This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.
Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.
We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:
-
Shaming.
-
Attempting to 'build consensus' or enforce ideological conformity.
-
Making sweeping generalizations to vilify a group you dislike.
-
Recruiting for a cause.
-
Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.
In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:
-
Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.
-
Be as precise and charitable as you can. Don't paraphrase unflatteringly.
-
Don't imply that someone said something they did not say, even if you think it follows from what they said.
-
Write like everyone is reading and you want them to be included in the discussion.
On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.
Jump in the discussion.
No email address required.
Notes -
An update on the J6 pipe bomber story. Steve Baker of Blaze Media is now claiming that a Capitol Hill Police officer planted the pipe bombs based on gait analysis from the videos released from the FBI of the pipe bomber and known footage of the Capital Hill Police officer. (https://www.theblaze.com/news/former-capitol-police-officer-a-forensic-match-for-jan-6-pipe-bomber-sources-say). Blaze Media also claims Shauni Kerkhoff the officer in question left the Capitol Hill Police to join the CIA mid 2021 to work on dignitary protection.
Blaze Media claims the gait analysis was a 94% match but I'm unsure as to how unique that match is or how suspects were selected for matching or how many suspects were tested. Apparently, the pipe bomber also walked with a limp and Shauni had a football related injury that required an operation. Hypothetically, if 1/100 people would score a 94% match then its very likely such a match could have been produced by just trying to match against all the Capitol Hill Police officers. However, if only one suspect was tested and this testing was based on some other lead then you would have more confidence that this match was not a coincidence. Shauni was the neighbour of a person of interest that was linked to the metro card that was allegedly used by the pipe bomber (https://archive.ph/wMRun).
Steve Baker was also arrested in relation to the J6 riots (https://loudermilk.house.gov/where-is-the-outrage-over-steve-bakers-prosecution-3/) and may hold some animus against Capitol Hill Police officers.
Congressman Massie has made a statement about the claim on X (https://x.com/RepThomasMassie/status/1987120156682953165):
@gattsuru makes some good points about the gait analysis, but I don't think you even need to go that far. I spent 2 1/2 hours at the DMV this afternoon and did some reading about gait analysis. I learned that the way it works is that analysts break gait down into components, and analyze those components into categories based on how prevalent they are in the population. Like a lot of other things, when a gait analyst says there's a certain percentage match, what they're saying is that, based on the attributes they observed, they can eliminate that percentage of the population. With that being said, from here on out I'll assume that the science is bulletproof, because I don't know that that even matters in this context.
I've read a lot of crime books in my life, and one of the things that's always interested me is suspect descriptions and how useful they are. I've read about cases where police failed to solve the crime because they seemed to focus on a description that wasn't very good, and others where they didn't solve the crime and dismissed good descriptions as being too vague. I've also seen authors excoriate police departments for not focusing on suspects who matched relatively vague descriptions. So during my time at the DMV I also thought about a rubric that could be used to categorize suspect descriptions.
A Level 1 Description would be one that eliminates 90–99% of the population. This may seem high, but anything less than that isn't really even a description. If the suspect is described as a black female, well, only 12–14% of the population is black, and about half of them are female, so that eliminates 93% of the population right there. If the suspect is described as a young, tall, white male, 40–45% of the population is white males, eliminating children and anyone too old to be reasonably described as young and you cut that in half, and cut it in half again to get rid of anyone shorter than average height, and you're down to 10%. These kinds of descriptions are of little to no use in a police investigation and are completely worthless in a trial.
A Level 2 Description is one that eliminates 99–99.9% of the population. These can be of some use in an investigation but are of little to no use in a trial. Suppose the man running from the scene was described as an African American teenager, short and extremely overweight. Take the 7% who are black men, teenagers being about 20% of them, divide by half again to get people shorter then average, then in half again to get anyone plausibly described as overweight (always use the larger numbers), and we're in that 0.1–1% range. But in most places there are going to be entirely too many short, black, overweight teenage boys for police to identify and question them all.
A Level 3 Description would eliminate 99.9–99.99% of the population, but still include between one person in a thousand and one in ten thousand. To give a few examples:
These kinds of descriptions are of value to police and may play some role in a trial, but no one could be reasonably convicted of a crime based on them. It's also worth noting here that some of the attributes are changeable, and this needs to factor into the analysis as well.
A Level 4 Description would eliminate 99.99–99.999% of the population, but still probably include a few people in any decent sized metro:
If you match a description of this specificity you should expect the police to come to your door, but it still wouldn't be enough to convict absent other information.
A Level 5 description would exclude 99.999% of the population or more, aka 1 person in 100,000 or less. This is the point where you stop combining combinations of independent variables that belong to lots of people and zero in on very specific attributes that are themselves fairly unique: A missing finger, a particular tattoo, one green nipple, etc. At this level you're on the defensive; if you match a Level 5 Description, you're going to need an alibi.
A Level 6 description is a description that applies to only one person: Fingerprints, DNA, being recognized by someone who knows you. A Level 6 Description is an identification.
I bring all this up because there's a certain level of obfuscation going on, both with the science and the use of percentages. Supposing we didn't have any gait analysis but a witness who told the FBI that he observed the person in the video and it was a man who appeared to be of East Indian descent, and Baker claimed that Rajneesh Sarna was the perpetrator on the basis that he's a male of Indian descent living in the DC area, everyone would find it ridiculous. Yet Indians only make up 3% of the population of the DC Metro, and assuming men are about half of those, and we're at a 98.5% "match". Actually higher because a certain percentage of that population is going to be children. All this 94% "match" means is that the police officer they're claiming is in the video has the same gait characteristics as 378,000 other people in the DC area, more if we allow for the possibility that the perpetrator was from elsewhere. Even a 98% match only gets us down to 126,000 other people.
Of course, that wasn't the only attribute mentioned in the article; it says that both the person in the video and Karkhoff are about 5'7". Being very conservative, about 10% of the population can reasonably described as 5'7". It's the point where the bell curves cross, which makes things convenient, and about 9% of men and women will be this height. I don't know how accurate the FBI estimate is supposed to be, but we'll assume it's pretty accurate and just bump the numbers up to 10% to allow for a little wiggle room (an inch on either end would make this closer to 25%). That gets us in to Level 2 description territory, but still includes over 5,000 people. The Blaze engaged in motivated reasoning by linking this to a Capitol police officer and working backwards from there. An honest assessment would have looked at any surveillance video from DC they could get their hands on and analyze the gait and height of as many people as possible. Of course, if that information was fed into their computer and they ended up identifying a 45-year-old cashier from Landover, Maryland as the only possible suspect, they never would have published the story, because it would have been ridiculous. And that there's one person in thousands who happens to have been employed in some law enforcement capacity in the DC area makes things really convenient for them.
If that were the end of it, we could put this nonsense to bed, but there's also the whole business with the Metro card. As per the article:
I apologize from the long quote, but I wanted to include it as-written to point out something here that's particularly dishonest. The article refers to "the suspect", and from the context it looks like it's referring to Kerkhoff, and it talks about how the suspect used the card to travel to Virginia after planting the pipe bombs. Not being familiar with this evidence, I presumed the story to be making this point: Kerkhoff was the person seen in the surveillance video released by the FBI. The video was taken the evening of January 5, and while it doesn't show the bombs being planted, it shows the person who likely planted the bombs walking with a backpack in the vicinity of the targeted buildings. This person then went into the nearby Metro station and took it to Falls Church, Virginia, where they were picked up by a friend whose Metro card they used. The FBI surveilled the friend's house but were pulled of for reasons that weren't explained to them, and weren't allowed to talk to the guy, who happened to be a civilian employee for the Air Force.
As many of you know, I'm a fan of reading official reports to get all the details as best as they can be known, as piecing things together from news reports and the like is time consuming and doesn't usually contain all the necessary boring details. In January of this year, a joint report on the pipe bomb investigation was issued by a group of congressmen representing subcommittees with names too long to mention here, chaired by Massie and Loudermilk. This report was incredibly critical of the FBI's response. What we learned about the whole Metro card thing was far different than what was implied in the article.
Surveillance showed a suspicious person (POI2) photographing a dumpster near where the RNC pipe bomb was planted before meeting with 2 other people and disappearing into the nearby Metro station. This video was from the morning of January 5. The FBI was able to link POI2 to the Metro card of a man living in Falls Church (POI3), and had the FBI surveil both. POI2 was a man. The FBI interviewed him and reviewed the pictures on his phone, which corresponded with his story that he had been taking pictures of numerals on doors and the like, including numerals on the RNC dumpster. The FBI was satisfied that the guy had nothing to do with the bombing and they eliminated POI2 and POI3 as suspects on January 19.
Seraphin's story about what happened seems a lot less suspicious in this context, and it's pretty clear that he was a minor part of the investigative team who didn't know the whole story behind what was going on. The FBI identified POI3 and put him on a surveillance team. In the meantime they interviewed POI2 and eliminated him as a suspect. POI3's status was contingent on POI2 being involved, and once POI2 had been eliminated there was no reason to interview POI3 or continue to watch his house. As for Kerhoff's allegedly living next to POI3, so what? It's an odd coincidence, but what does it really mean? Someone who is a 1 in 5,000 shot to be the person in the video happens to live next to someone who had nothing to do with the bombing.
This has to be some of the worst "journalism" I've ever seen. They're using "the suspect" to refer to people seen in two different videos, one of whom was already identified by the FBI, interviewed, excluded as a suspect, and not the same sex as the person they're accusing. I don't know if they're intentionally engaging in misdirection or if Baker is simply incompetent, but neither would surprise me.
I think that you are correct in you assessment of the probabilities. However, another consideration is that the gait analysis evidence was reported on simply because it was found to be somewhat significant, so it is subject to p-hacking considerations.
Fundamentally, I am not sure how a Bayesian should update on true but potentially adversarially selected evidence. As an intuition pump, consider two persons A and B which are in a relationship which is supposed to be exclusive. A suspects that B is cheating. B proposes to send A screenshots of their text messages. Normally, a random sample of texts which contain no evidence for misbehavior would be at least weak evidence of the absence of such misbehavior. But A will not get a random sample, but potentially a curated subset selected for being misleading. Thus A should not update at all on receiving harmless screenshots (beyond the signaling value of going through the effort of sending them, at least). If A is willing to update even a tiny bit on such a screenshot, B can take them for a ride.
On the other hand, "the evidence presented by my enemies was adversarially selected" is a fully general counterargument. Most of the evidence which we use to build a world-view does not come with a strict chain of custody to guarantee that it was randomly sampled and reported without publication bias etc.
I have no good way to resolve these two viewpoints. In criminal justice, the idea to allow both sides to make their best case certainly seems helpful.
More options
Context Copy link
More options
Context Copy link
More options
Context Copy link