Election forensics describes a growing body of work devoted to using statistical methods to try to determine whether the results of an election are accurate: whether the results are the collective choice implied by citizens? intentions given the election rules. Most work proposed in this area has worked negatively, attempting to detect election fraud by describing patterns “clean” elections are supposed to match and then asserting that failure to match the specified patterns indicates-or at least suggests-that election frauds have occurred. Recently what might be described as the first positive empirical model of election frauds appeared (Klimek, Yegorov, Hanel and Thurner 2012): positive in the sense that it presents a model that describes what a fraudulent election looks like and empirical in that it offers an algorithm to estimate the amount of fraud occurring in a particular election. While the model has conceptual limitations and the algorithm has technical deficiencies, the Klimek, et al. (2012) idea opens a path to being able not only to estimate the incidence and magnitudes
of any frauds that occur, but also to be able to recover what the election results would have been in the absence of frauds. Being able to do that is the ultimate goal of election forensics research. It is worthwhile to try to give the idea the best realization possible to see how well it does when implemented in a statistically sound way that is true to the election system the election is being conducted in.