[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums

New GV article

Posted By: Timothy Chow
Date: Wednesday, 16 December 2009, at 3:52 p.m.

In Response To: New GV article (Rich Munitz)

Rich Munitz wrote:

Why is 1296 a better number?

1296 is a better number simply because 1296 is greater than 108. 5184 would be an even better number than 1296, etc. (There's also the fact that 1296 is a power of 36 so that quasi-random dice are in effect, as Jason mentioned.)

I think if you were to take a hard "look under the hood" at the standard assumptions about Gaussian approximations, independence, variance reduction, finite equity, etc., then you would start to feel uncomfortable with 108 trials too. If everything really were nice and independent and Gaussian and so forth, then statistical significance (as it is usually understood in this context) would be a reasonable stopping criterion; but those assumptions are not really true in the real world.

But, as I said, it's a matter of personal judgment.

Messages In This Thread

 

Post Response
Your Name:
Your E-Mail Address:
Subject:
Message:

If necessary, enter your password below:

Password:
 

 

[ View Thread ] [ Post Response ] [ Return to Index ] [ Read Prev Msg ] [ Read Next Msg ]

BGonline.org Forums is maintained by Stick with WebBBS 5.12.