Short explanation: odds for are the ratio of "favorable" cases to "unfavorable" caes (odds against is the reciprocal). Probability is the ration of "favorable" cases to total cases. For example in flipping a fair coin, the probability of getting heads is 1/2 but the odds of getting heads is 1/1. On an ordinary die, the probability of throwing an ace is 1/6 but the odds against are 5/1.
Empirical data is used in genetics to compute probabilities all the time. It is interesting that you sight gambling to dispute the probabilities posed above. Evolution is just a big gamble after all and it continues to try its luck on different machines all the time but the house wins.
The article that these posts are supposed to be based upon even site a probabilities link. It doesnt work though what are the chances?
The clearest analogy to how variation and selection can lead to "progress" is quantum tunneling. There are everyday events in the subatomic world which cannot take place in classical physics -- particles having insufficient energy suddenly leaping from one state to another, apparently overcoming an insurmountable obstacle.
When an event has demonstrably happened, you cannot use probability theory to prove it could not have happened.