Some notes:
- Axelgaard didn't "win" because he mentioned ~27 riders per race, I only used his first 10.
- When only using 6 riders per prediction, the top 3 stay the same, but some predictors move up a bit.
- I haven't used all predictors for each stage, as you can see in the "Races Processed" column. I plan to be more consistent in the future.
I think the system works, but I'm definitely open to suggestions and improvements.
I also plan to post the rankings during the Tour de France (~weekly), and the upcoming website will show rankings for each specific stage as well.