What's this? Another confounded tree? 10/10 and a very merry Multistage Simpson's Paradox Machine* from all of us here at WeRateDAGs! *Pearl, J. (2014) 'Comment: Understanding Simpson’s Paradox' The American Statistician, 68(1):8-13

Signs that the author knows the person they're citing -- accidentally using the first name :)

It's astonishing how a technology called artificial intelligence has revealed the actual stupidity of so many people. This tweet hasn't been bettered

I think I need to write something on "over-control" -- cases where the ability to fine-grain intervene reduces the variance (i.e. risk aversion) in a way that's long-term suboptimal. Very related to Alison's work on children exploring (vs. exploiting), and I'm starting to see it everywhere.

Some cybernetic blogging to kick off the holidays: why does framing decision, design, and discovery as optimization remain so irresistible?

😂. The issue of disentangling the causal effects of age is conceptually rich, and entirely untouched by philosophers.

The richest man in the world has killed hundreds of thousands of the poorest children in the world I think this is one of the most evil--and definitely the most anti-christian--acts in human history

🚨🚨🚨 Our 52nd Annual Meeting will be held from June 18–20, 2026 at Johns Hopkins University in Baltimore, MD, with a pre-conference on Mental Control and Agency held at JHU on June 17 🚨🚨🚨 We are currently inviting submissions of papers (talks and posters)!

1/2 New open access paper in which we apply the Nyquist-Shannon thereom from signal processing to 2 EMA datasets to figure out the optimal sampling frequency for EMA assessments. 🧪 #psychscisky #statssky www.cambridge.org/core/journal...

Can't really explain it, but I think of watching TV as kind of like looking into someone's living room, reading a book as having a conversation, listening to a podcast as like overhearing to a conversation, and listening to the radio like listening to a public speech.

Our special issue on Evolutionary Functions of Consciousness, coedited with Tecumseh Fitch and Adina Roskies, now online royalsocietypublishing.org/toc/rstb/202... Contributions by (1) Irina Mikhalevich; (2) Eva Jablonka and Simona Ginsburg; (3) Nicholas Humphrey; (cont'd)

This week's episode is on the Manipulation Theorem, which is not well-known in the broader causal community, but has hugely influenced philosophers. I cover hard vs soft interventions, why causation doesn't require human agents, and whether causation requires open systems. youtu.be/fg1ApidilVM

"Are we the baddies?"

I was honored to be asked to comment on David's talk. Many of you know him from his philosophy of physics work, though my interaction goes back to his teaching the 1st course I took in philosophy of science. Full talk in Sean's link. For my commentary, start here: www.youtube.com/watch?v=8TwP...

Continuing our drive to rate the classics, here are two maximally LA-coded suburban DAGs from Judea Pearl. There's thirsty lawns and sprinklers, burglars and earthquakes, and neighbors who call you if your alarm goes off. 9/10 because there's no mention of the traffic.

Kids don't need us to be their buddies. They need us to be what our teachers were—firm, fair, and just a little out of touch. Someone's got to show them how to age with style.

I'm revising a paper, and I still haven't managed to force myself to remove a snarky comment that doesn't seem uber-professional, but which best captures the reaction I'm trying to pre-empt. What do y'all think?

We swear that Bielefeld exists! We even had a speaker coming from there. And his train was delayed. In Bielefeld! Which totally exists!

What does it mean to assign probabilities to causal parameters? In just one minute and 33 seconds, I explain why this is puzzling, and how you can nevertheless make sense of it. youtu.be/UMJSFkz2sCg

Next talk from the #NaturalPhilosophy Symposium is David Albert on Fundamental Laws of Physics and the Idea of Agency, with commentary by Naftali Weinberger @dagophile.bsky.social. www.youtube.com/watch?v=8TwP...

Now the thumbnail has a very recent pop culture reference to attract the youth. youtu.be/9Jwrw-oSUFU

In the latest episode, I relate Bostrom's simulation argument to debates over the causal faithfulness condition. I get into interpretations of probability, why "coincidental"≠"improbable", and what it means to assign a probability to a causal parameter. youtu.be/9Jwrw-oSUFU

"At 76.4 percent, it's the lowest turnout for a parliamentary election since 2012"

I can confirm the German accent -- no deep fakes here. Perfect color-matching would violate the positivity assumption, which makes me nervous...

