The NEMO initiative works towards agreeing on one set of head-related transfer functions (HRTFs) for use across various applications whenever incorporating individual(ized) HRTFs is not feasible. This initiative is grounded on the assumption that listeners benefit from adapting to a set of HRTFs that is different from their own. Naturally, one key step in this process is the selection of the particular set. We hypothesize that the available information in the set influences adaptation speed and/or post-adaptation localization performance and, therefore, is a potential factor in the selection. To test this hypothesis, it is essential to develop measures that quantify the information of an HRTF set. To this end, this contribution presents several potential measures of information content and compares them using multiple HRTF sets.