CSotD: The Parable of the Bad Shepherd

Phil Hands leads off today with a narrow accusation I want to expand into a continuing theme.

The refusal to wear masks, or, more broadly, the refusal to take Covid-19 seriously, requires a level of gullibility that should boggle the mind.

That is, it’s not just “bad science.” The anti-vaxxers can at least string together apparent evidence by failing to really understand what they’re citing, beginning with the fact that a lot of autism begins to exhibit at about the age when vaccinations happen.

It’s bad science, but it’s understandable.

Whether or not to wear masks is an idiotic strain of bad science, but, more to the point, if you look at the statistics, if you look at what’s happening in the rest of the world, the belief that Covid-19 is not real, or not any worse than flu, descends to the level of idiocy.

Simply comparing the number of deaths in past years with the number of deaths in the current year should be enough to convince anyone that something is going on.


This Washington Post story even provides a graph that you don’t need to be a statistician to understand.

Now, the denialist would say that people aren’t actually dying from Covid-19, the answer to which is, “Then what’s killing them?”

The only answer being “Those are fake statistics.”

Which satisfies my definition of “sheep.”

And if the only paranoid, absurd thing they believed was that the CDC and state health bureaus are conspiring to gin up false mortality statistics, well … okay, you’d still think they were nuts, but nuts in a well-contained area.

Alas, these nuts fall far from that specific tree.


It’s a level of loyal gullibility that makes this Tom Toles piece my favorite commentary on Trump’s inexplicable insistence on not just re-opening schools on time but on making it as “normal” as possible.

There is a certain Berlin Bunker desperation in his desire to make everything appear normal when it plainly isn’t, and Toles captures the evil intent in sacrificing our children to create an atmosphere of “Everything is fine. Ignore those other people. Big Brother loves you.”

Yesterday, Dear Leader proudly tweeted that he has a 96% approval rating among Republicans, though Rasmussen puts it, rather, at 80%, which is still pretty good until you realize that only 25% of voters identify as Republicans, so he’s bragging about having the approval of about one-fifth of voters.

If that were all the support he had, we’d be in much better shape, but only about a third of voters identify as Democrats, so his penetration within the Independents matters. A lot.


It results in this, according to Fivethirtyeight, and, while 9.6 percent isn’t a crushing gap, it’s well beyond the margin by which Clinton led him in 2016, and, while the margin of error let him take the Electoral College, she did get more votes.

Today, it’s as if he’s given up and shifted to doing whatever the hell he wants, which brings us to our

Juxtaposition of the Day

(Rod Emmerson)


(Deb Milbrath)


(Jimmy Margulies)

One of these cartoons was drawn in the past 48 hours, two were drawn last February.

The Margulies piece obviously dropped when Stone was sentenced; the other February cartoon is Emmerson’s.

The date in the margin aside, Milbrath offers a clue, because Dear Leader has been tweeting about “Law and Order,” which is a giveaway that he wants the latter because he clearly has no intention of obeying the former.


And if you had any questions about his commitment to good governance (you did?), Kayleigh MagaNinny’s press release on the topic tells you exactly where we stand.

There’s no need for me to go through it lie by lie and spin by torque, because CNN has done an excellent job, in an analysis posted under the no-shit headline “Debunking 12 lies and falsehoods from the White House statement on Roger Stone’s commutation.”

To which I would add that, finding 12 lies and falsehoods in a document of only 628 words doesn’t leave a lot of space for truth. As near as I can tell, the true statements are (1) the date and (2) the fact that it was issued by the Office of the Press Secretary (3) for immediate release.

It’s not quite the Saturday Night Massacre, because it’s only an indirect interference in the legal oversight of a corrupt presidency and that was quite direct.

But you have to award style points for the number of GOP loyalists Trump still appears to have under his sway: By this point, Nixon had been nearly abandoned.

I like John Deering’s take on the current situation. We’ll get into the pair of Supreme Court decisions another time, but they aren’t the only source of Trump’s “slow descent,” and tying those losses into his embarrassing performance at West Point — an event, BTW, that he insisted on staging — is brilliant.

Certainly, a wounded animal is dangerous, and people are speculating about what may happen between November and January if he loses the election.

First of all, I think what will happen is that he’ll find out the truth of the lyric, “Nobody knows you when you’re down and out.”

The other is that I have said before that his administration could end with tanks surrounding the White House, the question being which way the turrets would face.


I feel pretty good about that at the moment. Not only is our military leadership not backing him up on his intent to preserve white supremacy in the names of bases or the freedom to fly a traitor’s flag, but they have also come out against his wish to use the military to quash civilian demonstrations.

My impression, first hand and distant, is that the great majority of our military take their oath to uphold and defend the Constitution one helluva a lot more seriously than their CIC ever did.


They may not draw the same parallel Clay Bennett did, but they’ll  rally ’round the flag, whether or not they actually hang anyone from a sour apple tree.

Would there be dissent in the ranks? Sure.

But that’s part of what they’re sworn to defend.


6 thoughts on “CSotD: The Parable of the Bad Shepherd

  1. ” a graph that you don’t need to be a statistician to understand.”

    Ah, but this is a classic case of misleading graphing. The vertical scale starts at 40,000, not at zero as it should. This makes the excess look like about a doubling!

    There’s an important point there, and some valuable statistics, they shouldn’t have undermined it by a false exaggeration visually.

  2. At least they included, off to the side, a small representation of what the graph would have looked like with a baseline of zero. Otherwise, it’s what Huff called a “Gee-Whiz Graph” in his immortal treatise How to Lie with Statistics.

  3. Formatting the graph such that the main view extended all the way down to 0 would be a massive waste of blank space, in addition to making it harder to judge individual variation. This is the only correct way to present a data set that has high individual variation, but no points close to 0.

    Arguing that it’s misleading if one doesn’t read the labels is like saying knives are painful if you shove one into your stomach.

  4. Kip Williams, thanks for pointing out that inset off to the side. It indeed does ameliorate the potential distortion of the main graph.

    Nicolas Merritt, I can hardly believe the amount of heat you chose to direct on this point.

  5. Misleading would be failing to label it at all, or using a log scale without indicating such, or other similar failures to provide the necessary information in an obvious manner. Using a nonzero baseline when the data is clustered at a significant distance from zero is only good practice.

    One might argue that they could have instead graphed the percent change from expected for this year, which could easily use zero as the baseline (given that it appears to be no more than 15-20% on the high end), but that would run into the usual problems with only talking about a percentage – without the absolute numbers, it’s much harder to tell how significant of a change it actually is.

    Using both might have been better, but by this point we’re getting into whether or not the extra space taken up by the new graph (not to mention the extra time and pay to create it – no idea what their process is or if they have a good tool for that) would be worth the marginal increase in information conveyed to the readers. In my opinion, it would be better to simply state the highest percent change in the accompanying paragraph and leave it at that, as the graph itself isn’t a series of spikes.

  6. You never let me down
    I wish I could cartoon.
    I would a video reenactment of the Gong Show with Chuck Berris. I would gong every RETHUGLICAN candidate
    One by one gonging each one to defeat.

Comments are closed.