Hi all,
I've had some statistics coursework, probably just enough to be dangerous.
Here's my problem. By the way this is an actual problem, not theoretical.
I need to analyze the hold percentage on certain table games in the casino I
work at. The hold percentage is the % of money a gaming table wins / the
total amount dropped into the "drop box" (Win/Drop). This is done to
evaluate whether problems exist, i.e. theft, bad game protection, etc..
The data I have is for each day, I have the calculated hold percentage for
each of the individual table games. There are multiple table games of each
type, for example there are 7 blackjack tables.
Q: I want to calculate the standard deviation and confidence interval for
blackjack by week. Do I add the total win and total drop for the entire
week and establish a weekly hold percentage and use multiple weeks to
calculate a standard deviation and hold percentage?
Since the daily hold percentage can vary widely, a more useful number is a
year to date hold percentage. How do I determine whether the current hold
percentage is within an acceptable range? I will see fluctuations lasting
weeks where our hold percentage is lower than "standard" norms, but cannot
tell if these fluctuations are statistically significant, or just
fluctuations within an expected range.
Any help anyone can provide would be greatly appreciated. I will gladly
provide more detail if necessary.
By the way, if you have any questions, particularly technical ones, about
casino "secrets", feel free to ask. There really isn't too much that is
secret.
Mark Solberg
=================================================================
Instructions for joining and leaving this list and remarks about
the problem of INAPPROPRIATE MESSAGES are available at
http://jse.stat.ncsu.edu/
=================================================================
