Editor-in-Chief, @FiveThirtyEight. Author, The Signal and the Noise (amzn.to/QdyFYV). Sports/politics/food geek.
Is this like a $200 clue in Jeopardy where the answer is obvious but you have to phrase it in the form of a question? https://twitter.com/SNYtv/status/1140078115777105920
re: Rasmussen: worth noting that founder Scott Rasmussen was ousted from Rasmussen Reports several years ago (in 2013). While their polls have always had a GOP lean, Scott was a good guy who cared about polling accuracy. The guy who took over (Ted Carroll) is…a…different story.
RT @Herring_NBA: Latest for @FiveThirtyEight: The Lakers Land Anthony Davis. Ok; Now What? https://fivethirtyeight.com/features/the-lakers-land-anthony-davis-ok-now-what/
One interesting fact is that Rasmussen had Republicans winning the popular vote for the US House last year and Democrats actually won it by **9** points, which makes for about twice as bad a fuck-up as any pollster had in 2016. https://twitter.com/Rasmussen_Poll/status/1139644148955439105
A lot to give up but if the Lakers can sign a 3rd guy (haven't done the math but it looks like they have maybe $30m-ish available?) a team of LeBron + AD + FA TBD + Kuzma + 10 replacement level dudes might be the best team in a league where everything's in transition/turmoil. https://twitter.com/wojespn/status/1140023139142971392
In other words, some voters who like both candidates (or even mildly prefer Warren) might stick with Bernie so long as he seems to have the clearer path to the nomination. But the moment she appears equally/more viable, they could jump to her instead.
While Sanders supporters and Warren supporters are far from perfectly overlapping, if I were Bernie's campaign I'd nonetheless be concerned that if Warren passes me in the polls at some point, it could trigger a bandwagon effect that would be hard to come back from.
RT @laurabronner: How unlikely was this large a discrepancy between the two debates, given the DNC's stratified randomization methodology?…
I think I'd rather have Biden in my group if I were Warren or pretty much any of the other frontrunners. https://twitter.com/mattyglesias/status/1139571899267276800
Yeah I guess it’s not actually a tournament.
Yeah. Not sure how much there is to gain by winning the JV tourney
I can't articulate why this tweet is so funny to me https://twitter.com/EricRWeinstein/status/1139036884154310656
Is Taleb part of the Intellectual Dark Web?
I think he might be IDW adjacent
RT @NateSilver538: Just watched a Canadian team win a championship in an American sport while drinking beer from a Dutch multinational cong…
RT @FiveThirtyEight: No, the Warriors weren’t full-strength for these NBA Finals. But if you think Toronto became a champion solely because…
Empirically, the margin of error on general election polls conducted this far out is around... 25 points! https://twitter.com/micahcohen/status/1139492207373799429
RT @micahcohen: All this talk about polls testing hypothetical general election matchups -- Trump vs. Biden/Sanders/Warren/etc -- is making…
Just watched a Canadian team win a championship in an American sport while drinking beer from a Dutch multinational conglomerate at an Irish-themed bar in Brazil while wearing a cap with a Japanese baseball team's logo. Then took an Uber home. Have achieved peak @neoliberal-ism.
Don't asterisk this series. The Raptors are a great team and deserving champion and not really any telling who would be better at full strength. https://fivethirtyeight.com/live-blog/warriors-raptors-game-6/?lbup=14818#14818
Hey folks we're having a lot of fun liveblogging Game 6 if you want to join us: https://fivethirtyeight.com/live-blog/warriors-raptors-game-6/
Andrew, please don't become an #unskewer. Virtually all polls now either include cellphones or are conducted online. https://twitter.com/AndrewYang/status/1139002576119185409
and 2) The injury came early in the game on a routine play and 3) They were "all-in", with KD playing lots of minutes at a high pace vs. a very good opponent after barely having practiced. So yeah this really should prompt questions about the Warriors' decision-making process.
