Roulette: asking for simulation of random Vs big pattern

hi, pros,

here i want to ask if any one who could do roulette simulation for me to test random and pattern.

reasons to test: 1) it is true, roulette gets no memory. but how does it come into being for law of big numbers?
2) there is not pattern in random. it is true. but i just want to see if any 'big pattern' from random. which may not help with gambling actually.
3) seeking for pattern is human nature. which leads to progress for the creature.
4) by short-term test, like 100, i see: that seeking for pattern is about 1% more right than betting with closed eyes. of course 100 spins are too short.

how to do: 1) 1000 or 10000 simulations of 1000 or 10000 spins.
2) flat bet
3) 1st way: betting only red. by maths, it should be 47% wins.
2nd way: following. for 1st spin, it can be any, let's say, black, then follow it till it stops, the follow another color.
3d way, more human like, not sure if it is simulation possible: only bet when in 30 or 40 or 50 spins, no streak of 4. then do it like 2nd way.
4) when it is a streak of 5 or above. when streak stops for 1, bet 1 for that streak. only 1. for example: when 5 red or 6 red, 1 black, then place the bet on red. then no other bets until another streak of 5 or more appears, maybe 5 blacks or more... then 1 red, now putting another bet for black. because i see this pattern. maybe it is a trick in my brain.
5) if still possible. get a short random pattern: 5 spins or 6 spins or 7 spins or any. to fit into 1 million pins, to check if big numbers can be devided by small numbers (which takes over 50%)

notes: 1) i am sure that the big pattern cannot apply to gambling much.
2) i am not asking for testing of betting systems. i can do it in my blackjack software.

thanks for reading. though it is hard to get good answer, i will wait and see.

Obviously it's fairly easy to create random numbers (0-36 or whatever) and try out any betting ideas. In particular showing you got about 18/37 reds or blacks would only prove your random number generator was working - in fact you use such ideas when testing it! Thus it's better to use maths, as you know the probability of each spin.

Simulations are better where the state of the game, such a cards gone in a Blackjack game, may affect the result. In roulette we assume the spins are independent, so yes it can confirm difficult puzzles such as how often does event_X occur - e.g. a streak of 5 reds occurs - but you have to remember the next spins are independent. Thus I also cannot see how you would be able to test whether "roulette gets (sic) no memory".

What is more difficult is the pattern recognition problem, to test whether what the human sees is real. We tend to spot patterns (e.g. the same number was spun three times in a row seems quite rare, but in theory it should happen fairly often - compared to intuition). To make the test, say the one above, it's easy to use maths to prove what should happen - so a simulation isn't really needed.

Quote: charliepatrick

Obviously it's fairly easy to create random numbers (0-36 or whatever) and try out any betting ideas. In particular showing you got about 18/37 reds or blacks would only prove your random number generator was working - in fact you use such ideas when testing it! Thus it's better to use maths, as you know the probability of each spin.

Simulations are better where the state of the game, such a cards gone in a Blackjack game, may affect the result. In roulette we assume the spins are independent, so yes it can confirm difficult puzzles such as how often does event_X occur - e.g. a streak of 5 reds occurs - but you have to remember the next spins are independent. Thus I also cannot see how you would be able to test whether "roulette gets (sic) no memory".

What is more difficult is the pattern recognition problem, to test whether what the human sees is real. We tend to spot patterns (e.g. the same number was spun three times in a row seems quite rare, but in theory it should happen fairly often - compared to intuition). To make the test, say the one above, it's easy to use maths to prove what should happen - so a simulation isn't really needed.

thanks.
maybe we come to figure out this point:
1 yes, roulette gets no memory.
2 if no meory, why after 1 million of spins, red is about 0.49 million, black is about 0.49 million, not 0.46 million? when at true random.
it is maybe naive, maybe silly, maybe self-evident, maybe it is like how life came in being, maybe it is how nature goes.
yes, i take all of them.

On a single zero wheel you'd expect 18/37 of spins to be Red (i.e. about 486486 SD=500) with about 27027 zeroes. Thus 0.49 you see is actually more correct than the 0.46 you quote. (With a double zero it would be 473684.)

btw if you are getting consistently wrong results then you need to check the random number generator (or whether your program has an error). It is not easy to prove a random number generator is correct, but the easiest way is to play games where you know the expected results (and variance) and ensure you are seeing the same when you run the trials.