Our Best Stuff From Another Week of Fact-Check Whack-a-Mole

Plus vaccine delays, endless wars, and the greatest hits from our What's Next event.

By
132

One of my favorite things about doing this newsletter is hearing from readers in the comments each week. I can tell some of you have gotten to know each other (virtually), and it is a treat to see people sharing book recommendations and trading recipes and inquiring about one another’s well-being. Please, keep it up as we enter what is sure to be a long, dark, lonely winter. 

But you aren’t the only readers we hear from. And the people who reach out via our fact-checking email address have been on my mind a lot this week, as conspiracy theories related to the election keep cropping up like the world’s most annoying game of whack-a-mole. It’s heartening when someone emails and says something to the effect of “My [husband/neighbor/cousin/friend/aunt] keeps sharing this claim about [Dominion Voting Systems/Michigan vote totals/vote manipulation] and I can’t change their mind. Can you report on this?”

First off, it gives me the warm fuzzies when someone writes with a message like that, because it means people see the The Dispatch Fact Check as a trusted source. (It’s less heartening when people write in and call us names, but I digress.) And it’s great when a reader sends us a question and it sends our fact-checkers down a rabbit hole or two. (I hope I’m not giving away state secrets when I tell you that we did this fact check partly in response to a reader query.)

I’ll be honest, though. We have two great writers doing fact checks for us—Alec Dent and Khaya Himmelman—but they are on the receiving end of a firehose of misinformation. We can’t shoot down every crazy claim that’s out there. At least not until we figure out how to clone the two of them.

Keep reading with a free account
Create a free Dispatch account to keep reading Get Started ALREADY HAVE AN ACCOUNT? SIGN IN
132
By
Comments (132)
Join The Dispatch to participate in the comments.
 
  1. BIgMack

    Rachael, please keep this just between us! It's very confidential. But while I was in GA last month, I was visiting a friend very high up in the CDC, and this person gave me access to 3 syringes of antidotes to Covid, one was specially marked with a Microsoft logo that was to be administered to an acquaintance, I had recently met in Maine, my cousins live outside of Waterville. (they said they from the true Korea) anyway, I gave them the injection and as it so happens I am able to track their whereabouts just like you would track a package! Imagine that! (Right now they have been in Cabo for a week). I imagine they will be heading back to their office, for a new assignment soon, but what do I know of international politics...
    Thank you for your articles, I enjoy everyone of them!

    Collapse
  2. Avatar photo
    Rebecca A Mohun

    Thanks, I just joined the Dispatch and this is all very helpful. So for requests, this one has been making the rounds on my FB feed: https://noqreport.com/2020/12/03/bombshell-video-evidence-of-clear-voter-fraud-presented-in-georgia-for-the-first-time/?fbclid=IwAR0wmx6sJfM1FFQ1aCfM2QcyARc-22sIiaN8HzfSAYUlUQh4FYMTL8Wc9CU . I found a short article addressing it but I'd love to find something more extensive.

    Collapse
  3. JohnO

    I find it difficult to follow even the potentially plausible claims of Trump's people v. court statements and analysis. I realize the headline is 'have lost virtually very case'. But it doesn't seem anybody yet has a compiled an organized account of the claims actually made in court with each court's reason for rejecting them. I'm not talking about the broader array of generally wilder things alleged in press conferences, tweets and generally floating around the internet which the Trump lawyers have not be willing to claim in court.

    I accept (as Trump skeptic if not quite 'Never Trump' conservative) that if you or representatives say obviously crazy stuff or you don't complain loudly and constantly enough when people who aren't your representatives say those things, the rational part of the public will reject all your claims as nonsense. Which is in that case your own fault. But there could actually be valid kernels in the manure here which the body politic should be considering for the future. They might be found in cases where court rejections based on 'you're asking for a disproportionate remedy even if you proved this' besides 'and you *haven't* proved this' are the correct decisions as far as this election, so 'move on' yes, but 'nothing to see here' not necessarily entirely.

