Bayesian Cabaret

by Scott Alexander1 min read27th Apr 200912 comments

25

Art
Personal Blog

I'd heard rumors that some leading Bayesians had achieved rank in the Bardic Conspiracy. But I wasn't aware that every two years, some of the world's top statisticians hold a Bayesian Cabaret, full of songs, dances, and skits about Bayesian probability theory.

...no, really. Really. I think my favorite has got to be this one.

YouTube seems to be full of this stuff, including What A Bayesian World and We Didn't Start The Prior. Be sure to also check out some of the recordings and the Bayesian Songbook.

Eliezer's finished his sequences, and it's a new era for Less Wrong. Let's celebrate...Bayesian style!

12 comments, sorted by Highlighting new comments since Today at 5:19 PM
New Comment
[-][anonymous]8y 2

(In case anyone else is wondering, this comment used to show this picture.)

HAHAHAHAHAHA

The audio on that version of "I'm a believer" is really bad (or perhaps it's just my ears), but it sounds as if they kept the lines "without a trace / of doubt in my mind". FAIL.

Here are the lyrics via The Bayesian Songbook:

V1: I thought inference was just a fairy tale, / Confused by stats and probability, / Frequentist approaches (doo-doot doo-doot) / made no sense to me (doo-doot doo-doot) / Summarizing evidence by p?!

Chorus: Then I saw Tom Bayes — Now I'm a believer, / Without a trace of doubt in my mind, / [I'm a] Bayesian (ooooh) — Oh, I'm a believer — / I couldn't p now if I tried!

V2: I thought likelihood was just the only thing, / Turn the crank and get the MLE, / What's the use of thinking (doo-doot doo-doot) / Disconnect your brain (doo-doot doo-doot) / Play along and minimize the pain...

Chorus: (repeat)

Solo: (keys/guitar)

V3: SAS was out to get me (doo-doot doo-doot) / that's the way it seemed (doo-doot doo-doot) / Fixed effects and forced normality...

Chorus: (repeat 2x w/assorted hollering and out!)

I'm not sure if this is an event to celebrate, I think less posts by Eliezer is a bad thing.

It's...bittersweet. Like a graduation ceremony. It's nice to know you've learned a lot and reached a level where you can go out on your own. It's sad that you've reached the end of what formal instruction will teach you and you're going to have to part ways with your teachers and the college community. And it's scary that you've got to go out into the real world when you still know so little.

Less posts by Eliezer is bad.

Less work on Things Not To Be Discussed before May is much worse.

I once tried to write a take-off called "Damn It Feels Good to be a Bayesian", but I didn't get far.

Anyway, here a link to an extensive collection of math and stats quotes. (Warning: 2,000 quotes, 142 pages.)

For Bayes humor I recommend Guernsey McPearson.

Lyrics to Bayesian Believer: http://math.bu.edu/people/mg/music/believer.txt

For me, "Bayesian Believer" is a funny term. "Belief" has such strong connotations of "faith" for me. I always said that I didn't "believe" in anything - that all was open to doubt.

I wouldn't be at all surprised to hear that Tim Minchin was at least a novice initiate in the Bayesian Conspiracy