For an individual, evaluating 50 items can be a lengthy and monotonous task. However, for a collective, it becomes significantly simpler. Instead of having each participant provide a complete list, we can divide the rankings into manageable segments, then merge them after gathering sufficient votes. Here, we employ a modified ELO algorithm to achieve that.
With ELO, each item starts with a baseline score, which is then adjusted in every paired contest. The score change depends on the rank of the opposing item in the matchup; defeating the highest-ranked item yields a greater reward than defeating the lowest, and similarly, the penalties for a loss are more severe against top contenders. (ELO was initially created for chess, where the victory conditions are more stringent than ‘user preference,’ so we’ve adjusted the conventional approach to lessen the impact of significant upsets.)
This implies that every time you select between any two items on our Apple Top 50 list, you are contributing to the broader community rankings. Thank you!
CREDITS
- Design: Kristen Radtke
- Engineer: Graham MacAree
- Editors: Nilay Patel, David Pierce, Meredith Haggerty
- Concept: Victoria Barrios, Travis Larchuk
- Copy edit: Kallie Plagge
