statistics life hack: you can just remove the data points that look weird in order to validate your hypothesis