Well if it isn't Provably Fair you should just out this casino's name?
like i said before, this sample size is useless, a sample size of under 100 hands is in no way determinant of fairness or unfairness. you would need at least 100k hands to have any meaningful data to analyze, and drawing a conclusion from this data would be inconclusive. no need to fire off an accusation without being 100% sure.
Maybe for Blackjack it would require a large number of hands but I'm not so sure.
I know you can run a Chi-Test on Craps and if it is blatant cheating you can notice within a few hundred rolls without a shadow of a doubt.
you should see Michael Shackleford(Wizard of odds) and his test on BLR technologies craps...I don't care to take this testing this much farther, I guess just stay on Provably fair casinos and always change client seed...all you can do