statistics question

1HandInMyPocket

Unoffical Capital One rep
Mar 2, 2002
1,563
0
36
Mirror Universe
I'm trying to test out a theory if certain croupiers have a tendency to hit (or avoid) the zero/double zeros more than statistically possible. I know in the long run a zero/double zero will occur 5.26% of the time. Lets say dealer A has the following results: day1) 3/27 spins (11.11%) are 0/00; day 2) 0/40 spins (0%) are 0/00; and day 3) 2/18 spins (11.11%) are 0/00. I will continue to record more results from dealer A, and from other dealers on an on-going basis.

Would it make sense to use my recorded results (11.11%, 0% and 11.11%) from dealer A and subtract 1) 5.26% or 2) the current average of the 3 recorded days (7.41%), when trying to calculate the standard deviation?

Thanks in advance
1Hand

edit: Sorry, I forgot to mention if you havent figured it out, that I am referring to roulette.
 
Last edited:
The results are statistically consistent with the odds. Over three days 3+0+2/27+40+18=85.

5/85 = 5.88%. From 85 attempts, 4 times 0/00 would be 4.71%.