Statisticians call this kind of decision making, sequential decision making. You are forced to decide on the spot while you are still gathering data.
Assume that you desire to get married. There are 100 eligible suitors. You want the very best possible mate from this pool. How do you go about finding the best possible mate? Obviously you don't want to go through the trouble of dating all 100 of the suitors. The rules of this particular game require that you can only go out with each person one time. After that date you have to decide whether this is the one. You can never date that person again if you reject them after that first date. If you pick that one the game is over. So if you choose the first person you will never get to know whether later dates might have been better. If you reject too many those following may never equal the earlier suitors.
What is the best strategy for finding a mate under these circumstances?
Answer: According to H.W. Lewis, a retired physics professor and the author of the excellent book "Why Flip a Coin" the best strategy is to patiently date and rate and wait through 36 prospective partners before selecting the next one who is better than one of the preceding suitors.