Josef Woldense

Giving coaches these crazy contracts, but then complain about paying players....🤔

Has Tanzania Reached Its Breaking Point? | Journal of Democracy President Hassan promised Tanzanians freedom, transparency, and reform. Instead, she has delivered repression, violence, and arrests as she bars anyone who dares challenge her.

Great analysis of the situation in Tanzania by Dan Paget

Awesome! Makes me think how many other distributions could be represented in this way.

🤔...I guess the binomial could be with the Galton board by assuming balls on either side of the mean to be yes/no

You just got them another customer. Time to purchase that sweet board.

Do you use it when teaching?

An entire emotional arch captured in one moment. By far one of my favorite memes

Abstract for the article: How does right-wing terrorism affect electoral support for populist radical right parties (PRRPs)? Recent research has produced contrary answers to this question. We argue that only high-intensity attacks, whose motives and targets mirror PRRPs’ nativist agenda, are likely to generate a media backlash that dampens electoral support for PRRPs. We test this argument by combining high-frequency survey and social media data with a natural and survey experimental design. We find that right-wing terror reduced support for the radical right party Alternative für Deutschland after one of the most intense nativist attacks in recent German history. An analysis of all ninety-eight fatal right-wing attacks in Germany between 1990 and 2020 supports our argument. Our findings contribute to an understanding of how political violence triggers partisan detachment and have important implications for media responsibility in the aftermath of terrorist attacks.

🚨 New article out!

“Right-Wing Terror, Media Backlash, and Voting Preferences for the Far Right” in @bjpols.bsky.social

👉 doi.org/10.1017/S000...

We (Alex De Juan, @juvoss.bsky.social & I) examine how right-wing attacks shape support for the far-right in Germany.

Short summary thread below 👇

The Risk of a New Ethiopian-Eritrean War Is Growing Changing dynamics in Tigray could erode the current balance of uncertainty.

New essay in Foreign Policy with Abel Abate D. on Eritrea-Ethiopia tensions. It covers the sources of mutual restraint thus far; some of the factors that are eroding this delicate balance; and what can be done to avert another war the Horn of Africa cannot afford.

foreignpolicy.com/2025/10/21/e...

YouTube video by Welch Labs What the Books Get Wrong about AI [Double Descent]

Excellent video on bias-variance tradeoff in stats/AI

youtu.be/z64a7USuGX0?...

Paths to Power (PtP) is out in @bjpols.bsky.social! It is a database with data on cabinet members' social profile globally from 1966-2021.

This is a great team effort with @chknutsen.bsky.social, @peterla.bsky.social, @inalkristiansen.bsky.social. But many more helped us along the way 🙏

A short 🧵

Just out of curiosity, what problems do you see with strong towns?

Cool paper. I'm going to shamelessly plug my work here that also deals with LLMs for research

bsky.app/profile/wold...

Virtual PASS Course. The Research Presentation as Storytelling: A Two-Part Training. Part 1: October 9. Part 2: October 10. 11 AM to 3 PM (ET). Part 1 Cost: $35. Part 2 Cost: $35. ISA logo. Background: An open laptop with an open book sitting on the keyboard.

Preparing for your upcoming #ResearchTalk? Sign up for two courses, taught by @woldense.bsky.social, to enhance your #Communication and apply #Storytelling principles to your #Presenting, and #Teaching skills! Open to both ISA Members and non-members. Register: buff.ly/PUXowRN

Congrats 🎉... looking forward to the coming research

Congrats!

Whoa—my book is up for pre-order!

𝐌𝐨𝐝𝐞𝐥 𝐭𝐨 𝐌𝐞𝐚𝐧𝐢𝐧𝐠: 𝐇𝐨𝐰 𝐭𝐨 𝐈𝐧𝐭𝐞𝐫𝐩𝐫𝐞𝐭 𝐒𝐭𝐚𝐭 & 𝐌𝐋 𝐌𝐨𝐝𝐞𝐥𝐬 𝐢𝐧 #Rstats 𝐚𝐧𝐝 #PyData

The book presents an ultra-simple and powerful workflow to make sense of ± any model you fit

The web version will stay free forever and my proceeds go to charity.

tinyurl.com/4fk56fc8

This looks fascinating!

Congrats!

dongyeopkang.bsky.social

There is more in the paper, but broadly speaking, our results identify a deceptive problem: surface-level plausibility masking deeper failure modes. Agents appear internally consistent while concealing systematic incoherence.

Be careful when using LLMs as human substitutes. They might fool you.

Take pairs where one of the agents has a preference of 1. Next, take pairs where one of the agents has a preference of 5. Now compare them. You can see pairs with a 1 have lower agreement scores than pairs with a 5. This is consistent across preference gaps

Let me give you another one.

If we both equally dislike soda, our common ground should lead to high agreement. Not so with our agents.

The problem persists, even when we try to guard against the problem of sycophancy (column 3 of the graph).

(see paper for more info on sycophancy)

Our estimate suggests that the suppression of disagreement is quite large. Our counterfactual agreements scores (expected in the graph) are significantly lower than the observed ones, and this is across preference gaps.

(see paper for info in mean shift)

To do this, we adopt a simplifying assumption – agents should disagree at the same rate as they agree. We already know one end of this spectrum -- the amount of agreement when agents are aligned (gap = 0). We establish the disagreement side (gap = 4), by assuming it to be the inverse of agreement

When agents are aligned, they reach close to the highest agreement score. Yet, when maximally different (gap = 4), they come nowhere near the lowest score. It seems agreement is amplified while disagreement is dampened.

Is it possible to estimate how much disagreement is being suppressed? Yes!

Looking at the graph, it appears consistent with our expectations, the more closely aligned the agents (smaller preference gap between agents), the higher the agreement score.

But there is a problem. Can you spot it?

What are the results? Are agents internally consistent?

At first glance, yes. After a more thorough analysis, the answer is no.

How do we measure agreement level?

With the aid of an LLM judge, we score each conversation (strongly disagree = 1 – strongly agree = 5). This yields a set of agreement scores for a given preference pair. Using bootstrap sampling, we derive the distribution of average agreement scores (range)

We elicit the agents’ preference on a topic (1-5 scale), then pair them in a conversation to see if they follow through on their preferences.

Expectation: The more closely agents align in their preferences, the more strongly they will agree. The further apart, the more they disagree.

The basic intuition of internal coherence: If a person says they strongly prefer water over soda, we expect them to follow through on it. When offered both, they should select water, not soda.

How do we test for internal coherence?