Peer Ranking Algorithms
Peer ranking algorithms allow peers in a field to rank each other. This provides a quantitative score of a person’s expertise in this field. This is meant to provide an “expert consensus” which can be compared to the default
Popular Consensus algorithm.
- ”Chicken and egg problem” A two step algorithm is used to bootstrap around the chicken and egg problem. The mathematically minor (see #2 below) first pass democratically determines who the people think are the experts. The mathematically much more rewarded (see #2) second pass totals the rankings of these peers to come up with each expert's canonized peer rank.
- “Each pirate only ever votes for his self” problem, when a popular expert votes for at least one other person, besides themselves, their vote is multiplied by 5 resulting in a significant penalty for only voting for “his self”.
For more information see the actual open source code being used in the
Algorithm.php file in github.