teaching nhst always makes me glad to be bayesian
November 8, 2026 at 2:42 PM
teaching nhst always makes me glad to be bayesian
if they were actually subjective utility maximizers strapped to a bayesian updating system then we'd be having a completely different conversation... but they're not!
November 8, 2026 at 8:16 PM
if they were actually subjective utility maximizers strapped to a bayesian updating system then we'd be having a completely different conversation... but they're not!
Hey, the Bayesian bootstrap is pretty neat!
November 10, 2026 at 12:12 AM
Hey, the Bayesian bootstrap is pretty neat!
Adding the word Bayesian to a procedure that is otherwise not particularly Bayesian still substantially improves the procedure.
For estimating standard errors & confidence intervals, the Bayesian bootstrap (where you assign random weights to units) is better than the regular bootstrap (where you randomly resample with replacement)
pubs.aeaweb.org/doi/pdfplus/...
pubs.aeaweb.org/doi/pdfplus/...
November 7, 2026 at 4:13 AM
Adding the word Bayesian to a procedure that is otherwise not particularly Bayesian still substantially improves the procedure.
In Praise of the Bayesian Bootstrap! Insightful + fun new JEP from Isaiah, Sarah, and Jesse
t.co/gIXDonDTo3
t.co/gIXDonDTo3
November 6, 2026 at 5:18 PM
In Praise of the Bayesian Bootstrap! Insightful + fun new JEP from Isaiah, Sarah, and Jesse
t.co/gIXDonDTo3
t.co/gIXDonDTo3
Adding Bayesian Bootstrap to my empirical methods class next semester.
Very useful.
Very useful.
In Praise of the Bayesian Bootstrap! Insightful + fun new JEP from Isaiah, Sarah, and Jesse
t.co/gIXDonDTo3
t.co/gIXDonDTo3
November 7, 2026 at 1:56 PM
Adding Bayesian Bootstrap to my empirical methods class next semester.
Very useful.
Very useful.
She loved her first Bayesian book so much that she was trying to hold it and eat it all the time.
November 8, 2026 at 12:04 AM
She loved her first Bayesian book so much that she was trying to hold it and eat it all the time.
You know you can also just be Bayesian
November 7, 2026 at 3:16 AM
You know you can also just be Bayesian
He's really taking on naive Bayesian updating, and he has a point.
On the verge of winning Mississippi, Scott Colom points a finger at the Cook Political Report and other election prognosticators and says they hamstrung candidates like him and Louisiana’s Jamie Davis by rating their races unwinnable
November 10, 2026 at 12:11 AM
He's really taking on naive Bayesian updating, and he has a point.
unfortunately, I like eyebrows, em-dashes, purple, and hierarchical bayesian models
November 6, 2026 at 11:26 PM
unfortunately, I like eyebrows, em-dashes, purple, and hierarchical bayesian models
For estimating standard errors & confidence intervals, the Bayesian bootstrap (where you assign random weights to units) is better than the regular bootstrap (where you randomly resample with replacement)
pubs.aeaweb.org/doi/pdfplus/...
pubs.aeaweb.org/doi/pdfplus/...
November 6, 2026 at 10:40 PM
For estimating standard errors & confidence intervals, the Bayesian bootstrap (where you assign random weights to units) is better than the regular bootstrap (where you randomly resample with replacement)
pubs.aeaweb.org/doi/pdfplus/...
pubs.aeaweb.org/doi/pdfplus/...
It is an abuse of Bayesian statistics, and it is pseudo-statistics, thats what it is.
These Bayesian models are kind of clever in that they allow you to make predictions of future events with some probability. But you must update your model in light of new information. But Shor won‘t.
These Bayesian models are kind of clever in that they allow you to make predictions of future events with some probability. But you must update your model in light of new information. But Shor won‘t.
November 6, 2026 at 3:39 PM
It is an abuse of Bayesian statistics, and it is pseudo-statistics, thats what it is.
These Bayesian models are kind of clever in that they allow you to make predictions of future events with some probability. But you must update your model in light of new information. But Shor won‘t.
These Bayesian models are kind of clever in that they allow you to make predictions of future events with some probability. But you must update your model in light of new information. But Shor won‘t.
Gotta admit it's funny how Nate Silver's data-based empirical Bayesian analysis is "Ossoff just won his election by a lot and he's a looker."
Whoa, incredible, no way the rest of us could've come up with that kind of trenchant commentary. It helps to be popular and hot? Huh. Wow.
Whoa, incredible, no way the rest of us could've come up with that kind of trenchant commentary. It helps to be popular and hot? Huh. Wow.
November 7, 2026 at 9:21 PM
Gotta admit it's funny how Nate Silver's data-based empirical Bayesian analysis is "Ossoff just won his election by a lot and he's a looker."
Whoa, incredible, no way the rest of us could've come up with that kind of trenchant commentary. It helps to be popular and hot? Huh. Wow.
Whoa, incredible, no way the rest of us could've come up with that kind of trenchant commentary. It helps to be popular and hot? Huh. Wow.
pumpkin spice? it's almost mint mocha season bro you gotta update your bayesian priors
November 9, 2026 at 4:48 PM
pumpkin spice? it's almost mint mocha season bro you gotta update your bayesian priors
Yudkowsky's writing is all about powerful bayesian calculators who have internalized logical assumptions about all of ethics, morality and causality and so they become so powerful that they're able to reason about the future and the past simultaneously.
