This doesn't follow. What if their sample size was 100,000 conversions? | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		stavros on Feb 24, 2014 \| parent \| context \| favorite \| on: Most Winning A/B Test Results are Illusory [pdf] This doesn't follow. What if their sample size was 100,000 conversions?

rubiquity on Feb 24, 2014 [–]

Did you even read the article? The third point is "regression to the mean."

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact