Methodology

MoralMatchups uses the Elo rating system — the same algorithm used in chess — to rank billionaires by perceived positive impact.

Every person starts with a rating of 1500. When you vote, the winner gains rating points from the loser. The amount exchanged depends on how surprising the result is: beating a higher-rated opponent earns more points than beating a lower-rated one.

New participants have a K-factor of 32 (provisional, under 100 votes) — ratings move quickly early on to find their true level. After 100 votes, the K-factor drops to 16 (established), making ratings more stable.

Each matchup is selected using a weighted strategy designed to produce informative comparisons:

• 60% — similar-rating pairs (tightest contests)
• 25% — under-voted pairs (ensure everyone gets coverage)
• 15% — most divisive pairs (surface controversial comparisons)

Your browser tracks which pairs you have already seen this session. When you have seen every pair, the list resets so you can vote again.

Divisiveness measures how split the community is on a person — a score of 100% means votes are perfectly split 50/50 in their matchups; 0% means near-consensus.

Formula: divisiveness = 1 − 2 × |p − 0.5|, averaged across all matchups weighted by vote count. Recalculated nightly.

Skipping a matchup records the skip but does not change either person's Elo rating. Skips are counted and displayed on profile pages.

To keep ratings trustworthy:

• Each session can only cast one vote per pair
• Votes require a time-limited HMAC challenge token issued at matchup fetch time
• Burst detection blocks more than 10 votes per 60 seconds per session

Biographical information and net worth figures are sourced from publicly available Wikipedia articles and Forbes estimates. Net worth figures are approximate and dated — see each profile for the reference date. This site does not make claims about the moral worth of any individual.