Three things, even though this isnt my league
1) fouls per minutes is the more interesting statistic than fouls per game, and thats easily enough calculated
2) id use private league data and friendlies as well, just because the data still seems relevant
3) if the player is on a middling to slightly losingish team, and plays alot of minutes, the player could rack up extra garbage time fouls from the match engine, which are what they are. i try to ignore foul rates, which has cost me some (i have one of the worst fouling average teams in the country), but overall the bargain prices generally make it worth it.
the only spot i refuse to compromise on fouling averages is if im playing a player thats had a year or two of training, with the intention of reselling in the future. at low levels of prices the fouling doesnt make a difference, but from a gross return on investment the fouls do make a difference at the higher levels.
i.e: if you buy a 19 year old now with 2 seasons of training for 800k and intend to sell for 3.2 mil, you could maybe buy the foul prone player for 700k, but you will sell for alot less than 3.1 mil or even 2.8 mil, so that matters.