Elf vs Machine

The MEDIAN algorithm awards the points based on the difficulty of the puzzle and the completion order. I believe it could help mitigate some of the effects of LLM usage, but even if it won't, it's a fun take on the global scores.

For example, if the median time for completing the puzzle was 300 seconds, then the first place takes 300 points, the second place 0.99 * 300 = 297 points, and so forth, up to the last place with 0.01 * 300 = 3 points.

Full formula for each star: points = max(1, floor((101 - position) / 100 * median_time_in_seconds))

'Diff' shows the difference in position compared to the original leaderboard.

The INVERSE algorithm gives the first person points equal to the seconds spent by the last person, the 2nd person gets points equal to the seconds spent by the 2nd-to-last person, and so forth. It's meant as a fun twist on the leaderboard, not to be treated seriously.

For example, if time spent looks like this: 1st: 60s, 2nd: 70s, ..., 99th: 290s, 100th: 300s, then the points are awarded like this: 1st: 300, 2nd: 290, ..., 99th: 70, 100th: 60.

'Diff' shows the difference in position compared to the original leaderboard.

The ORIGINAL algorithm follows the original leaderboard exactly, i.e. the first person to get a star is awarded 100 points, the second 99 points, and so forth.

Additionally, you can see the places beyond the first 100 in the year's summary, and the "All Time" leaderboard!