During the 2020 Democratic presidential primary, I ran a contest in which participants predicted the drop out order of the final 14 candidates still running as of January. I will be hosting the contest again for the 2024 Republican primary, with a better submission form, a sleeker website, and much more automation.
index.html, css/style.css, and js/script.js are the files behind the contest leaderboard.
python/dropout-contest-scoring.ipynb is a notebook for computing each participant's score as presidential candidates suspended their campaigns. Yes, I manually re-ran this notebook each time a candidate dropped out...I had a lot to learn. At least it was only 12 times 🤷.
python/heatmap.ipynb is a notebook for generating a heatmap visualization of all the drop out predictions. Reads data from the submission dataset, which is not published in this repository. A first draft for what I later recreated using D3.js:
