I probably mentioned this example earlier in the thread, but I will repeat it here.
The game of craps is a good example to use. The chances of rolling a 2 or 12 is really low, and rolling a 7 is much more probable than any other result when rolling two die. However, if you place a bet on getting a 2 or 12 this does not increase the chances of rolling a 2 or 12, nor do you increase the chance of rolling a 7 by making a bet that would win on a 7. The bets on the table don’t influence the result of the dice. In this case, the roll is random with respect to fitness.
Mutations appear to be the same. There is no detectable connection between what the organism needs and the mutations that happen. Some mutations are more likely than others, but it doesn’t seem to have a direct connection to the specific needs of an organism in a specific environment.