<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Compere — Blog</title><description>Notes on pairwise comparisons, UCB pair selection, and Elo ratings.</description><link>https://compere.skelfresearch.com/</link><language>en-us</language><item><title>Stopping rules: when have you compared enough?</title><link>https://compere.skelfresearch.com/blog/stopping-rules/</link><guid isPermaLink="true">https://compere.skelfresearch.com/blog/stopping-rules/</guid><description>There is no universal answer, but there are three honest tests. We walk through each, with the queries you can run against the compere API.</description><pubDate>Thu, 28 May 2026 00:00:00 GMT</pubDate></item><item><title>Reading the Elo output without lying</title><link>https://compere.skelfresearch.com/blog/reading-elo-output-without-lying/</link><guid isPermaLink="true">https://compere.skelfresearch.com/blog/reading-elo-output-without-lying/</guid><description>Compere ships Elo, not Bradley-Terry. The numbers it produces are easy to misread. A field guide to what an Elo gap actually means.</description><pubDate>Wed, 20 May 2026 00:00:00 GMT</pubDate></item><item><title>Why 50 pairwise votes beat 500 ratings</title><link>https://compere.skelfresearch.com/blog/fifty-pairwise-votes-beat-five-hundred-ratings/</link><guid isPermaLink="true">https://compere.skelfresearch.com/blog/fifty-pairwise-votes-beat-five-hundred-ratings/</guid><description>Rating scales drift, anchor, and lie. Pairwise comparisons survive all three. Here is the math we use, and where it actually breaks down.</description><pubDate>Tue, 12 May 2026 00:00:00 GMT</pubDate></item></channel></rss>