4 ms·

This is a follow-up to the item about how long submissions stay on the "Newest" page as a function of time of submission: http://news.ycombinator.com/item?id=2

• stcredzero 16y ago
  My hunch is that there is a threshold for social news sites of this general model (voting on items on a "new" page) where the volume of traffic swings the balance of power sharply in favor of upvoting cabals. (Rapid fall of items off the "new" page dilutes the power of the ordinary voting user, but has less effect on the targeted actions of cabals.) After that point, the site's "market" gets severely distorted. Deliberate manipulation by disparate parties all motivated to redirect the maximum number of eyeballs to their site, and this tends to pull excellent sites strongly back towards the average. This would also imply that upvote cabals can manipulate conditions at a given site in their favor simply by submitting more links.
  • RiderOfGiraffes 16y ago
    This is why PG has to have logic in the HN code to detect and punish or lock out voting rings. The Sybil attack [1] is well-known and there are methods to detect it. Harder is when a circle of genuine users all decide to watch each others' posts and upvote them purely on reputation. [1] http://en.wikipedia.org/wiki/Sybil_attack http://en.wikipedia.org/wiki/Sybil_attack
  • nitrogen 16y ago
    Sounds similar to the tactic of submitting fake profiles to dating sites: http://news.ycombinator.com/item?id=2065604 http://news.ycombinator.com/item?id=2065604