Where does this quote come from? “I don’t trust anything until it runs. In fact, I don’t trust anything until it runs twice.”

Google the above quote and you’ll see it attributed to me. And it does sound like the kind of thing I might say. But I don’t recall actually having ever said it, and a search does not turn up any original appearance of this saying.

Does anyone know the original source? Is it something I said in a talk and have forgotten? Or maybe it was misattributed to me, but then who said it, and when?

P.S. I was the one who said it! Back in 2017. Commenter Ahm reports:

I put it in my “Ideas to Inflict Upon Co-authors and PhD Students” notebook after hearing it in the talk “Theoretical Statistics is the Theory of Applied Statistics: How to Think About What We Do” at the 2017 New York R Conference.

The talk is on Youtube (timestamp 27:51). Link here.

The full-ish quote is “Computer programing is basically the last bastion of rigor in the modern world. I don’t trust anything until it runs. In fact, until it runs twice on the computer.”

15 thoughts on “Where does this quote come from? “I don’t trust anything until it runs. In fact, I don’t trust anything until it runs twice.”

  1. If you search with the option -gelman, even for the first sentence, you get only this post on multiple search engines.

    I am afraid that your options are:

    1. living with this (hey, you can still say it ex post, and no one will question you),

    2. hire an RA specialized in philology to track this down.

  2. I’ve attributed stuff like this to you before.

    Rhat has convinced me that, by default, interfaces to iterative algorithms should run and compare multiple solves. Having stuff baked into software is so handy. A lot of people could have said something like that, but the Rhat + MCMC story is nice as far as stories go.

    That doesn’t answer the question, but maybe someone else just used you as a drop in expert.

    • Could be. Sometimes an idea will come up in casual conversation and then when I write the idea down, I’ll attribute it to the person I was talking with, even if it might not be that person’s exact words. So only fair that it could happen to me too.

  3. Also in this theme I attribute a post “The importance of ‘bumblers’ and ‘pointers’ in science, and the division of labor” from November 28th, 2022 to you. I have it open on my phone, but it’s not showing with my laptop.

  4. Just revel in the glory of having a neat quotation attributed to you!!! About the only thing that was ever attributed to me was “I am a slave to fashion” which appeared in the Washington Post for an article on men who were wearing suits on the hottest day of the year (our office was right around the corner and I was headed out to a meeting when cornered by the reporter and photographer). At least in my case there is a published attribution :-) and my daughters gifted me a t-shirt with the quotation on it. Hey, that’s an idea; get some tees made up and gift them to grad students and research fellows in your department.

  5. It’s a real quote!

    I put it in my “Ideas to Inflict Upon Co-authors and PhD Students” notebook after hearing it in the talk “Theoretical Statistics is the Theory of Applied Statistics: How to Think About What We Do” at the 2017 New York R Conference.

    The talk is on Youtube (timestamp 27:51). Link here: https://www.youtube.com/watch?v=cuE9eHSbiNI&t=1671s

    The full-ish quote is “Computer programing is basically the last bastion of rigor in the modern world. I don’t trust anything until it runs. In fact, until it runs twice on the computer.”

Leave a Reply

Your email address will not be published. Required fields are marked *