You have put your finger on the tyranny of statistics. The biotech statistics for the groups are both predictive and accurate. Unfortunately that accuracy falls apart when it comes down to predicting what will happen to a single entity as opposed to the group as a whole.
Yes, I try to keep this in mind when pondering .. What are the odds that _________? (fill in the blank). For example, what are the odds that a hang-glider pilot will have a fatal crash? I believe it's close to 1 in a thousand per year. Haven't checked in awhile but there are (or were last I checked) roughly 6,000 certified hangies and about 6 on average die each year in the U.S; early on (up to the late 70's) it used to be around 1 in 20. Let's pretend it's actually one in a million; Like winning the lotto. What I try to always keep in mind is this: It doesn't matter if you're the one!
btw, the first hang gliding (friendly, not professionals) competition I ever participated in was called the Team Challenge in the Sequatchie Valley a 1600' cliff launch. Day 1, First pilot off the hill failed to hook-in to his glider. He didn't survive. More than a dozen pilots were there on launch including the Meet Director, Safety committee; well over 100 years of collective experience there and nobody caught it. What are the odds? It doesn't matter if you're the one.
Statistics won’t tell us which stock is the winner or loser, only the probability.
That is down to skill, experience and good fortune. Having a diversified portfolio where all that is applied should keep one in business longer, but likely in a more boring manner.