Stephen Wild
@stephenjwild.bsky.social
I try to put straight lines through things but usually fail. Try to be Bayesian when I can. Views my own. RT/like != endorsement.
On the reading list
January 4, 2026 at 12:01 AM
On the reading list
Look, the man influenced Rommel. We know because he said so in his biography of Rommel.
Many people do not know that Mearsheimer wrote a whole book about a military thinker who was catastrophically wrong about everything all the time, but then rehabilitated himself anyway
books.google.com/books/about/...
books.google.com/books/about/...
January 3, 2026 at 11:01 PM
Look, the man influenced Rommel. We know because he said so in his biography of Rommel.
Paging the nerds
January 3, 2026 at 10:55 PM
Paging the nerds
Some good suggestions here. Givr me MoAr
Hive mind! I am looking for biography recommendations. I would like something off the beaten path, about a person or topic you find interesting.
All suggestions welcome.
All suggestions welcome.
January 3, 2026 at 8:36 PM
Some good suggestions here. Givr me MoAr
Reposted by Stephen Wild
Histograms are also non-parametric density estimators, with the bin-size as the smoothing parameter. The typical advice is to prefer under-smoothed histograms (aka too many bins than is pretty) because that’s less likely to lie to you.
January 3, 2026 at 5:28 PM
Histograms are also non-parametric density estimators, with the bin-size as the smoothing parameter. The typical advice is to prefer under-smoothed histograms (aka too many bins than is pretty) because that’s less likely to lie to you.
Many such cases 😉
was looking at a paper, got confused about its statistical methods, went down a rabbit hole on Bayesian stats, and now I both have a headache and don't trust frequentist methods
January 3, 2026 at 1:22 PM
Many such cases 😉
Reposted by Stephen Wild
The 1.3.0 version of my {brms} + {tidyverse} translation of Kruschke's "Doing Bayesian data analysis" is up!
solomon.quarto.pub/dbda2
#rstats
1/8
solomon.quarto.pub/dbda2
#rstats
1/8
Doing Bayesian Data Analysis in brms and the tidyverse
solomon.quarto.pub
January 2, 2026 at 5:00 PM
The 1.3.0 version of my {brms} + {tidyverse} translation of Kruschke's "Doing Bayesian data analysis" is up!
solomon.quarto.pub/dbda2
#rstats
1/8
solomon.quarto.pub/dbda2
#rstats
1/8
One time on Twitter my "For You" tab was full of body builders talking about how they made huge gains once they "got on T." Apparently I need to talk to my doctor about it. Also go through cycles of bulking and cutting. And lift. But get on T.
I have no idea why the feed was full of it. And yet.
I have no idea why the feed was full of it. And yet.
January 2, 2026 at 4:17 PM
One time on Twitter my "For You" tab was full of body builders talking about how they made huge gains once they "got on T." Apparently I need to talk to my doctor about it. Also go through cycles of bulking and cutting. And lift. But get on T.
I have no idea why the feed was full of it. And yet.
I have no idea why the feed was full of it. And yet.
Reposted by Stephen Wild
I blogged with #rstats about New Caledonia's nickel exports and how they have collapsed in the past few years, from a combination of global economic conditions, local political economy, and security and civil unrest problems in 2024. freerangestats.info/blog/2026/01...
January 1, 2026 at 10:44 AM
I blogged with #rstats about New Caledonia's nickel exports and how they have collapsed in the past few years, from a combination of global economic conditions, local political economy, and security and civil unrest problems in 2024. freerangestats.info/blog/2026/01...
Finally, a good WWII take on this website
I actually would disagree with this take - though he might not have known it, Hitler has essentially set the conditions to make eventual defeat and regime extinction very likely (and so his great power position very poor) by November 1940 at the latest, even with the success of Fall Gelb/Rot.
From a great power diplomacy perspective, Hitler had a good run until the twin idiocies of invading the Soviet Union and declaring on the US.
His luck might have run out with the invasion of Poland if the invasion of France hadn’t worked. However, he did have a de facto alliance with the Reds.
His luck might have run out with the invasion of Poland if the invasion of France hadn’t worked. However, he did have a de facto alliance with the Reds.
January 2, 2026 at 1:22 PM
Finally, a good WWII take on this website
"Not all commentary invites reply" must have been in the pre-social media days 😉
Good post.
Good post.
January 2, 2026 at 11:53 AM
"Not all commentary invites reply" must have been in the pre-social media days 😉
Good post.
Good post.
🤔
Interesting numbers.
Interesting numbers.
I don't think we should trust computers to do things with numbers anymore
January 1, 2026 at 1:15 PM
🤔
Interesting numbers.
Interesting numbers.
These look nice
Some lecture notes on spatial and temporal modelling, illustrated using R - darrenjw.github.io/spatio-tempo... - #rstats #rspatial #quarto
Spatial and temporal statistics
darrenjw.github.io
January 1, 2026 at 1:01 PM
These look nice
January 1, 2026 at 1:05 AM
Share the resources if you have it.share all the resources.
