I have a bit of a tech-related issue and ask related to #algorithms in gov:
context: I’ve built an application that conducts a Lottery each week - to assign spaces in a plaza for artists to sell at. Based on a number of registrations, results are randomly generated and users are notified.
issue: Many users have reported “unfairness”, “lack of randomness”, and more.
context: overwhelmingly, reports come after a user receives a “bad draw”. I’ve worked with the agency to “increase randomness” in the algorithm, had that code “audited” by public staff, and created a graph that shows a) one User’s results over time, and b) all User’s results over time - and the city is still getting pressured to “do something”
ask: has anybody run into examples around algorithmic randomness in gov? or user experience related to randomness?