http://en.wikipedia.org/wiki/Random_number_generator
Some of you are missing the point that “random” doesn’t actually exist. So, it’s no wonder that the average programming language doesn’t do it well.
“Random” is just something we use as a mathematical model for reality.
If I roll dice, there will be a 1/6 chance that any side will come up. The reality is that the physical forces exerted on the dice will *exactly* determine which side will come up. The way I throw it… the surface it hits… bouncing off of a felt wall, etc.
If you could figure out all of the forces that would happen before I actually throw, you could tell me exactly which number will come up. That’s too complicated. Because of that, we use the concept of “random” to model such an event so we can make an educated guess. Using the concept of random, we determine that every side will come up 1/6 of the time over the long run… say 1,000,000 throws. That way, we can still make an educated guess on what we think will happen.
Probability theory doesn’t give you answers. Physics will give you answers. Probability theory gives you a framework so you can make educated decisions when the physics are too complicated, mostly because the variables are too many.
Payout percentages can be misleading. For one thing, they will be different based on how people play. Craps for instance… there are a ton of bets, and 2 bets are way better than most others. If people are playing bad bets, the payout percentage will drop. If people are playing better bets, the payout percentage would be higher. The reason you don’t see wide variation is because there are so many rolls. Probabililty theory won’t give you much of an answer at one point like calculus will. The more events you sample, the better.
Something touched on in another thread, but glossed over. There’s a reason why many people don’t cash out from casinos. It happens to be the same reason why progressive betting strategies don’t work. On a game where you can keep doubling your bet, there are betting limits. Casinos take advantage of short run variations. Outcomes will vary widely in the short run, and we could call any one person playing for one night a short run.
Some will win a lot in a short period. Some will lose a lot in a short period. When you lose a lot in a short period, the casino doesn’t have to grind you down slowly. Game over, you’re out of money. That’s why they have table limits. At some point, they’re going to force you into a long run scenario. If not for that, a “double every losing bet and leave the casino when win you win one” would actually be a winning strategy.