Our association (l’APRES) organizes discussions every Sunday either on the internet, in real life, or both. Every Monday, a week in advance, we randomly pick someone to choose the next discussion’s theme. We thus needed a cheap and secure way to carry out this weekly selection. This post demonstrates how this is done.
The first step consists in going to this website. On this page, you need to enter a “seed” number to produce the output. As a seed we use the CAC 40 value at the market opening of that day – Monday. This value is publicly available and easy to determine and verify. Importantly, the same seed produces the same output (i.e., the same selected person) but at the same time two very close by seeds produce different, wholly unrelated, outputs (see why here).
In the demonstration shown in the following images, I employ a few names of people from this blog.
The process cannot be tampered with because no existing method can predict or influence – with sufficient accuracy – financial markets. Moreover, the selection can be reproduced by everybody ad infinitum. In conclusion, we have here a robust and inexpensive way to select randomly.
Filed under: Applications, Sortition |
Small comment, the list ought to be sorted before selection. Tampering can occur “after the fact” by manipulating the order of the list after you input in the seed. So you need to add in
people.sort()
LikeLike
You are right changing the list would change the outcome (the simplest case would be to remove a name from the list). This is exactly why it is crucially important to make this list public before, during and after the vote.
LikeLiked by 1 person