This is an awful use of statistics--it compares apples to oranges.
Besides, the odds of an event happening must be factored by the number of
chances the event has to occur. For instance, when I try to express a
foreign gene in bacteria, the transfection efficiency is very low, maybe 1
in a million. Nevertheless, if I use several hundred million bacteria and
several million DNA molecules a few bacteria will successfully take up the
DNA. Then I provide a selective environment so that the successful
transfectants have a growth advantage. So, long-shot odds do not tell us
the true chance of seeing an event happen.