November 8, 2026 at 7:26 PM
Yudkowsky's writing is all about powerful bayesian calculators who have internalized logical assumptions about all of ethics, morality and causality and so they become so powerful that they're able to reason about the future and the past simultaneously.
this sounds like the nerdiest joke ever
-When is bayesian updating difficult?
-when the posterior is highly nonlinear. I'm here all week, try the veal.
-When is bayesian updating difficult?
-when the posterior is highly nonlinear. I'm here all week, try the veal.
When is Bayesian updating difficult? When the posterior is highly nonlinear: small changes in signal accuracy swing it sharply. That's where people err most. Timing matters too: in low-difficulty tasks, seeing signals at once eliminates base-rate neglect.
buff.ly/x4sUNCi
buff.ly/x4sUNCi
November 5, 2026 at 2:04 PM
this sounds like the nerdiest joke ever
-When is bayesian updating difficult?
-when the posterior is highly nonlinear. I'm here all week, try the veal.
-When is bayesian updating difficult?
-when the posterior is highly nonlinear. I'm here all week, try the veal.
Me, teaching geoscience graduate students about the wonderful world of Bayesian statistics:
paleobiologists 🤝 geophysicists
paleobiologists 🤝 geophysicists
November 4, 2026 at 4:36 PM
Me, teaching geoscience graduate students about the wonderful world of Bayesian statistics:
paleobiologists 🤝 geophysicists
paleobiologists 🤝 geophysicists
Bayesian phylogenetic inference accounting for recombination in whole bacterial genomes https://www.biorxiv.org/content/10.64898/2026.11.04.764181v1
November 5, 2026 at 1:17 PM
Bayesian phylogenetic inference accounting for recombination in whole bacterial genomes https://www.biorxiv.org/content/10.64898/2026.11.04.764181v1
Empirical Antifungal Treatment of Critically Ill Patients With Influenza-Associated Acute Respiratory Distress Syndrome: A Bayesian Perspective
🔁 Repost
https://t.co/xIS84kOkFS
🔁 Repost
https://t.co/xIS84kOkFS
Empirical Antifungal Treatment of Critically Ill Patients With Influenza-Associated Acute Respiratory Distress Syndrome: A Bayesian Perspective
t.co
November 8, 2026 at 5:12 PM
Empirical Antifungal Treatment of Critically Ill Patients With Influenza-Associated Acute Respiratory Distress Syndrome: A Bayesian Perspective
🔁 Repost
https://t.co/xIS84kOkFS
🔁 Repost
https://t.co/xIS84kOkFS
I used this approach in my job talk and it nicely bridges frequentist and Bayesian so even the hard core Bayesian in the room were cool with it lol
Also love how this app looks
Also love how this app looks
November 8, 2026 at 5:25 PM
I used this approach in my job talk and it nicely bridges frequentist and Bayesian so even the hard core Bayesian in the room were cool with it lol
Also love how this app looks
Also love how this app looks
Mathematical models of probability are transitioning from tools for idealized logic to powerful frameworks for explaining how human minds process…
Bayesianism by Michael Rescorla
https://doi.org/10.21428/e2759450.c0cec761
#CognitiveScience
Bayesianism by Michael Rescorla
https://doi.org/10.21428/e2759450.c0cec761
#CognitiveScience
November 6, 2026 at 7:54 PM
Mathematical models of probability are transitioning from tools for idealized logic to powerful frameworks for explaining how human minds process…
Bayesianism by Michael Rescorla
https://doi.org/10.21428/e2759450.c0cec761
#CognitiveScience
Bayesianism by Michael Rescorla
https://doi.org/10.21428/e2759450.c0cec761
#CognitiveScience
here's the one i was thinking about
November 8, 2026 at 4:45 AM
here's the one i was thinking about
*the master hit the bayesian on the head w/a stick after this and well it didn't make em reach enlightenment but fuck we got a good bit of peace for a while*
November 8, 2026 at 4:47 PM
*the master hit the bayesian on the head w/a stick after this and well it didn't make em reach enlightenment but fuck we got a good bit of peace for a while*
I mean, I'm not a Bayesian at all aside from being a fan of the Bayesian toolkit in certain specific, well-defined domains (a lot of statistics, parts of cognitive science, etc.)
November 8, 2026 at 5:10 PM
I mean, I'm not a Bayesian at all aside from being a fan of the Bayesian toolkit in certain specific, well-defined domains (a lot of statistics, parts of cognitive science, etc.)
As someone who has built a career doing Bayesian decision theory at startups, I don't often wish a startup would fail. Still, man, I'd throw a party if Polymarket, Kalshi, DraftKings, Underdog, etc. went under, because I firmly believe their business model should be a federal crime.
November 8, 2026 at 2:00 AM
As someone who has built a career doing Bayesian decision theory at startups, I don't often wish a startup would fail. Still, man, I'd throw a party if Polymarket, Kalshi, DraftKings, Underdog, etc. went under, because I firmly believe their business model should be a federal crime.