I would do a little change in this point... tactics are chosen randomly as effective or not, so one option can be "visibly better" ,,,
then, win bonus can add a random "little chance" in the same % as attend in terms of increasing chances...
training bonus is part of the tactics, activate one or another or have at least the 10% trained, affect in the live match... is almost like a change of mentality really, despite that each training bonus affect to a few random number of players....
and condition/morale can be chosen as decisive or not when impact the scenario.... and those determine "initially" a path that will affect the simulation as the diff' between 2 AvQs... thats a static fact when the match starts... is a "starting differance" between 2 teams that pre-determines a starting scenario...
hmmm....