In this paper we present an axiomatic analysis of several ranking methods for tournaments. We find that two of them exhibit a very good behaviour with respect to the set of properties under consideration. One of them is the maximum likelihood ranking, the most common method in statistics and psychology. The other one is a new ranking method introduced in this paper: recursive Buchholz. One of the most widely studied methods in social choice, the fair bets ranking, also performs quite well, but fails to satisfy some arguably important properties.
|Place of Publication||Tilburg|
|Number of pages||23|
|Publication status||Published - 2011|
|Name||CentER Discussion Paper|
- paired comparisons
- fair bets
- maximum likelihood