Scenarios vary depending on the rating variations, and the attend/don't attend factor too sets a parameter, is like
team A - attend power 5 /not attend power 7 too
team B attend power 3 /not attend power 6
Team B will loose always if attends, if do not, the oppo A will need to attend to get a posible win... and this gives you a simple explanation of how the force is distributed depending on the variables that the game uses to create the scenarios, which are rating variations of players to do a simplification of the team, for each game so players have different rating behavious, like 7-5-7-6-7 10-10-9-8-10 (here and always 9-10 is the same value, as is the max rating and means the same 9 or 10) so the engine will create unique scenarios selecting few players and then depending on the variations, they will improve or play worst than the previous game. As I explain somewhere you can force a drop of ratings, intentionally to get a push for a upcoming game.
Then the game have some dozens of scenario conditionants that are always in on-off so, condition at 0% matters or not (on-off. morale on-off, win bonus on-off. and like this for all conditionants, which determine the weight of that starting value that the teams have in terms of chances (power team A 5-7 / B 3-6) meaning this number the variables selected by the engine + how far can reach your team in terms of chances compared with the opponent once the engine converts the pre-determined factors in available animations to resolve the games.
Try always to understand the engine behaviour, and play in its favour because the game needs a system to work, and there is no other way to interpret this.
It's like trying to cheat casino roulette rather than playing management game.
Ok, there is a manager presence variable that affects the result. In my opinion, its power should be higher. Lose teams, whose managers do not play (or barely play) is very disappointing.
If the game "decided" that you will lose, you will lose.
Sorry for the English. This is google translate.