FWIW, my view is that some of the increase is real, but most if it is probably measurement non-invariance over time.
onlinelibrary.wiley.com/doi/abs/10.1...
FWIW, my view is that some of the increase is real, but most if it is probably measurement non-invariance over time.
onlinelibrary.wiley.com/doi/abs/10.1...
December 31, 2025 at 2:40 PM
Share the resources if you have it.share all the resources.
FWIW, my view is that some of the increase is real, but most if it is probably measurement non-invariance over time.
onlinelibrary.wiley.com/doi/abs/10.1...
FWIW, my view is that some of the increase is real, but most if it is probably measurement non-invariance over time.
onlinelibrary.wiley.com/doi/abs/10.1...
This is a pretty interesting episode
I enjoyed this podcast, which combines a bit of myth busting about Medieval Europe with some sociology of infectious diseases.
youtu.be/nECpdIs_9Cg?...
youtu.be/nECpdIs_9Cg?...
Did Medieval Society Collapse During The Black Death?
YouTube video by Dan Snow's History Hit
youtu.be
December 31, 2025 at 1:36 PM
This is a pretty interesting episode
My predict-next-token machine predicted the next token. It is sentient.
My Bayesian regression also does continual learning of I feed it new data, now let it vote plz
Sure people on here might trash AI rather extremely, but if this is your take on AI, it’s probably better to get off the Internet for a while
December 31, 2025 at 12:05 PM
My predict-next-token machine predicted the next token. It is sentient.
This is a good life philosophy.
The meme asks for a comparative claim which makes sense cos everyone can have some answer, but honestly it just makes me reflect on how little actual impact I have on anything in the grand scheme of things. This is nice! There's peace in irrelevance.
"What was your biggest impact on the Internet?"
I drew the original of this. (I storyboarded and assistant directed on this episode.) (Fry, "Shut up and take my money")
I drew the original of this. (I storyboarded and assistant directed on this episode.) (Fry, "Shut up and take my money")
December 31, 2025 at 11:41 AM
This is a good life philosophy.
My normie friends use it to write letters from Santa. They think it's awesome.
On here it's a fight on how they're destroying everything or changing the world. No one seems happy.
On here it's a fight on how they're destroying everything or changing the world. No one seems happy.
It is illuminating to see how my normie friends use LLMs and how people on here do. And their reactions to LLMs are entirely different than what you see on here
December 31, 2025 at 12:41 AM
My normie friends use it to write letters from Santa. They think it's awesome.
On here it's a fight on how they're destroying everything or changing the world. No one seems happy.
On here it's a fight on how they're destroying everything or changing the world. No one seems happy.
Alan: That's not a purse. That's a satchel!
Chow: It's a purse, okay?
Chow: It's a purse, okay?
December 30, 2025 at 9:34 PM
Alan: That's not a purse. That's a satchel!
Chow: It's a purse, okay?
Chow: It's a purse, okay?
Thr answer (I hope) is it depends on the source of the measurement error.
blog posts: should we adjust for a mismeasured X ?
You know the population distribution for X (e.g. vote choice in 2024).
But you only have a reported X* in your survey.
Should you adjust for it ?
Later today: exploring toy examples to see.
You know the population distribution for X (e.g. vote choice in 2024).
But you only have a reported X* in your survey.
Should you adjust for it ?
Later today: exploring toy examples to see.
December 30, 2025 at 12:45 PM
Thr answer (I hope) is it depends on the source of the measurement error.
It's fine because it is the gimmick of the show. It's like a medical drama having all sorts of rare diseases.
But the characters' motivations and actions need to be consistent and realistic within the gimmick. Sadly this is often not the case. Boooooo.
But the characters' motivations and actions need to be consistent and realistic within the gimmick. Sadly this is often not the case. Boooooo.
The James Bond discussion reminds me of a long Discourse we had in grad school re whether cozy murder mystery shows (think Murder She Wrote or Midsommar Murders or Detective Conan) are meant to literally have the crime rate implied by on screen murders.
another funny thing that came to mind watching basically all the James Bond movies in order
it amuses me how people came to the conclusion that "oh James Bond is a codename, that's why they're all different guys" when I watched them and went. "Ah yeah I know how this works. floating timeline"
it amuses me how people came to the conclusion that "oh James Bond is a codename, that's why they're all different guys" when I watched them and went. "Ah yeah I know how this works. floating timeline"
December 29, 2025 at 1:15 PM
It's fine because it is the gimmick of the show. It's like a medical drama having all sorts of rare diseases.
But the characters' motivations and actions need to be consistent and realistic within the gimmick. Sadly this is often not the case. Boooooo.
But the characters' motivations and actions need to be consistent and realistic within the gimmick. Sadly this is often not the case. Boooooo.
