Reproducibility

Our open sourced backend repo

It's impossible to reproduce the exact numbers we get each time we collect data to update the leaderboard -- there is randomness from various network and provider fluctuations. However, for transparency purposes, we open sourced our backend which includes scripts that collect the data displayed on the leaderboard. When run under similar conditions, the script should produce similar numbers on average to those we collect.

If you have further questions about how we collect the metrics or how to reproduce our collection, feel free to refer to our repository! While we do not have the bandwidth to respond to every issue immediately, we welcome feedback from the community :)