    Collapse
  4. Avatar photo
    BikerChick

    “It’s likely that we’ll one day look back at the speed with which multiple companies brought COVID-19 vaccines to market as some kind of miracle. It’s a process that normally takes years, and as we’ve pointed out previously in The Dispatch.” How can it be miraculous when it’s due to the suspension of federal regulations that have allowed this “miracle” to happen? There is a reason we have the regulations...to ensure safety. That’s precisely why 40% of healthcare workers are hesitant to get the vaccine. Please face check Dr. Michael Yeadon’s concerns, he’s filed a petition to halt approval because of his safety concerns (former Pfizer research scientist.)

    Collapse
    1. Avatar photo
      BikerChick

      https://www.npr.org/sections/health-shots/2020/12/01/940158684/some-health-care-workers-are-wary-of-getting-covid-19-vaccines

      Collapse
  5. Avatar photo
    GREG BROWN

    Call this what-aboutism (I call it, "reverse what-aboutism"): From the instant Trump won there began theatrical displays of outrage that were unheard of prior. Half the country suddenly felt threatened by the other half and lost their goddamn minds. The fury, the chubby girl wailing, the pussy hats, the despicable slander of Brett Kavanaugh, despicable outrageous lies about Russian "collusion" on and on. The half of the country that voted for Trump, however reluctantly, saw clearly that the half in opposition would stop at nothing, were openly indifferent to the most crucial ties that bind us to one another. Is it really any surprise that Trump supporters now doubt the election outcome?

    Collapse
    1. Avatar photo
      Aylene Wright

      One argument by Trump supporters that I do sympathize with is, "if all these Trump haters REALLY believe that Trump is a fascist tyrant, who is the Second Coming of Hitler Himself, and that a 2nd Trump term means that the US turns into Russia with Trump as President-for-life, then why *wouldn't* they try to steal the election?"

      However, I am also sure that some Trump supporters are so terrified that the "Leftists" & "Antifa" and "socialists" will be running the show behind the scenes using Biden as a puppet, or that Biden is going to be conveniently offed after a few weeks to pave the way for President Harris, that, even if the Democrats did NOT engage in mass fraud to steal the election, that Trump is perfectly justified in "stealing" the election himself.

      Collapse
      1. CW

        "if all these Trump haters REALLY believe that Trump is a fascist tyrant, who is the Second Coming of Hitler Himself, and that a 2nd Trump term means that the US turns into Russia with Trump as President-for-life, then why *wouldn't* they try to steal the election?"

        You don't know us. We love democracy with our hearts and our souls and if we stole an election we would no longer be a democracy. Why do you think we want everyone to vote? We don't sit around devising schemes to block Republicans from voting like they do to us. We knew that the only way we could win was to outvote them and we worked our butts off to make that happen despite all the obstacles thrown our way and Biden has now won by 7 million votes.

        Yes, we REALLY believe that our democracy would not have survived four more years of him, and we are justified in that thinking. But we don't destroy democracy because we fear democracy's demise by the despot. If he had won we would have continued the fight.

        Collapse
      2. Avatar photo
        GREG BROWN

        I can't speak for Antifa or socialists, but I can somewhat speak for my own (the right - and Trump supporters): we are incapable of taking out the garbage, let alone scheming to steal an election

        Collapse
    2. GregS

      Oh yeah it had nothing to do with trying to ban an entire religion from traveling here, or once President, being explicitly hostile to the half of the country that didn't vote for him. So of course now they think that the Democrats would cheat because they were angry that Trump won. The Republicans were just sweetness and light when Obama won and didn't try to claim that he was an illegitimate President because he was lying about his birth and was not even a natural born citizen. I'll repeat something I've said before - when you start doing things with the excuse that "the other guys would do worse if they had a chance" you've gone down a very very bad path that quickly becomes a one-way street to societal collapse.

      Collapse
      1. Avatar photo
        GREG BROWN

        Ima say, you want to equate caviling about a birth certificate with accusing a sitting president of treason - well, you can do that. We can agree that each time one side escalates we draw closer to societal collapse. Maybe they should have thought of that during the Bork confirmation hearings (whatever). The Chinese watch this food fight like the leopard on the branch watching the chimpanzees below squabbling over nothing

        Collapse
        1. Avatar photo
          Greenknight

          Considering that the 1,000+ page GOP-led senate report indicated 140+ direct meetings between the Trump campaign and Russian government representatives, including the passing on of internal polling data to a Russian operative (to better target the influence campaign that the Russians were running on Trump’s behalf), I can see the justification for accusing Trump of treason.

