The average turnout in US Presidential elections since 1932 is 56% with a small standard deviation of just 4%.

“Turnout” in 2020 looks like 70%. If you think turnout is normally distributed (it seems to be) then that is 3.5 standard deviations above the average.

The most plausible explanation for such an extreme statistical anomaly is fraud on truly a massive scale using the mail-in system.