Sign in with Twitter

Username:

Editor-in-Chief, @FiveThirtyEight. Author, The Signal and the Noise (https://t.co/EYTxvN6BLY). Sports/politics/food geek.

1,320 Following   3,351,660 Followers   31,070 Tweets

Joined Twitter 8/27/08


Deaths are up versus last Sunday, which is a little disappointing after several days of plateaus/declines. With tha… https://t.co/2UUJDS4TKHUS daily numbers via @COVID19Tracking: Newly reported deaths Today: 616 Yesterday: 1,088 One week ago (8/2): 498… https://t.co/0Kganv2YWQEvery election year is "fun" in its own way. This year's fun will come in the form of the general public likely thi… https://t.co/tRw62RLHONBut the empirically-driven intuitions that (1) elections tend to be close in a highly polarized era and (2) Trump w… https://t.co/8Ejv1GBvICOf course, you can use a prior other than zero based on the economy or incumbency or what not if you want. Given re… https://t.co/uvycyuGz7zSomething else to keep in mind is that presidential elections tend to tighten down the stretch run. So an 8-point B… https://t.co/zHf7JLgUgm“I’m really not making predictions here, but I suppose I buy the conventional wisdom that Harris is the favorite,”… https://t.co/weSw4gHbMz
Retweeted by Nate Silver @gelliottmorris I've been giving you good advice that I don't give to a lot of people. It's your call on whether to… https://t.co/aQx0jbxRrn @gelliottmorris There is no "validation set" other than reality. You can't determine predictive accuracy until a mo… https://t.co/94y5f6yhXX @gelliottmorris Again, I don't think you're getting my point. Given "reasonable" choices of parameters, backtesting… https://t.co/Pmsmk0Mv6K @gelliottmorris This is such a fundamental point that it depresses me a bit about the status of "data science" education tbh. @gelliottmorris You're really not getting it. You didn't publish a model in 2016. It's neither interesting nor rele… https://t.co/bVPTx8ZOO1 @gelliottmorris To put it another way, the easy part of forecasting is fitting a model and the hard part is knowing… https://t.co/Kg1PFluopa @gelliottmorris Backtesting does help to an extent but then you encounter negative returns if you overfit. But do… https://t.co/FX0RbSF3Gr @gelliottmorris Fundamentally, the failure to understand that *fitting a model* is not the same thing as *out-of-sa… https://t.co/k9xdUEWwzq @gelliottmorris It wasn't really about you or anyone in particular but if you're talking about a historical record… https://t.co/BR6v0HVC2uIt is generally better to think more carefully about how to pragmatically account for real-world uncertainty in a f… https://t.co/RjjEzNARdn
8/9
2020
It's good to see deaths and cases decline vs. last Saturday. With that said, a number of states are having reportin… https://t.co/7HWv2NZXw9US daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,088 Yesterday: 1,333 One week ago (8/1): 1,1… https://t.co/31K5k0tz1eThe thing about the VPstakes is there are a lot of actions that are consistent either with creating some subterfuge… https://t.co/keJFDK9abiIn other words, if Harris allies would not be very happy with the choice of Rice or Bass, and vice versa ... well t… https://t.co/jox1n2fAKRIn the abstract, Whitmer would be a logical enough choice as Biden's VP. What I wonder about is the extent to which… https://t.co/0nyBj3wiw5
8/8
2020
Same story as in recent days. Deaths still high, though the 7-day average death count is holding roughly flat. Case… https://t.co/lX17IvUlPnUS daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,333 Yesterday: 1,251 One week ago (7/31): 1,… https://t.co/90xXEDxoJDDuring that time, Trump's approval for his handling of COVID has worsened. But, the economy has gotten a bit better… https://t.co/dir28mIaE9It may be worth recalling that Biden was leading Trump by "only" 6 points when the George Floyd protests began. So,… https://t.co/Vf8BdTL6BEThis is interesting. There's a fairly high correlation between pollster quality (as judged by our pollster ratings)… https://t.co/oUYdTEEDzh🥳Job Alert🥳 I'm looking for a part-time video and motion graphics producer to join the @FiveThirtyEight team! For… https://t.co/yczygS2XIE
Retweeted by Nate SilverMidtown Manhattan feels a bit like (pre-pandemic) downtown Houston or Dallas or something in terms of the overall l… https://t.co/UIRITFqtyT
8/7
2020
