A generalization of Thurstone method for considering advantages of certain participants
Keywords:
páros összehasonlítás , Thurstone-módszer, előny figyelembe vétele, maximum likelihood becslésAbstract
In this paper we investigate a paired comparison method which takes into account the advantage of a member of the pair. We use the method of Thurstone on the basis of latent random variables, we allow three options for the decisions. The parameters are estimated with maximum likelihood (ML) method. We present a statement which provides sufficient condition for the existence and uniqueness of the ML estimation in the case of incomplete comparisons. The method is applied to the results of the European Open Chess Team Championship (2011) and our results are compared with the results of other methods.
Paired comparisons are frequently used in cases where the compared objects are evaluated subjectively. We can mention psychology, politics, marketing, decision making and sports as well, because the results of matches can be interpreted as results of the comparison of the pairs. If the number of the objects to compare is large, then it is necessary to apply incomplete comparisons.
In real life we can realize that one member of the pair might have advantage during the comparison, such as the homeland team in football, or the kick off in chess.
