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

BGonline.org Forums

53P41@GS extended-> $ = S

Posted By: Daniel Murphy
Date: Sunday, 18 October 2009, at 8:52 p.m.

In Response To: 53P41@GS extended-> $ = S (Andreas)

Let me ask, also, how is that a mere doubling of the number of trials from 10,368 to 20,736 (instead of a quadrupling) reduced 95% confidence intervals by more than 50% (from ±00.010 to ±00.004 and from ±0.011 to ±0.004)?

Is that not unusual? I expect CI reduction with trial extension to usually be similar to the position below rolled out 0-ply:

Blue on roll, cube decision?

White66


2X2X2X2X2X2X1X1X1X ' ' '

2O2O2O2O2O2O1O1O1O ' ' '

Blue66

Position ID: 27ZVAADbtlUAAA Match ID: cAlgAAAAAAAA

Trials 36 72 144 288 576
SD (ND) 0.02080.01630.01140.00720.0055
SD (D/T) 0.02790.01900.01270.0090 0.0063
Equity Difference0.14270.16220.15440.1609 0.1676
JSD 4.100 6.486 9.046 13.930 19.882

F.ex., quadrupling trials from 36 to 144, or 72 to 288, or 144 to 576, approximately halves the reported SDs (and approximately doubles the reported JSDs).

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.