Every ranking system is at risk for gaming, and the TechMeme Leaderboard is no different, as Dave Winer points out regarding today’s pile on to Jason Calacanis Web 3.0 meme. (Ironically, Jason himself anticipated the problem when the Leaderboard debuted.) This was a problem before the Leaderboard, but providing a lasting reward to this behavior certainly runs the risk of exacerbating it.
So here are some suggestions for Gabe to callibrate the TechMeme Leaderboard so that there is less incentive to game it. It’s possible that some or all of these are already included in the algorithm, but I’m throwing them out in public to suggest that the right disincentives can reduce the game of a ranking system like the Leaderboard.
- More weight to the source of the first headline, and less to subsequent subsidiary headlines
- More weight to items published earlier, i.e. less incentive to add redundant contributions late in the game
- Less weight to sources that do nothing but latch on to existing headlines and never produce their own original headlines
- Less weight to all participants in an attention grabbing pile on — objective markers might include a lot of convoluted cross-linking
There’s an opportunity for the TechMeme Leaderboard to be an interesting experiment in using game theory to optimize a dynamic, networked ranking system — Gabe has always been very savvy in revealing some of the “rules” of TechMeme while keeping others hidden.
Eating my own dog food, of course, this post should contribute little or nothing to my position on the TechMeme Leaderboard.