I'm saying that while there is a large random multiplier in everyone's outcome, the risk of total failure varies enormously with the founders. Some founders have a near 0% chance of total failure. Others (a larger number) have a near 100% chance. So while this means on average the change of a startup succeeding is 1/n, quoting that stat as part of the description of a startup would be misleading, because for a small number of people it's just not true, just as for a small number of fast runners it would not be true that they'd run slow in the shoes, whatever the average person did.
If so, then that changes nearly everything that people understand about startups.