          Collapse
  6. Avatar photo
    Bill Nelson

    Here is an explanation of the current favorite, the Georgia video:
    https://www.wsbtv.com/news/politics/georgia-election-officials-show-frame-by-frame-what-really-happened-fulton-surveillance-video/T5M3PYIBYFHFFOD3CIB2ULDVDE/
    Again, someone took a video segment, clipped it, knowing that it was being used in a false accusation.

    Collapse
  7. Vitaly

    I wanted you guys to comment (wherever possible on the the following paper published on Substack:
    https://votepatternanalysis.substack.com/p/voting-anomalies-2020

    The author do a statistical analysis of vote updates reported on Election night and next morning. They analyze these vote updates for both the absolute difference between the votes for Trump and Biden, and also the ratio of the votes in each update. They identify four statistically anomalous updates, all reported in the wee hours of the morning on Nov. 4th. While there had been numerous claims about election fraud, this one looks to me the most persuasive. Personally, I see two potential problems with this analysis:
    1) They rely on publicly available data from the New York Times. How reliable is the data? Shouldn't we be looking at the county data that was reporting?
    2) They assume that the number of votes reported should be over time somewhat randomly distributed. This may not be the case with very large counties that could report a single update after tabulating a large number of votes for a single candidate.

    I wonder whether someone who is really good at statistical analysis and has access to good county-level data could look at this.

    Collapse
    1. Mark

      I recall seeing this article pop up a few days ago, and I had several issues with it. One fundamental issue is that it doesn't clearly state the model assumptions: they make a claim that absolute vote margin are negatively correlated with the percentage of vote share won, but don't actually state that they are claiming the votes are randomized across vote updates. Of course, the reason they don't do this is that such a claim would be obviously false because we know that larger vote updates tend to come from precincts with higher population, and precincts with higher population tend to vote Democratic. They also don't take into account the partisan differences ballots from those who voted by mail and those who voted in person. They don't do this because in the end they aren't actually modelling the election - they are just highlighting data points that could be made to look like they support their claims. And when I say "made to look" it is because they rely heavily on transforming the data in ways that often do not make any sense unless your goal is to make the vote updates in question look irregular.

      I don't have time to go through each and every claim they make, but I can give an example or two of the authors committing statistical malpractice to fit their narrative. Example 1: after talking about using ratios as a way to identify unusual vote updates, they show a bar chart (for Michigan) of the ratio of non two-party votes *multiplied by the absolute number of votes for that update*. This transformation invalidates the core of their argument, because the data being charted is no longer a ratio! Of course, vote update in question has a larger number of absolute votes than almost any other update, so this transformation makes the accused update look much more irregular than it actually is. Thus we are left to conclude that either the authors are incompetent statisticians, or they are deliberately manipulating the data to make claims of irregularity look more compelling than they actually are.

      Example two: when graphing vote updates all across the nation, they explain they are standardizing the data by state, but make no attempt to justify this decision. There is a massive problem with this: by standardizing by state the authors are implicitly claiming that each state has voter distributions which are identical in shape, only differing in magnitude. That is quite the assumption! I do not see how anyone in their right mind would actually stand by it when called out on it, but in any event the authors decline to offer any justification for their decision beyond saying that "it provides precisely the functionality we need here." And that of course is no justification for standardizing by state, and it just reeks of finding a transformation to fit a narrative rather than to analyze the data.

      Collapse
    2. Avatar photo
      TimNYC

      I'm not a professional statistician, but I work with usage statistics and data analytics for the web as my day job. My main training was in English, theatre, and communications. That said, I see a few red flags here:


      1) They're really leaning in to the stats and math jargon.

