Unlikely Voter

I notice a trend: when people mock reasonable sample sizes (as though multiplying out 500 terms isn’t powerful enough math for anyone), it only seems to be when their preferred candidates are losing, and never the other way.