Why? Because 1) There's not a lot of data on rushing guys back from injuries (especially if you want data on particular types of injuries, e.g. calf injuries) since NBA teams are usually pretty careful, so the Warriors' priors can't have been that strong.
Listening to the @ZachLowe_NBA @HowardBeck podcast convo about what Durant's risk of reinjury was and whether the Warriors just got unlucky. My 2c: While people are often too results-oriented, I think this probably *is* a case where the result suggests the Warriors screwed up.
The key when dealing with chronic bullshitters like @davidsirota is to recognize that they're arguing in bad faith and do quite a bit of trolling back, in addition to calling out their BS. Dudes can't figure out how to handle and wilt under the pressure every single time.
LOL .@davidsirota can't take the heat when you call out his bullshit.
Democrats are underdogs to win the Senate but *conditional on winning the presidency* they're probably slight favorites to win the Senate. So the modal outcome in a D presidency is probably like a 50-50 Senate nominally controlled by D's with Joe Manchin often voting with R's. https://twitter.com/dylanmatt/status/1139167172075429889
There's a decent case that Warren has emerged as a clear No. 2. Also a case (which I mostly don't buy) that Harris has fallen off the lead lap. Her favorables are good but she isnt a lot of people's first choice. But let's wait and see what things look like after the debate.
Overall, the gap between Biden and the rest of the top tier has closed somewhat. This is a bit predictable in national polls (i.e. Biden's bounce has faded somewhat just like everyone else's bounce did) but IA/NH could be challenging for him even if SC/NV potentially pretty good.
A lot of early-state polls starting to look good for Warren & sometimes also Buttigieg. https://fivethirtyeight.com/features/silver-bulletpoints-iowans-seem-to-like-warren-and-buttigieg/
Or you could set up some more complicated formula where experience was weighted alongside other stuff, but then people would complain about how opaque the formula was.
Automatically giving a slot to statewide office holders seems like an OK idea in principle but then people would complain about how Gravel got into the debate. https://twitter.com/mattbai/status/1139133491063468036
RT @FiveThirtyEight: The St. Louis Blues really, actually won the #StanleyCup. https://fivethirtyeight.com/features/are-the-st-louis-blues-really-actually-going-to-win-this-thing/
SandersBrothers should therefore embrace the sort of probabilistic thinking espoused by 538, which aptly assesses that he's among the stronger candidates but not the favorite, and that everyone but **maybe** Biden is inherently something of a longshot with 23 runners in the race.
To be fair it's tricky because if you assume that primaries are 100% predictable based on polls, then Biden wins, but if you assume that primaries are totally unpredictable, then Bernie has like a 1/23 chance but it's the same as Bill de Blasio or whatever.
this is very good point, multi-candidate races (plus the key role of perceptions of electability) means things genuinely can flip all over the place very quickly.
SandersBrothers can't seem to decide if they should tout how good their polling is or claim that polls are unreliable and should be ignored. https://twitter.com/joshtpm/status/1138986032819507200
RT @geoffreyvs: 20 Democrats have qualified for the first debates on June 26-27. Barring any late polls tonight or donor surprises tomorrow…
p.s. Although I think his campaign is engaged in some bad-faith arguments about polling, I still think Bernie is probably the ~3rd most likely person to win the D nomination. (It's pretty close between No. 2 and 5, IMO.)
I'd guess that 80-90% of the time that a campaign gets into a public argument about polling, it isn't doing itself any favors.
It's sorta funny @davidsirota that you respond to a critique of mine about how you're cherry-picking polls to cherry-pick one result out of the thousands of forecasts we've made. https://twitter.com/davidsirota/status/1138868586779435008
Here's a list of literally every election (and sports) forecast we've ever made, they're pretty good in the aggregate although some are wrong of course! https://projects.fivethirtyeight.com/checking-our-work/
Also @davidsirota it's sorta funny that you think this makes me mad. Literally nothing* makes me happier than calling out flagrantly dumb bullshit like you sometimes poop out; it's why I got into this business. * In a work/professional context, lots of things are better IRL.
