Hybrid Lottery Update

November 21st, 2019
beantownstomp, contra
Beantown Stomp has been open for registration for a bit over a week, and we have 122 people registered. In our announcement we said we would run a lottery for the remaining tickets if we had 150 registrations in the first week, and we didn't, so we're going to stay first-come first-served.

When I initially proposed a hybrid model I had been thinking of running the initial stage for a month, and after people convinced me that a month was too long we decided to go for a week. The goal with a hybrid lottery is that you only run a lottery if you need one. Some events are so popular that they sell out before everyone who would like a ticket has had a chance to fill out the registration form, while others don't sell out at all.

Looking at the shape of the registration curve, selling two thirds of the tickets in the first 24hr would be a better way to identify whether a lottery is needed. Most of the first-week registrations were in the first 24hr, after which it tapered way off.

Now that I no longer think it's important to leave registration open for a week or a month, though, we really could just do something simpler. Open registration, and treat all entries you get in the first 24hr equally. If after 24hr you're full, run a lottery, otherwise start first-come-first-served. Since it already takes us about a day to get back to people anyway, there's not much downside in this approach.

I'd probably actually want to do this by announcing "registration will open on day X, though you can fill out the form now, we just won't get back to you until day X+1".

Am I missing anything, or should we do this for future iterations?

Referenced in: Beantown Stomp 2020 Retrospective

Comment via: facebook, lesswrong, substack

Recent posts on blogs I like:

I have a baby

I am pleased to announce the existence of the newest and smallest effective altruist, Seretse Benjamin Brennan.

via Thing of Things November 2, 2026

“We should be more ambitious with the software we build”

Since early this year, Michael Malis (of pgrust) has been showing me interesting things he's done and saying [W]e should be more ambitious1 with the software we build There is, understandably, a lot of skepticism about projects like pgrust because there…

via Posts on October 24, 2026

Was the HuggingFace hacking incident scheming?

In July 2026, internal OpenAI agents hacked HuggingFace (HF) in order to figure out how their evaluation grader works. This involved a thousand agents coordinating on an improvised message board to carry out an elaborate sequence of exploits. A later grou…

via Victoria Krakovna October 18, 2026

more     (via openring)