Let's try to figure out one thing:

The simulator counts
1- League situation (general scenario, this can be, for ex- keep a close distance vs the 2nd, to have a scenario with a close league, usually, lose vs a team in the middle table... etc of scenarios)

2- Pre-match scenario that counts the distance of Main AvQ between the 2 teams, and chose some orders to be determinant, and some not (this linked to the point 1)
Remember that from X distance of avq, the old high scores where cutted, still the situation is equal in the case to face a similar avq, or a distant one, we can, have a 5-0 vs a similar avq, and then a 1-1 vs a -30%,,, but same time happens that we can draw with the similar team 2-2 and win 4-0 vs the -30%.


3-Live match -once the scenario is set, here there is a impact coming from the internal programming, -hidden skills- and hidden X players that contribute by default (ex-theres always a best player for FKs, one for corners etc... chose him and the game will react properly-
Hidden skills means:
You can have a 5-6* player, with all atributes in 1% rank and just 2 skills +340% and, as the internal programming is linked to the main AvQ and raises with this %, then thats why it is enough to oversize 2 skills.
The impact of each individual skill in the internal parameters is minimal and thats why if one train each ST till 9*'s never obtain the same result, because there's this internal limit of contribution, that raises with the main player AvQ.

And more or less thats the point I wanted to share.