Think we will leave it right here @NateSilver538 #micdrop
And the pessimistic-but-reasonable case is that he's probably still in second place nationally, but his overall trajectory is downward, his early-state polls aren't looking so hot, and that Warren and Biden present a fairly difficult two-front problem.
If you're looking at the polls in any sort of comprehensive way, the optimistic-but-reasonable case for Sanders is that he's rebounded by a point or two off his post-Biden announcement lows, and closed his gap with Biden, although he's still a ways off his peak from the spring.
RT @NateSilver538: Like seriously, Sirota's characterization of the polls is so obviously wrong that it almost verges on gaslighting. https…
RT @NateSilver538: This is complete BS. Sanders was polling at 19-20% in the RCP average at the end of April, and he's at 17.3% now. Rather…
Because I've spent literally 11 years telling people to look at the polling average instead cherry-picking data to fit their preferred narrative? https://twitter.com/davidsirota/status/1138861522913042437
I’m pretty sure it’s your corrupt ties to the corporate oligarchy.
I’m sorry to @NateSilver538 and others who are so upset by these Q Poll numbers since Biden got in the race in April. It ain’t “gaslighting” — it’s just the actual data. Why does it make you so mad? Sanders: +8 Warren: +3 Biden: -8 Buttigieg: -2 https://poll.qu.edu/national/release-detail?ReleaseID=2627
Between these two candidates, who (1) would you prefer to see win the Democratic nomination (2) do you think is more likely to win the nomination?
RT @Herring_NBA: I could tell you why bats get into the Spurs arena so frequently, and why the Raptors seem to get so many friendly rolls o…
It's probably stuff like this that's pushing Sanders's campaign into unskewed polls territory. They have real problems with Biden on the one hand and Warren on the other hand. https://twitter.com/SteveKornacki/status/1138853592369192967?s=19
Do you not agree that a single quality poll is always more indicative than a composite of the trend in all polls over time?
Like seriously, Sirota's characterization of the polls is so obviously wrong that it almost verges on gaslighting.
This is complete BS. Sanders was polling at 19-20% in the RCP average at the end of April, and he's at 17.3% now. Rather than surging, he's lost a couple points. https://www.realclearpolitics.com/epolls/2020/president/us/2020_democratic_presidential_nomination-6730.html https://twitter.com/davidsirota/status/1138834760208474113
Overall, I remain taken by surprise at how taken by surprise the Sanders campaign seems to be that Biden is polling in front of them. A lot will change between now and Iowa and polls could be off even then, but they seem to have a lot of wishful thinking in lieu of a strategy.
For instance, Sanders is doing worse with voters who are paying a lot of attention to the campaign, which might be a sign that he'll *lose* ground with the actual electorate (or in likely voter polls) as compared with the registered voter polls you see now.
This adds error/uncertainty, and primary polling is generally a rough enterprise, but the polls are probably about as likely to be overestimating Sanders as underestimating him.
Younger voters are harder to reach, but pollsters attempt to compensate for that by upweighting the younger voters they do reach to match their projected composition of the electorate, as @fshakir surely knows.
Pretty irresponsible by Sanders campaign manager @fshakir to endorse the theory that polls are "oversampling" older voters. Short THREAD follows: https://twitter.com/StatusCoup/status/1138508125127024642
Seems right in time for the unskewing part of the primary season.
I’m not sure what story @NateSilver538 and everyone are arguing about, but I have a wise and judicious take on the controversy.
If I worked for a campaign, I’d be sorely tempted to charge large sums of money to produce “internal polling” that just agrees with public poll averages. Seems way more profitable and probably pretty accurate.
RT @NateSilver538: We've re-launched our generic ballot tracker. Democrats currently lead by ~6 points, down from 8-9 points as of the 2018…
© 2009 creamsocial