| Abstract: | At first, it was demonstrated that social comparison theory (Festinger, 1954) predicting a need for moderate (instead of maximal) superiority, could reconcile a number of disparate results of earlier Maximizing Difference Game (MDG) experiments. Using the same theory, it was then further predicted that high power players in an asymmetrical MDG would compete less than their inferior opponents or than equal power players in a symmetrical MDG and that the inferior and equal power players would not differ in competition. The data of an experiment, involving an asymmetrical (8/6) MDG matrix and two symmetrical MDG matrices (8/8 and 6/6), generally confirmed these predictions, but it was observed serendipitously that the high 8/8 symmetrical matrix yielded more competition than the lower 6/6 symmetrical matrix. An extended replication of this variable with five linearly related MDG matrices (4/4, 5/5, 6/6, 8/8 and 12/12) showed a similar result, namely most competition in the highest matrices and least in the lowest ones. The finding was interpreted in terms of the competitive motivation of the players and the low cost of competition in the highest matrices. |
