shows Biden (and to some extent Bernie) performing quite well vs. Trump in head-to-head matchups in the Midwest. I wouldn't put a ton of stock in that data, either. But it's not just one poll in Ohio.

    13 replies 18 retweets 127 likes
    Show this thread
    Show this thread
    Undo
  3. 6 hours ago

    All of which is… fine… to get a rough approximation. But the real-world margin of error on those estimates is (at least) several points wide, which is much too large to come to many firm conclusions about the tipping point, especially if you're not working probabilistically.

    2 replies 5 retweets 122 likes
    Show this thread
    Show this thread
    Undo
  4. 6 hours ago

    Your hypothesis relies on data that's (i) 8 months old (ii) was collected mostly online (fine but not gold-standard) (iii) involves a lot of statistical extrapolation (iv) is calibrated to 2018 rather than 2020 turnout (v) is based on approval rather than head-to-head matchups.

    My hypothesis is that the president's approval rating in Ohio is probably better than this one poll in Quinnipiac, based on election results and high-quality, large, calibrated election surveys that are easily forgotten but deserve to be remembered
    27 replies 26 retweets 181 likes
    Show this thread
    Show this thread
    Undo
  5. 9 hours ago

    It just seems like there's a lot of ambiguity around your hypothesis, which of course there there is because it's 16 months until the election, and we don't know who the D nominee will be, or what turnout will look like, and you're determined to cherry-pick your way around it.

    Morning Consult is Trump's worst pollster right now, with a 40% approval among RVs. They also don't have state-by-state samples, it is an MRP model--something you've critiqued. And Civiqs, with a similar method, is more reasonable
    Show this thread
    52 replies 21 retweets 219 likes
    Undo
  6. 9 hours ago

    Ehh, Morning Consult, which has much larger sample sizes, also shows Trump underperforming, or at least performing mediocrely, in the Midwestern swing states, including being -4 in Ohio and -8 in Iowa.

    The exits put Trump approval in Ohio at 53/46. The GOP won the US House vote and the governorship. And polls generally underestimated the GOP there in '16 and '18. So it is hard for me to just take this poll at face value-Trump at 43/52--and run with it
    74 replies 69 retweets 336 likes
    Undo
  7. 9 hours ago

    I don't think you should engage in overly precise Electoral College analysis at this point (WAY too early) but if you ARE going to do it (don't), seems important to note some candidates (Biden, sometimes Bernie) are polling way better than other D's in white working-class states.

    55 replies 72 retweets 319 likes
    Undo
  8. 9 hours ago

    Harris's debate bounce and Biden's debate crash have reverted to the mean a bit, which is pretty much what you should have expected because things like this usually do.

    80 replies 133 retweets 572 likes
    Undo
  9. 19 hours ago

    Why not just report the news and not reach for a weird like 5th-order conditional/counterfactual narrative

    While Americans might not read the book, the argument went, they would watch the movie. If so, the movie Americans tuned into was not the blockbuster Democrats had sought nor was Mueller the action star they had cast.
    196 replies 960 retweets 7,474 likes
    Undo
  10. Jul 24

    Random tidbit from the NBA metrics we're working on: We think most of the existing catch-all metrics are biased toward overrating bigs (PFs and Cs), because they don't account for the fact that most bigs rely heavily on assisted field goals instead of creating their own shots.

    23 replies 12 retweets 219 likes
    Undo
  11. Jul 24

    Also to folks who are like "WHY IS THE MEDIA FOCUSING ON OPTICS INSTEAD OF SUBSTANCE?!?!?" the whole reason the Democrats held this hearing was because they thought it would make for good optics; Mueller made clear ahead of time that the substance was all in the report itself.

    166 replies 210 retweets 1,444 likes
    Show this thread
    Show this thread
    Undo
  12. Jul 24

    Mueller signaled very explicitly ahead of time that he didn't want to testify and that if he did, he'd just stick quite narrowly to the report. So if Dems think the hearing went poorly, they should be blaming leadership and others who called for the hearing, not Mueller.

    Much as I hate to say it, this morning’s hearing was a disaster. Far from breathing life into his damning report, the tired Robert Mueller sucked the life out of it. The effort to save democracy and the rule of law from this lawless president has been set back, not advanced.
    174 replies 497 retweets 2,614 likes
    Show this thread
    Show this thread
    Undo
  13. Retweeted
    Jul 24
    Replying to

    makes sense in that “Franken was railroaded” is basically dc cocktail party cannon.

    12 replies 12 retweets 143 likes
    Undo
  14. Jul 24

    Same

    From colleague John Roberts. He reports that Trump is watching the Mueller hearing but “also working on other things”
    34 replies 88 retweets 1,478 likes
    Undo
  15. Jul 24

    M U E LIVEBLOG L E R:

    3 replies 26 retweets 139 likes
    Undo
  16. Retweeted
    Jul 24

    We’re live blogging Robert Mueller’s testimony this morning. Follow along!

    3 replies 21 retweets 49 likes
    Undo
  17. Jul 24

    What's so gross about the Franken story is how much it relied on character testimony from Franken's rich and powerful friends while downplaying the fact that 8 (!) women had accused him of misconduct.

    Let's not forget: "Franken’s resignation came after many women had accused him of misconduct" -- not just one.
    107 replies 479 retweets 2,357 likes
    Undo
  18. Jul 24

    "Would you say your name for the record, sir?" "Robert Mueller" "And would you SPELL your name, sir?" "M-U-E-L-L-E-R". "How do you expect us to believe anything you say, Mr. Mueller, if you pronounce your name M-U-L-L-E-R, but spell your name M-U-E! [dramatic pause] L-L-E-R?"

    56 replies 195 retweets 2,032 likes
    Undo
  19. Jul 23

    The public option (90%) is also way more popular than Sanders M4A (64%) *among Democrats* in this poll.

    77 replies 188 retweets 889 likes
    Show this thread
    Show this thread
    Undo
  20. Jul 23

    So a supercharged public option (e.g. Biden) is way more popular than a Sanders-style Medicare-for-all plan.

    NPR/PBS/Marist poll: • 41% favor a Medicare for all plan that replaces private insurance • 70% support “Medicare for all that want it,” which keeps private insurers
    Show this thread
    203 replies 397 retweets 1,588 likes
    Show this thread
    Show this thread
    Undo

@NateSilver538 hasn't Tweeted yet.

Loading seems to be taking a while.

Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

    You may also like

    ·