A simple "statistical" question that makes a big difference in the game play: Let's suppose you have two axemen, one well promoted and the other one just with Combat 1 and you are finding yourself with the task of capturing an enemy city with a lone archer inside. You check the percentages, and you see that the "good" axeman has 60% chance and the "bad" one has 30% chance (you can also change the numbers, if you want). The question is what do you think is best - attack first with the "good" one and if he loses then with the "bad" one on the injured archer, or make first an attack with the "bad" axeman and then (if he loses) with the good one? It seems that the second method gives better chances for a "sure" ultimate win, but has the disadvantage of 1) losing more units if you conduct wars like this, and 2) since your "good" units are always attacking injured opponents, they will never reach too high levels in experience (reduced experience bonus). Is it a matter of style, or has anyone thought of a strategy about it? The idea can be applied to many places in the game, not only in early rushes.