Pack, your love of superstition is rather touching, I'll give you that. Your notion that "math only counts in the long term" is bizarre to say the least. What I think you're saying is that only on the very long end of an actuarial curve will percentage probabilities be relevant and that some mysterious force like "luck" or "chance" takes over in the short term.
This is an incorrect way to view statistics for any independent trial activity, eg Roulette, Craps, Sic Bo etc etc
Think about it this way:
Joe Six Pack goes to an Indian casino and bets $ 100 on black in roulette. He holds his lucky rabbit foot tight in his hand. The ball bounces into a black square an incredible 25 times in a row! Joe cashes out his $2500 and goes home to tell his wife about the incredible streak he just stumbled upon.
But it's not a streak. The outcome is a result of 25 independent events each of which has an 18/38 or 47.4% chance of occuring. Spin # 25 has the same percentage chance of hitting black as does spin number 5 billion. Each event occurs independent of the previous event.
The actuarial curve will show a 100% probability of hitting black if only the first 25 spins are measured. As the number of spins move toward infinity the curve will slowly show a 47.4.....% probability. This does not mean that the odds change as the number of spins increases, only that the more trials that occur the greater the true probability is illustrated.
Hope that helps