Quote:
Originally Posted by jungmit
The problem is it happens in huge streaks this is why it looks rigged and why it also could be. Most things live don't happen in streaks like this.
The test I proposed tests for this.
Quote:
Originally Posted by jungmit
But if u check I am 100% sure your aces win the amount of times it should.
That's not what the test I proposed tests for.
Let's say every day for 3 weeks, you're involved in 20 coinflips, and we count how many we win every day. We're expected to win a mean of 10 each day, with a standard deviation of sqrt(5), or about 2.24. I just did a random simulation, and came up with the following:
11 11 11 11 8 13 12 9 12 12 7 10 8 8 6 12 12 14 14 10 13 (
mean = 10.67, standard deviation = 2.67)
Now, let's say your rig is true. Then you might get some numbers like this (many more highs and lows caused by "streaks"):
16 4 15 3 18 5 17 5 18 3 1 2 19 3 16 16 17 6 17 4 5 (
mean = 10.0, standard deviation = 6.87)
Notice how the means are both close to what they're expected to be (10.0), but the standard deviation of the second set is
way higher than the expected value of 2.24? The chi-squared analysis can tell you if it's too high to be truly random. For example, I just stuck the above data into
this calculator, and it tells me there's a 0.00000 probability that the above data could have resulted from a non-rigged RNG (it only goes to 5 decimal places).
Last edited by madcatz1999; 03-02-2016 at 07:32 PM.