One source of uncertainty is that quite a few states are having reporting issues of one kind or another right now,… https://t.co/IWULt4dbnRWe are starting to see the death numbers plateau, which would make sense given that cases began to plateau a couple… https://t.co/KuW4xjHckMUS daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,251 Yesterday: 1,403 One week ago (7/30): 1,… https://t.co/sNEcsykEP7There's a fairly big gap once you get beyond Georgia/Texas/Iowa/Ohio to the next set of states that Biden could win… https://t.co/anOf42nETKHot take: Whether or not the predictions are important unto itself, working on predictive modeling is one of the be… https://t.co/XpG1IBTCNR @BrendanNyhan @mattyglesias @cwarshaw My next contention: someone reading the peer-reviewed literature on this topi… https://t.co/7nylbwMMWm @BrendanNyhan @mattyglesias @cwarshaw Let's get specific then. What percentage of voting behavior do you think can… https://t.co/tlOETE2R0o @BrendanNyhan @mattyglesias This is so weird Brendan. It's a strawman to call my position a strawman! I'm saying th… https://t.co/o2YYhXr8PB @BrendanNyhan @mattyglesias I think you're missing my point here, which is that a lot of political science theories… https://t.co/qFf5tWdLje @mattyglesias If polling-based models considerably outperform fundamentals-based models in true out-of-sample predi… https://t.co/IDgNjLDPfZ @mattyglesias Yes, I disagree that there's a difference between "modeling designed to improve forecast accuracy and… https://t.co/BWrQooPgshEhh. There's no get-out-of-jail-free pass for designing a p-hacked/overfit model because you say you're not using i… https://t.co/jDV9ZHs3qt @jonathanchait Well, that's why it's a subtle shift! But generally speaking there's a lack of articulation of exit… https://t.co/4Jtyy3XFneIt's a subtle shift, but at some point in the US discourse, closing/locking things down transformed from something… https://t.co/xF0U7SjYra
8/6
2020
Our national polling average has tightened a bit to Biden +7.6. We haven't had that many super high quality polls… https://t.co/eNfNdfBVRoWednesdays seem to be producing high death counts, on the other hand. A very bad number today (1,401) after another… https://t.co/qbV9JfEQYlA decline in cases vs. last week. But this is the fewest # of tests reported in several weeks, which likely reflect… https://t.co/GU4SEdX2r2US daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,401 Yesterday: 1,176 One week ago (7/29): 1,… https://t.co/x6qoY2xtNIFalse negatives and false positives may not be all that common as a share of the general population, but once you g… https://t.co/y6bu0ywf24I've seen a few too many mainstream media stories of "unusual" COVID cases where the most likely explanation is a f… https://t.co/AX1JZ9fEra @CathyYoung63 Yeah, I should have been clearer about that. Since the start of the pandemic.How often are you seeing (in person) friends and family who don't live with you? (Please answer only if you're in the United States.)I do think there could be pretty decent TV viewership. People will be sitting at home without a lot of alternatives… https://t.co/Xlk14Ncylb…both parties have now scaled back their convention plans to the point where I don't think we can pretend its busin… https://t.co/Ci0rUr3LULThis year is forcing a lot of ad-hoc decisions for election modelers. One of them concerns the party conventions—ou… https://t.co/6na61nNxuGHe Incorrectly Predicted That Trump Would Win The Popular Vote In 2016. But, Hey, Who Cares About Details Like That? https://t.co/FkWyWFcfAc @Redistrict If only the opponents were Paul Kroger (D) and Mary Shoprite (I) then this would basically represent my… https://t.co/FoqAgu4YFQ
8/5
2020
A somewhat mixed bag. Fewer cases than last Tuesday, but also fewer tests, with some states' testing impacted by th… https://t.co/J95VZrmiixUS daily numbers via @COVID19Tracking: Newly reported deaths Today: 1,176 Yesterday: 519 One week ago (7/28): 1,1… https://t.co/75pC2TxK3w @jonmladd @BrendanNyhan Also, from a comparativist standpoint, many world leaders have become *more* popular despit… https://t.co/Yp77ZIuenT @jonmladd @BrendanNyhan And if you plug in the before-and-after economic data into most of these models, they predi… https://t.co/t9iLYDheSC @jonmladd @BrendanNyhan Yes! I think that's absolutely an incorrect story and one not well supported by the evidenc… https://t.co/EggZBr6b0g @jonmladd @BrendanNyhan I'm saying that, in addition to doing a poor job of out-of-sample prediction, the fundament… https://t.co/NQ6ijBZIiY @BrendanNyhan @jonmladd Since you think the Abramowitz model is strong, happy to agree to make it the standard by w… https://t.co/eGE6Pij63C @BrendanNyhan @jonmladd I really wish you'd responded in more detail to that article that I wrote 8(!) years ago. I… https://t.co/F1TUQTafFg @BrendanNyhan @jonmladd The fundamentals models have performed **terribly** out of sample and they're also a **comp… https://t.co/pGa57RzpijThe reason we generally prefer polls>economy is because different modeling choices on the economy can yield radical… https://t.co/oJsnOuyjjTSo