Re: SF: Back to theory

From: Nora Baron (norabaron_at_hotmail.com)
Date: 02/27/05 

Date: 27 Feb 2005 06:27:30 -0800

Yes - that was my intended point. However to my surprise, some
experiments by Tim Peters in this thread seem to indicate that
factoring T^6 = f * g and letting X = f - g and does significantly
better than chance. The reason is not clear. I think X = f^2 - g^2
may do better still.

Nora B.