|
From: | Philippe Michel |
Subject: | Re: [Bug-gnubg] Alternative weights files |
Date: | Wed, 19 Sep 2012 08:45:22 +0200 (CEST) |
User-agent: | Alpine 2.00 (BSF 1167 2008-08-23) |
On Wed, 19 Sep 2012, Joseph Heled wrote:
You should use -q1, where matches are played in pairs, and in each game the same dice are fed (reversed) to the two players. That way, if the two players make the same choices, you eliminate luck completely :). If not, you reduce its effect. But now you have 1/2 the trials, with 3 outcomes per trial, -,0,+ (X winds both, they split, O wins both), so the stats will be different. But I would repeat with a -q1. 50.69% is huge. I don't expect it to go away.
50.69% was with 2ply evaluations, it will take some time to reproduce the experiment :-).
In the output with --q1 : # X - 504 O - 496 (50.40+%) (262 121 117)The 50.4% gross result is unbiased, right ? But estimating the variance is not clear to me. Compared to the plain random dice there must be a additional term close to (121+117)/500, something like that.
[Prev in Thread] | Current Thread | [Next in Thread] |