iiinova — where your digital self keeps evolving

Andy Seaton

@aseaton.bsky.social

155 followers 139 following 124 posts

Unemployed convalescent Honorary Fellow at the University of Glasgow

Andy Seaton @aseaton.bsky.social · 24/02/2026

Trying to write about "good statistical practice" and applying statistics to models with immense predictive power is the final boss. I feel like this is a weird place to draw a line but I have to draw it here to be consistent with my opinions on messier applications of stats. #statsky

A screenshot of a paragraph from a piece of writing about statistics that reads:

When the international community of physicists agreed on the statistical analysis they would perform on the data they collected at the Large Hadron Collider, they were doing statistics. But when they agreed what the conditions were for declaring the Higgs boson real or not, they were not really doing good statistics anymore. They were agreeing a position as a scientific community about when and how to leave an uncertain world and enter a new world with one new piece of certainty attached to it. The statistical work was one (small) piece of this decision making process. This community-driven process, viewed as a whole, should not, in my view, be seen as a part of statistics, though I understand that it looks like it could be. I should be really explicit here and emphasise that saying something is not statistics is not the thing same as saying that it is bad science.

040 0

Andy Seaton @aseaton.bsky.social · 10/02/2026

I've been reading Good Statistics by I. J. Good. He's entertaining as a writer. He's also pretty good at letting you know exactly what he thinks in the most unimpeachably polite British way possible. #statsky

150 0

Andy Seaton @aseaton.bsky.social · 07/02/2026

Reading some Montaigne and mostly I enjoy his little melancholic meanderings. Then sometimes he casually drops in shit like this. In the short essay On Idleness

100 0

Andy Seaton @aseaton.bsky.social · 03/01/2026

Playing with Bayesian non-parametrics for distance sampling (preprint out soon). Had a little moment where my feelings about the model were worringly conditional on how many histogram bins I used. These four plots are from the same model fit, just different numbers of bins summarising the data.

171 0

Andy Seaton @aseaton.bsky.social · 17/01/2025

Up to no good again

010 0

Andy Seaton @aseaton.bsky.social · 05/12/2024

Well this is news to me

000 0

Andy Seaton @aseaton.bsky.social · 17/11/2024

The academic grind never stops even on a Sunday. Got to network, get your research out there.

120 0