this is why dice control and roulette can only have theories, not actual proof, unless of course you did a million trials, or the wheel was severely off
Not quite the case. Due to the larger house advantage, the results on roulette and craps should overcome standard deviation far quicker than they would in a game of blackjack. You are still talking large samples of numbers for any form of proof, but not nearly as large as you would be for blackjack.
The main issue is getting a dealer who claims to be able to target the ball to step forward and do exactly that for 5 figure + spins. Although i'm fairly confident that if it was possible, that would have happened by now.
RJT.