      The issue with that is, that often, but not always, using jargon during the presentation of an argument is an effective tactic to make a weak argument sound stronger than it is. It doesn't affect the substance of the argument, of course... but if the jargon used is complex and impenetrable enough that the reader *cannot understand what they're reading* — as happened to me at a few points in this piece — the natural inclination of most readers is to *assume that the author knows what they're talking about, since obviously they know more than you*.

      This then engenders trust on the part of the reader, and makes them more inclined to go along with other, more transparent and (perhaps) spurious claims that an author might make, or not question those claims as strongly.


      2) There's actually no information given on the author. None whatsoever. If someone truly believed in their analysis — and if they had the credentials to back it up — wouldn't they be willing to provide at least basic information about who they are? What their credentials are, something that's verifiable? Instead... nothing. No way to judge their credibility.


      3) They don't actually seem to offer any *direct* evidence for fraud. Over and over, this piece looks at statistics — with math that I presume is correct or at least justifiable — to highlight a statistical anomaly. Unfortunately, a statistical anomaly does not mean that fraud occurred.

      Showing that something unusual happened doesn't mean it's fraudulent. It just means something unusual happened.

      And even then, a while back I remember reading an article about a rather frail but charming-seeming 96-year-old gentleman who survived his fight with COVID-19 and was able to leave the hospital successfully, despite being ventilated. That was a rare occurrence — and a fortunate one, for him. Such things are... unlikely, but things that are statistically unlikely *occur all the time* if your sample size is large enough. And isn't an American election a pretty large set of data to choose from?


      4) Delving into the actual math a bit, they claim that anomalies in *when votes were reported* suggest that the vote counts were manipulated.

      However, this is a specious argument though, because 2020 is inherently an outlier when it comes to the tabulation and reporting of votes, due to the *overwhelmingly large* number of people who voted absentee compared to previous years.

      Further, as has been amply demonstrated in the media, Trump's repeated bashing of absentee ballots had the effect of generally driving Trump voters away from that method, and towards in-person voting... while Biden voters, who generally would not be temperamentally inclined to listen to Trump much on a matter of political preferences, were more likely to vote absentee.

      *This alone is an adequate explanation* for why so many of the absentee ballots counted sided with Biden this year, as compared to the much more even split among absentee voters in previous years. Absentee ballots are counted using different processes than in-person ballots.

      We used absentee ballot counting methods on a much larger percentage of the votes than normal, this year. Why should anyone be shocked we saw results that were in line with those processes when vote counting was completed?

      They can claim it's suspicious all they like... but it isn't actually suspicious.


      5) Statistics are meaningless without context. And the way this is written *reeks* of p-hacking.

      What is p-hacking, you ask? It is, effectively, when statisticians and analysts pick the data analysis methods that lead to a desired result.

      For example, at my day job, I analyze usage data for the company website and app, and pull together the most important stats that would succinctly capture the big picture for my boss and his superiors. Usually this would involve things like, comparing the number of successfully logged in users to the number of reported problems. Obviously, I would present that data faithfully and accurately... but there were times when (due to circumstances beyond our control) unintended problems would occur.

      For example, once, one of our partners (that has mirrored servers for the company website all around the world, to help reduce load times) had a major problem. This increased load times dramatically.

      Further, because local traffic tended to be routed through their servers before it reached us, and many of their servers were down, the company website would spontaneously appear to "die" for many customers. Who would then call or email us... about an issue we were powerless to fix ourselves.

      The few days that that was an issue, our number of reported complaints was through the roof when compared against that week the year before. And when it came time to do my reporting, I had the problem of how to communicate it in a way that didn't cause *genuinely unnecessary* alarm.

      Now, I could have omitted that stat... but instead, I (correctly) showed my bosses that there had been a 527% increase in customer complaints relative to successful users for that week year over year... and I also gave them the necessary context to *understand* that number.

      When taken in the full context of the business, it was clear the increase in the complaint/success ratio this week was due to an issue with one of our partners, and was nothing that any action by anyone at the company could have reasonably prevented; that our customers were being responded to in a timely fashion as usual, and that the issue had already been corrected and was not expected to recur.

