Hacker News Reader

Probably because the votes weren't all that suspicious: two semi-regular HN readers who upvote a story in quick succession, but certainly not simultaneously - I have no idea how you'd like to "catch" those (or even whether that would be all that desirable).

BTW, have you ever thought about making other people's "saved" views (i.e. their upvotes) public (maybe w/ opt-in)? It would be an awesome data set for collaborative filtering ...

The submission and the first two upvotes were simultaneous at the 1 minute resolution we use for those things, and that is usually a sign of something fishy. Usually there's something else fishy too, and it's the other things we currently look for.

Glad to hear you're actively doing something about this. It's been discussed a lot here, and it seems to be a real problem.

I always ask my friends to upvote my articles. What are friends for?

As an aside, I'd rather focus on having more people exposed to fresh stories; that'd make "cabal upvoting" much tougher.