Today, three adorable missing data DAGs. O* is missing when Rₒ masks it and O otherwise. Now, can we d-separate Rₒ and O? 12/10 Ten for the graphs, plus two for casually flipping "causal inference is a missing data problem" From Mohan & Pearl (2021) "Graphical Models for Processing Missing Data"

Hello Bluesky! We rate DAGs. Some are great. Some are... not so great. But we rate them all. Let's start with a famous powerpoint hairball a.k.a. "the Afghanisdag", presented to Gen. Stanley A. McChrystal around 2010. His own rating? 1/10 "When we understand that slide, we'll have won the war"

I've now posted an episode on the causal faithfulness condition. Faithfulness says much more than just that causal paths don't cancel, and raises foundational questions about causation, probability, and underdetermination. youtu.be/k8p0OS-8u2Q

A sneak peak for next week's episode: Does Causation Imply Correlation? youtu.be/5ocnRfA3PX8

I'm excited to post this new episode on the causal Markov condition! youtu.be/eQNfpaAURdU?...

Very few directors could pull off having these 45 second in the middle of a movie in a way that's both unrelated to any other plot point, but also integrated into its whole world in a way that adds to it. www.youtube.com/watch?v=AW-B...

Aydin Mohsehi @center4philsci.bsky.social. "A Bayesian Reduction of Causation in Causal Models". The part about reduction was less original than I anticipated, but I really liked the link to exchangeability and how it links causation to Bayesian probabilities. www.youtube.com/watch?v=3_PT...

What was the word count of the longest referee report you've ever written?

Wait. Is The Martian just Saving Private Ryan in space? How many people does it take to save Matt Damon?

I just posted a video on the Front-Door criterion, which enables you to identify an effect using mechanisms. This video is more technical than the others, but really illustrates how causal graphs create a bridge between causes and probabilities. youtu.be/xRFOj71FPU0

Very exciting. "Frugality" is a term I coined because "sparsest Markov permutation" was a mouthful, and it I wanted to explicitly contrast it with "Faithfulness". But up to now it's been too computationally intensive to implement at scale. www.sciencedirect.com/science/arti...

Has anyone else watched Zoolander since watching Severance and picked up on the stylistic similarities?

Thinking about Mechanisms is the top cited article in @philscijournal.bsky.social (est. 1935). Their website says 1847 citations, Scholar says 4032. Says a lot about philosophy citations compared to other fields. What was the most cited in @thebjps.bsky.social? www.cambridge.org/core/journal...

After a bit of a summer pause, I'm back to making episodes. In this episode, I explain the notion of confounding, and clarify why confounders should not be thought of as alternate explanations of an observed effect. youtu.be/kAgS7cltBhM

In DAGs, confounders are variables that bias an effect measurement (i.e. they hinder identification). I regularly see a vaguer understanding of confounding as alternate explanations of an effect (e.g. I thought X caused Y, but in fact Z does). Does anyone have good examples of the latter usage?

This clip is the part of the last episode that has so far generated the most discussion. youtube.com/clip/Ugkxeuv...

An Episode in honor of Independence Day... youtu.be/7h6tnV1mEEE?...

I decided to make an introductory episode, to provide some orientation for the later episodes. In it, I introduce DAGs and explain how causation is a strategy for managing complexity. I spent a lot of time improving the sound quality, so I hope that paid off. www.youtube.com/watch?v=7h6t...

Jupyter Notebooks: Stop trying to make fetch happen

How can a drug raise the chance of recovery in a population, while lowering the chance in both men and women? The answer is Simpson's paradox, which still confuses people. Here I clarify the paradox, w/ examples involving COVID-19, and testing police discrimination. www.youtube.com/watch?v=MMnA...

Coming next week on Causal Foundations: Simpson's Paradox! Subscribe to my YouTube channel in order to not miss episodes when they're released. www.youtube.com/@dagophile

What is common knowledge in your field but shocks outsiders? “Falsifiability” does not demarcate science from non-science.

Upcoming episodes (~1/week): - Simpson’s paradox - confounding (again) - causal Markov condition - collider bias - the manipulation postulate - faithfulness - varieties of interventions - causal parameters - mediation And much more. youtu.be/FABdstSaUA4?...

Illuminating the process of producing rigorous science by selecting on the dependent variable

Episode 3 "Causation Without Interventions" is now up. In it, I discuss what the most underrated concept in causal inference: identifiability. It's what enables one to measure causal effects from data. I'm still figuring out the AV side of things. Feedback welcome! youtu.be/FABdstSaUA4?...

Well, y'all wouldn't want to disagree with Huw! Episode 3: "Causation without interventions" premieres today at noon EST Prior episodes: 1: Correlation does imply causation 2: d-Separation without tears www.youtube.com/@dagophile