No, we already know the outcome of those groups that may or may not have existed and that is that they had no outcome on the current gene pool meaning that in essence they did not exist as far as survival goes.
In your analogy it's like buying 40 lottery tickets knowing that 39 of them will not win for sure vs buying just 1 ticket.
Sure, but you only know that after the fact. Basically, talking about `risk' after the fact is a murky business. (And can invite a lot of philosophical debate about frequentist vs Bayesian interpretations of probability. And determinism.)
In either case, the most likely outcome in the end is still that I don't win the lottery. But the chances were different.