      When I'm reading this piece... it feels an awful lot like I'm looking at a report like that, with big, scary looking numbers called out — but without the necessary context to understand it. (This also circles back to point #1; namely, how jargon-heavy the writing is.)

      And lastly, and this is the big one:


      6) There's no historical context, or any attempt to prove the reliability of their methodologies in analyzing other elections!

      Their analysis works on the assumption that patterns in vote reporting can expose fraud... but there is no evidence offered to back that claim *anywhere* in the piece that I was able to find.

      If a claim like that is verifiable and based in evidence, the *very first thing they would do* is present their analytical methods correctly flagging cases of known fraud in past elections. Even if fraud in American elections isn't at enough of a level to routinely sway elections, if their methods were as valid as they'd like us to think they are, it would be a repeatable, measurable statistical phenomenon when looking at historical data.

      They don't do that though. There's no evidence like that shown. And absence of evidence is evidence of absence — or that the claim they're making does not stand up. If they actually had the ability to point out the necessary context (to show how *deeply unusual* this is... or how *not actually unusual* one aspect of it is) they would do that; it's a central part of communicating, well, anything really.

      Instead, all they're doing is throwing numbers at us.

      So, pardon my language please, but this, to me, looks an awful lot like bull$#!t. It positively reeks.

      Collapse
      1. DJ

        Statistically anomalies happen all the time if you look for them. Trump winning in 2016 was an anomaly! The idea that a sitting president who never got over 46% approval lost the election is the regression to the mean.

        Collapse
  8. Avatar photo
    Patti Scriffiny

    Thanks for the fact checks. They really help me get a sense of what is happening since they reference specific people who are in a position to know the right information to answer a concern.

    I was called upon to use some of that information recently. I am a HS math teacher and a colleague sent me a link to a lengthy article that focused on an analysis of the overnight vote updates from Nov 3 to Nov 4. He said it seemed legit when he read it and he wanted my take on the mathematical reasoning (I teach statistics).

    I used information I’ve learned from TMD, various fact checks, and information from David and Sarah (the interview with Ben Ginsburg among other things) to answer his concerns. I also pointed out that the presence of outliers proves only that there are outliers - even randomly generated data can produce those. He was simply loathe to give up his suspicions. COVID hypocrisy has made him distrustful of anything that doesn’t come from his team and I don’t know if that will change. The one thing I shared with him that made him pause was my response to his comment that he couldn’t believe how he went to bed on Nov 3 thinking that Trump was leading and would win, only to wake up the next day to see things had changed. When I told him that was the result I expected, he came up short. I explained how different states had counted ballots in different sequences, and so I’d expected those swing states to move towards Biden overnight. He’d never heard that because no one in his silo told him. But at least he asked me. I don’t think it made much difference, but maybe it did.

    Collapse
    1. DJ

      Statistical anomalies show up in practically any large data set if you squint. Trump's win in 2016 was a statistical anomaly!

      Collapse
      1. Avatar photo
        Patti Scriffiny

        Yeah. I spend a decent amount of time and effort in stats to help kids see that 1) we are terrible at recognizing what randomness should look like and 2) that outliers do occur in random processes.

        Collapse
  9. Scott Mc

    Rachael another wonderful weekend treat. Thank you. Apologies for name calling but it was how I was raising eons ago. You are awesome, funny, brilliant, wise, reflective, caring and well maybe I shouldn’t call you too many names. Love your work. Best to the role at home too lots of work to balance a big job here and a huge job there. These times (relativity does matter) are hard and can set people on unusual courses. Being focused on their course and yours is critical and in time the laundry gets done and no one today actually can smell the zoom difference. Have a great weekend and week ahead.

    Collapse
  10. Avatar photo
    Robert Knudson

    I really appreciate the fact checks, although the folks I know who spread conspiracy theories are beyond the point where facts are an answer for the questions that they are "just asking".

    I subscribe to the Dispatch because I find the editors and contributors to be knowledgeable, trustworthy and insightful about politics and policy and the fact checking is important for me because it speaks to your commitment to truthful reporting. Ultimately pursuing the truth embodies good journalism for me, and that commitment to journalism is part of what sets you apart from other sites.

    Collapse
Load More