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

I agree with your first sentence for sure. Another way to implement it would be to avoid showing new questions with no upvotes to many users--only show them to a small sample of users and see if in the first ten minutes the question gets voted up or down to gauge its quality. That way we'd limit the "shotgun" effect of sending crappy questions to thousands of people all at once (we sometimes see 6 or 8 downvotes on a single question before it gets closed or deleted, which implies a ton of people saw it despite that it was crap).


I like it! Would be sorta like A/B testing the questions...


This pattern is so common I'm not sure why it's not implemented more often. It's just like feature flags with incremental rollout except for content.




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

Search: