Shiri's Margin Attractor

How media controversy maximization drives election differential to the margins

A reader challenged me in the comments section of an earlier article to defend an axiom I’ve been using, but honestly not checking. Today I checked it. The results are fascinating and useful in both overall political theory and in crafting action plans for advocates of issues of any stripe. You’ll like this. We’ll begin by describing the axiom in the most esoteric way possible using Slate Star Codex terminology, HWFO terminology, and regular old muggle words as well. Then we’ll test the hypothesis in section two, graph it, and finally talk about how policy advocates might apply it. With guns of course, because that’s what we do around here.

Shiri’s Margin Attractor

In the mathematical field of dynamic systems, an “attractor” is a state or set of states towards which a system tends to evolve. I posit I’ve found one.

Fans of Slate Star Codex will recall some seminal articles by Scott Alexander discussing culture war and media behavior, from which we can build the Shiri’s Margin Attractor theory directly by blending them with some HWFO material. In December of 2014, Scott published The Toxoplasma of Rage, a long look at his own blog’s traffic concerning what went viral and what didn’t.

Scott on the Rolling Stone UVA fake rape story debacle:

Only controversial things get spread. A rape allegation will only be spread if it’s dubious enough to split people in half along lines corresponding to identity politics. An obviously true rape allegation will only be spread if the response is controversial enough to split people in half along lines corresponding to identity politics – which is why so much coverage focuses on the proposal that all accused rapists should be treated as guilty until proven innocent.

And on Michael Brown:

I propose that the Michael Brown case went viral – rather than the Eric Garner case or any of the hundreds of others – because of the PETA Principle. It was controversial. A bunch of people said it was an outrage. A bunch of other people said Brown totally started it, and the officer involved was a victim of a liberal media that was hungry to paint his desperate self-defense as racist, and so the people calling it an outrage were themselves an outrage. Everyone got a great opportunity to signal allegiance to their own political tribe and discuss how the opposing political tribe were vile racists / evil race-hustlers. There was a steady stream of potentially triggering articles to share on Facebook to provoke your friends and enemies to counter-share articles that would trigger you.

And on his own blog:

Compare this to the three most viewed category of post. Politics is self-explanatory. Race and gender are a type of politics even more controversial and outrage-inducing than regular politics. And that “regret” all the way on the right is my “things i will regret writing” tag, for posts that I know are going to start huge fights and probably get me in lots of trouble. They’re usually race and gender as well, but digging deep into the really really controversial race and gender related issues.

The less useful, and more controversial, a post here is, the more likely it is to get me lots of page views.

By 2014 the cat was out of the bag to Slate Star readers that publishing controversial opinions gets you the most clicks, but I wasn’t a Slate Star reader at the time. HWFO started in 2018 talking about why nobody seems to understand the truth about gun policy mathematics, and landed on exactly the same conclusion in this post, which led to the very name of the blog you’re reading now:

From analysis of my own numbers, I determined that I was earning about a third of a penny per click, and so was Vox, and their organization is massive while mine’s not purely because of traffic. Which explained why Vox was lying in their gun graphs. They were doing it for traffic. I later confirmed the “third of a penny per click” number with Ken LeCorte, the former director of Fox News’s web division, in a podcast that appears to have been banned from Youtube. Censors really don’t like Ken LeCorte.

Later in 2018, Scott drove the point home with a fictional account of a tech company who trained an artificial neural network against Reddit posts to create the most controversial statement possible, a statement so controversial it could tear families, friendships, and societies apart. The story is Sort By Controversial, and the fictional software in it was called Shiri’s Scissor, which they inadvertently released upon themselves leading to a dust-up that caused half the software company to quit.

At no time in our five hours of arguing did this occur to us. We were too focused on the issue at hand, the Scissor statement itself. We didn’t have the perspective to step back and think about how all this controversy came from a statement designed to be maximally controversial. But at 8:01, when the argument was over and we had won, we stepped back and thought – holy shit.

We were too tired to think much about it that evening, but the next day we – Brad and the two remaining members of the coding team – had a meeting. We talked about what we had. Blake gave it its name: Shiri’s Scissor. In some dead language, scissor shares a root with schism. A scissor is a schism-er, a schism-creator. And that was what we had. We were going to pivot from online advertising to superweapons. We would call the Pentagon.

And so forth. It’s a good yarn, and what we drag from this yarn is the idea that modern media companies are not calibrated for truth in any way, they are calibrated to find things that are maximally controversial to generate the most thirds-of-a-penny possible. This is a repeated[1] theme[2] on HWFO, and foundational to our sensemaking approach[3].

This drive to create the most controversial content exists at the media layer, and the sharing layer, and the social media argument layer, and the three layers connect in three phases of the evolution of media itself.

Concocting controversy occasionally sold in the old era of traditional media with one newspaper and a hand full of broadcast TV channels, but it also turned people off because the media wasn’t targeted. One newspaper for the same town went to Republicans and Democrats and Socialists and Libertarians and People Named Steve on a subscription basis, and every car dealer published their ads in the same paper. The first step in Shiri’s Attractor was when 24 hour cable news factionalized with the founding of Fox News. Before, CNN advertised for Subaru station wagons and Dodge Ram pickup trucks equally, so they were beholden to provide content that each of those consumer demographics might watch. After Fox News started, Dodge bought Fox ads, Subaru doubled down on CNN, and such. This marked the first phase of demographic targeting of news, what we’ll call the Partisan News era, beginning in 1996.

The second phase was the creation of the internet, which democratized and decentralized news. Anyone could be a news agency, anyone could seek out a news agency that confirmed their biases, and the fertile ground was laid for nigh infinite A-B testing for controversy. The Internet Era started in 2000, concurrent with the Partisan News Era, and our modern echo chambers begin to form, but the overall media consumer organism hasn’t maximally evolved for controversy.

The third phase, social media, allows consumers to more easily use the internet to share things with like minded people, and argue with opposite minded people, and the arguments carry references to more media, and the most controversial things get shared. This is is the actual A-B controversy testing function that the second phase lacked, and is what led Scott Alexander to notice the trends in his traffic, and I in mine, and Vox in theirs. The Social Media Era begins in 2008.

Shiri’s Attractor then comes together in the ballot box. If people vote based on their world, and their world is slowly being absorbed by attention whoring smartphones, and the attention whoring smartphones are full of media, and the media is seeking to make the most thirds of a penny out of maximum controversy, and the users of the smartphones are constantly A-B testing for controversy in their sharing behavior, then the world views of the public are going to divide along lines of controversy. That’s Shiri’s Attractor. If this dynamic mathematical system were indeed transpiring, we should be able to track it with data. The ballot box should be trending towards 50/50 splits. Or in other words, the gap between the winner and loser should be shrinking since 1996, and then generally flattening out.

Is it?

Testing the Attractor

If you take the presidential election popular vote totals for every year since 1900, isolate the two major parties, and chart the percent difference between the winner and loser, the results are noisy and look like this:

You can see a wide variation in results, with frequent 20%+ blowouts mixed into some races which were so close they might have been influenced by Guidos, but the wide variation all transpires in the pre-internet days. It’s difficult to make too much sense of this because of the variation, but with widely varying trend data we often do a rolling average to smoothing it and pick out the real trend. If we instead start our graph at 1908 and calculate a three-election rolling average from then on, we begin to see the real trend emerge:

Here I’ve colored the eras mentioned above, where Shiri’s Attractor should start to come into play. Before 2000 the eight year rolling average of the popular vote gap never goes below 7.3%. After 2000 it never goes above 5.1%. Even the individual year data after 2000 without any rolling averages never gets higher than 7.2%, during Obama’s “blowout” win. It’s clear from the data that something is suppressing the gap, meaning something is attracting the popular vote results towards the 50/50 boundary.

If you point this trend out to a partisan of either flavor they will disagree and blame the “other news” for being fake and increasingly misleading an ever larger number of people away from the truth - the “truth” that their news is telling them. While that’s possible, Shiri’s Attractor theory predicts that they’d both say that.

This is only one test because I’m not Nate Silver and don’t have a bunch of interns working for me, so I don’t consider this test to be conclusive proof of the theory. If Nate wants to run with it I welcome him and hope he’ll credit me or offer me a job. But presuming it’s true, we can take a look at how to use it.

Application: Guns

I’m agnostic to almost all political issues other than guns, but I do think quite a bit about gun activism in my spare time, and HWFO has mentioned before that Guns have effectively won the culture war:

Post 2020 “Guns” has the entertainment media, the courts, half the federal government, somewhere between 47% and 53% household ownership rate, more semi auto rifle owners than the entire Asian American population, around as many 10+ capacity mag owners as the entire Black population, Second Amendment sanctuary counties and states, and is making huge inroads in the gay community, the Asian community, the Black community, and the Jewish community. Half the states in the union have no permit to carry. The kids all want to buy the thing they used in Fortnite when they turn 18. This is what winning looks like. Shiri’s Attractor theory can show the gun rights community how to seal the deal.

Shiri’s Attractor puts a dynamic bound on elections at somewhere around 6%. The Republican contingent is so all-in on gun rights that they’re performatively wearing AR-15 pins in congress, so that puts the floor of gun rights at 50%. If Shiri’s Attractor establishes the election bound at 8% delta, that’s 54/46, exceeding even the delta on the 2008 Obama win. If one in ten regular Democrat voters began to value guns more than other issues, then the Democrat establishment must choose to permanently drop gun control from their platform or else lose forever. Either is a permanent win for guns.

This sort of analysis works for other issues too. The Democrats could easily apply this sort of analysis to abortion rights, for instance, and might make inroads into the Republican base by softening their ask and going for something more like what most of Europe has. The value in the Shiri’s Attractor analysis is it establishes the bound over which your issue wins, and while those bounds used to be very large, they’re now sucked in to the margin.