Power Indices: Shapley-Shubik OR Penrose-Banzhaf?

Shapley-Shubik and Penrose-Banzhaf (absolute and relative) power measures and their interpretations are analysed. Both of them could be successfully derived as cooperative game values, and at the same time both of them can be interpreted as probabilities of some decisive position (pivot, swing) without using cooperative game theory at all. In the paper we show that one has to be very careful in interpretation of results based on relative PB-power index and not to use it without absolute PB-power index, what is frequently the case in many published studies.