Breakthroughs in both the theory and practice of the statistical analysis of geometric partitions have been driven by applications the analysis of potentially “gerrymandered” political districtings. In this talk, we will survey the some of these developments, including the role of Markov chains for both sampling and outlier detection.