Our first post was twenty years ago today.
It was followed by posts on The Electoral College favors voters in small states; Why it’s rational to vote; Bayes and Popper; and Overrepresentation of small states/provinces, and the USA Today effects.
Later that month we had our first post on the Red State/Blue State Paradox, a guest post on statistical issues in modeling social space, and a stab at partial pooling of interactions.
On 27 Oct came one of my favorite early posts, The blessing of dimensionality, and early the next month came Sam Cook’s post on her now-classic work on Bayesian software validation, which we now call SBC, for simulation-based calibration checking, an early post on cross-validation for Bayesian multilevel modeling—a topic on which we made lots of progress in the decades since.
Also in Nov 2004 came our first formulation of the important (to me) concept of institutional decision analysis, my discovery about correlations in before-after data, and a debunking of a claim of possible election fraud in Florida.
Those were the days!
We’ve published around 11,000 posts since then. If you want to read the old ones, just go into the blog archives or subscribe to the StatRetro twitter feed, which spits out links to old posts, in order, every eight hours. It’s currently at 2009. Once it catches up to us—I guess that won’t be for many years—we can restart it, if we’re able to do so.
Thanks so much to you, the commenters, who’ve supplied us with so much insight and feedback over the years. Even when you’re confused, it helps us by revealing flaws in our communication.
And to all you readers out there, keep reading—you’re why we do it!
And thanks to my co-bloggers, listed below in increasing order of number of posts so far:
Paul Buerkner
Rohan Alexander
Josh Miller
Sean Talts
Sander Greenland
Witold Więcek
Yu-Sung Su
Sebastian Weber
Lonni Besançon
Joe Bafumi
Lauren Kennedy
Jon Zelner
Michael Malecki
Jeff Lax
Leo Egidi
Juli Simon Thomas
Yuling Yao
Mike Betancourt
Masanao Yajima
Charles Margossian
Shira Mitchell
Jonah Gabry
Dean Eckles
Lizzie Wolkovich
Eric Novik
Daniel Lee
Sam Cook
Dan Simpson
Keith O’Rourke
Aki Vehtari
Jessica Hullman
Phil Price
Aleks Jakulin
Bob Carpenter
The blog wouldn’t be the same without you. Here’s to another 20 years.
Happy birthday! Thanks for all the great posts
20 years! Thanks for keeping up the blog! I’m constantly amazed at how consistently informative *and* entertaining it is — I visit every day! Thank you!
I’ve been interacting usually daily since around 2005. I really appreciate the blogs ability to bring people with a similar set of interests together to have interesting discussions. That it’s still going strong 20 years later is fabulous.
I second Daniel’s comment. I myself don’t comment much but only because I know so little yet learn (and un-learn) so much.
My quest to understand all this began after a gray-haired Ivy League epidemiologist testified (paraphrasing): “If p.05 then the study refutes causation, the same rule applies if you combine the studies, the Mantel–Haenszel test is the gold standard of causal inference, and, the Sir A.B. Hill causal criteria are how a competent expert makes causal judgments.” The courts largely adopted this scheme, then made ad hoc exceptions, and wound up with a bizarre sort of statistical numerology. I’ve had little luck getting the courts to stop writing stupid things but thanks to your blog I write and say fewer, or at least different, stupid things.
“subscribe to the StatRetro twitter feed, which spits out links to old posts, in order, every eight hours.”
Feels like this is pretty much the speed at which you write blog posts ;)
Happy 20th to the blog.
So it’s still a year until the blog can legally drink. I know someone who can get the blog a fake ID (apparently registered in Iceland) if you want to go out out on the town for beers and Jamaican meat pies.
You’ve built something with many wonderful things to read. Not only that, but you have helped sustain my interest in Bayesian statistics in my mostly classical day-to-day world of econometrics. Much appreciated and thank you for it.
If truth and wisdom are things you can sometimes see
perhaps this blog has helped me in providing a key
that may have unlocked a view that I did not see
and may have contributed to setting something free
If truth and wisdom are things you can sometimes feel
perhaps this blog has helped me in breaking a seal
that may have opened a sense, or made it more real
and may have contributed to letting something heal
Whether this is all true or not may not even matter most
Maybe this blog is merely a mirror that reflects what I post
But it could be that a particular mirror is sometimes needed to reflect a ghost,
or to provide a key, or break a seal, which may all deserve a toast
So here’s to 20 years of this blog, and here’s to the host!
The most useful posts for me are where I ask for help and experts chime in. It’s magic. For example, one of my first posts asked what we should do for inference in Stan (we were considering just doing some hardcore engineering on JAGS, from which we could’ve easily squeezed out an order of magnitude speedup and scalability from Matt’s early experiments on vectorization). Once people suggested HMC, another post asked how we should compute derivatives (autodiff was super obscure when we started Stan in 2010—we hadn’t even heard of it). Most recently, I asked how we should code models in JAX and the replies were super helpful.
The second most helpful posts for me are the ones where Andrew does a complete simulation of a simple fact of statistics. I don’t care if it’s unexpected (for me) behavior of p-values, a trigonometric model of golf putting, or a subtle point about demographics relating to elections. I love those bite sized examples—examples broken down into code are sooo helpful for a CS person like me. I don’t think I could’ve learned stats without the R code in Jennifer and Andrew’s first regression book or without BUGS and more particularly, the three volumes of BUGS examples.