I would have to agree with rastaman70 on this one. Only statistics are affected by in game changes, not the results. At least this is what my 15-match experiment showed me 2 seasons ago. I used 2 teams of equal quality. One always started the match with legal formation and the other with illegal formation and then swap formations asap. 12 times out of 15 the team that started with the legal formation won the match. The rest of the results were 2 draws and 1 lose. I don't know if they have fixed the engine now but I have stopped bothering too much since then.