This paper proposes an alternative solution to the standard approach used in the aggregation problems that arise when creating rankings.

This approach has advantages and disadvantages with respect to more traditional methods, but it is undoubtedly attractive because it allows to accommodate in an alternative way issues such as the normalization of individual indicators or the treatment of missing information (missing values). In addition, it is designed in a way in which it does not allow compensation between individual starting indicators. Although, in specialized literature, these methods are attributed to Social Choice Theory, this approach addresses the mathematical structure of the problem directly, making the connection to this theory afterwards.

The suggested techniques are demonstrated with an application to the creation of university rankings using the U-Ranking database.