Ever since the discovery by Doddington, et al. that the vast majority of biometric errors come from a relatively small number of “weak users”, significant research has been devoted to understanding and correcting for the so called biometric menagerie problem.  The majority of work in this area revolves around the assumption that match scores are partially dependent on the particular subjects involved in the comparison operation.  The focus of this research is to promote a fundamental shift in this understanding by demonstrating that identification rates are noticeably affected by certain intrinsic properties of the subject, such as his or her ethnicity, gender and eye color as well as characteristics of the image, notably the wavelength of light utilized by the sensor.  To support these claims, a unique data collection in which these variables are controlled is created.  Experiments involving this dataset indicate significant differences in match score behavior as these factors change throughout the subject pool.  These results are crucially important to understanding the expected performance of iris biometric systems as they are utilized on larger and more diverse population sets.