Read on Twitter
Political Scientists began to develop tests to identify election fraud which is based on Benford’s Law (human attempts to manipulate results leave traces behind). Benford’s law “describes a pattern for the frequency distribution of digits in numbers that occurs in many settings”
This law is mostly used to detect financial and accounting fraud which argues that people who fabricate numbers have a tendency to do so uniformly. Hence the significant deviations from the comparison of frequencies could indicate “suspicious” behavior and possible fraud
There are criticisms to this view. Some scholars also say that anomaloys vote count can also be due to normal behavior such as strategic voting. In response, some scholars argued that a relationship between proportion of invalid votes and the second digits of the ballot box vote
totals signify suspicious behavior rather than strategic voting. References: (Hicken and Mebane 2015; Cantu and Saiegh 2011; Deckert, Myaglov, and Ordeshook 2011; Mebane 2014; Hicken and Menane 2015)
An interesting thread on recent developments in Political Science 
“The science of Election Forensics”
“The science of Election Forensics”
