-3

There is a common misconception is that if there is a long period of unlucky outcomes, it's more likely that there will be a lucky one. But of course the probability of each one is identical, and such a long period of same colors is a nice argument against it.

7
  • 1
    I doubt records are kept.
    – user15620
    Commented Feb 15, 2018 at 16:45
  • As a coder, what have you found?
    – justCal
    Commented Feb 15, 2018 at 16:53
  • 4
    As a statistical problem, it's fairly simple, but that's a question of math, not history.
    – user15620
    Commented Feb 15, 2018 at 18:18
  • 2
    @justCal Simulating, or computing such a probability is easy, but real historical evidence adds more social context - people betting on the other color, believing it must hit next now etc.
    – Petr
    Commented Feb 15, 2018 at 18:42
  • 2
    A simple Google search for "longest roulette winning streak" gives the data you are looking for. Since this means that the question is for "easily searchable trivia", I vote to close. Commented Feb 16, 2018 at 2:58

1 Answer 1

2

This casino site says:

Think your number or color has to hit soon? The longest recorded streak of one color in roulette in American casino history happened in 1943 when the color red won 32 consecutive times. In a row. The people convinced black had to hit next were absolutely right. Eventually.


Interestingly, I wrote a quick simulation of this, and can get to 29 in a row rather quickly, but to get to 32 it took several minutes:

68717891 Flips, in Elapsed Time: 0 min, 2 sec ,  Current longest run=25
68717892 Flips, in Elapsed Time: 0 min, 2 sec ,  Current longest run=26
102946854 Flips, in Elapsed Time: 0 min, 3 sec ,  Current longest run=27
102946855 Flips, in Elapsed Time: 0 min, 3 sec ,  Current longest run=28
4078170601 Flips, in Elapsed Time: 2 min, 1 sec ,  Current longest run=29
4078170602 Flips, in Elapsed Time: 2 min, 1 sec ,  Current longest run=30
4078170603 Flips, in Elapsed Time: 2 min, 1 sec ,  Current longest run=31
7945382548 Flips, in Elapsed Time: 3 min, 56 sec ,  Current longest run=32
7945382549 Flips, in Elapsed Time: 3 min, 56 sec ,  Current longest run=33
15853962961 Flips, in Elapsed Time: 7 min, 52 sec ,  Current longest run=34
15853962962 Flips, in Elapsed Time: 7 min, 52 sec ,  Current longest run=35
15853962963 Flips, in Elapsed Time: 7 min, 52 sec ,  Current longest run=36
6
  • 1
    Ignoring spins that result in green for simplicity, the chance of any sequence of N spins coming up red is 1 in 2**N. So roughly 1 in every 4.3 billion sequences of 32 spins should be all red. The other bit you'd want to make the mathematical calculation is the number of spins in "professional casinos".
    – user15620
    Commented Feb 15, 2018 at 18:22
  • So it looks like my simulation (just a coin flipper) was a little slow, taking 7.9 billion flips.
    – justCal
    Commented Feb 15, 2018 at 18:28
  • 2
    @justCal - Computer "randomness" is almost always from an algorithm (IOW: not really random). I've been bitten by this personally. I'd be more interested in expected amount of "flips" before this happens (a prob-stat person could calculate this for you), and perhaps how long you'd have to sit at a single table for it to roll that many times (given how long it usually takes the ball to roll around, bets to be made and collected, etc.)
    – T.E.D.
    Commented Feb 15, 2018 at 21:29
  • @T.E.D. yes, If you don't Randomize the generator to start each run of a program, you can even get identical Random results! This was just a two-minute curiosity code to see what results looked like. At Steven Burnap 's 1 in 4 billion odds, I think you might grow old waiting to get that 32 reds in a row...
    – justCal
    Commented Feb 15, 2018 at 22:30
  • 1
    I once had a RNG I was using and modding by 2 to get a 50/50 flip (which coincidentally may be roughly what you are doing here). Due to some quirk in the algorithm that result always toggled between 0 and 1. Always. The seed didn't matter. The algorithms are much better now (Mersenne Twister wasn't invented until 1997. Mine was cribbed from the 1984 first edition of A Book On C), but still I don't really trust them to model randomness properly.
    – T.E.D.
    Commented Feb 15, 2018 at 22:40

Not the answer you're looking for? Browse other questions tagged or ask your own question.