/sci/ why don't we create a dataset for voter fraud by simulating the election and then train a convolutional Neural Network to recognize fraudolent elections? It would be a much more powerful approach than doing retard tier excel analysis