while we spend a lot of time thinking thru different model specifications, it's mostly *not* about trying to max… https://t.co/gDT5jwjl4ELike all folks modeling presidential elections, we're forced to use small samples (e.g. n=12) to infer what the bro… https://t.co/KN7tTxQeIh @CT_Bergstrom How does this interact with the supposed heterogeneity in rates of transmission? It seems like "some… https://t.co/npsy2Nd9r0One could have a high degree of confidence that Biden will win based on their read of news events (i.e. that COVID… https://t.co/g3n9B9W5RgSure. But we're also in the midst of a global pandemic which is ~unprecedented in recent times, extreme economic vo… https://t.co/7agpwnVoVHOne thing people may not realize, too, is that campaign time is bit nonlinear. Polls don't really get much more pre… https://t.co/Y4MnlhB3pjIt's still early, folks (although several of these were impacted by convention bounces). https://t.co/4JLFBLLjMKAllow me to translate: someone told him that he's scaring his elderly supporters in Florida away from vote-by-mail… https://t.co/fg96Z8I129
Retweeted by Nate SilverAnyway, I talk more about this on our podcast, where I rail a bit against the claim that elections heavily boil dow… https://t.co/qiGYso6zaIRather, it's COVID that's Trump's big problem, plus other stuff like his handling of the race relations, overall pe… https://t.co/OgBTmAC8R5Of course, if negotiations for the next round of stimulus payments fall apart, that could present big risks for Tru… https://t.co/mS3VORHrjQIf it seems crazy to treat current economic conditions as tantamount to merely a mild-to-medium-sized recession, ta… https://t.co/JDoQtR2OXaInstead, with the government having put so much money in people's pockets, the index looks more like a league-avera… https://t.co/aBrFnZbwL8The economic index we use in our model has 6 components: —2 of them (inflation, stocks) have been basically "fine"… https://t.co/dHCjvaPdrHI don't think it should be underestimated how much the CARES Act and the fairly generous stimulus it provided has b… https://t.co/fkC5KAzzqMNo one bad take is dispositive—I certainly have had plenty of bad takes myself—but "Kanye West will have a huge imp… https://t.co/PuEZQLHjE0
8/4
2020
It does seem like when we're talking in 5 years about how we got out of this, improvements in indoor ventilation sy… https://t.co/iL28GZbbODAlso: Looks like Florida's numbers will be a bit depressed for a few days here because of the hurricane. https://t.co/eENuVIG36a🎧 NEW POD 🎧 -- It's *not just* the economy, stupid (when it comes to forecasting elections) -- The potential cons… https://t.co/84CwmtfTB0
Retweeted by Nate Silver...and we've now gone 9 days in a row with a week-over-week decline in the number of new cases. There is still a *… https://t.co/lElEAY2V84* Technically, there were 1,059 deaths reported last Monday, but that reflected a one-off adjustment by Texas; it w… https://t.co/aNARP7r4HxUS daily numbers via @COVID19Tracking: Newly reported deaths Today: 519 Yesterday: 515 One week ago (7/27): 428*… https://t.co/2y645Lrm0LThere's a variation that's both more encouraging and more accurate IMO that gives people a sense of long-term optim… https://t.co/H4YqRdicy3I'm seeing a lot of "don't assume a vaccine is a magic bullet!" stories lately, which seems like a pretty rational… https://t.co/JRNq8lZCkI @CoreyT_WX Ahh yeah, that was it. Thank you!Did... someone post on here the other day numbers showing what % of voters intended to vote by mail in each state?… https://t.co/4kUnPiIP3HA quick announcement: In addition to covering the NBA for GQ, I'll be contributing a little bit to @FiveThirtyEighthttps://t.co/IqT6my1Vxr
Retweeted by Nate SilverTrump has been talking a lot about crime this summer. That could tap into Americans' inaccurate belief that crime i… https://t.co/AIlY1H93uY
Retweeted by Nate Silver @DanRosenheck No, not clear. We have this as not being as bad as 2008. Why? * CARES act means real disposable inco… https://t.co/kZZNWhFsXa*** I'm way late on this today, so please see here for additional details and caveats: most importantly, that Texas… https://t.co/rzNluUeQMwUS daily numbers via @COVID19Tracking***: Newly reported deaths Today: 515 Yesterday: 1,172 One week ago (7/26): 5… https://t.co/Q3MrTJL225
8/3
2020
How the Media Could Get the Election Story Wrong https://t.co/kDEzgI0ejl
Retweeted by Nate SilverNew: The weekly @TheEconomist/@YouGovAmerica surveys ask about voter enthusiasm in three ways, which allows for a r… https://t.co/MmXlurOJry
Retweeted by Nate SilverNotably, these maps reflect cases where Biden could win on the basis of Florida + North Carolina (+ Arizona in one… https://t.co/wZerkBBjcU
8/2
2020

0