BGonline.org Forums

GNUbg vs XG as a learning tool update

Posted By: Timothy Chow
Date: Sunday, 25 November 2012, at 9:38 p.m.

In Response To: GNUbg vs XG as a learning tool update (konstantin)

konstantin wrote:

does gnu actively tries to enter complicated Positions?

No. GNU 2-ply and XG 3-ply agree 98% of the time. And the 2% difference is just due to differences between what their neural nets think is optimal play, not to any sort of contempt factor or fish MET.

Messages In This Thread

• GNUbg vs XG as a learning tool update
  leobueno -- Sunday, 25 November 2012, at 8:31 p.m.
  • GNUbg vs XG as a learning tool update
    konstantin -- Sunday, 25 November 2012, at 8:46 p.m.
    • GNUbg vs XG as a learning tool update
      Timothy Chow -- Sunday, 25 November 2012, at 9:38 p.m.
    • GNUbg vs XG as a learning tool update
      Simon Knight -- Tuesday, 27 November 2012, at 1:37 p.m.
      • GNUbg vs XG as a learning tool update
        mtuhtan -- Tuesday, 27 November 2012, at 2:43 p.m.
        • GNUbg vs XG as a learning tool update
          Michael Petch -- Tuesday, 27 November 2012, at 5:27 p.m.
          • GNUbg vs XG as a learning tool update
            Michael Petch -- Tuesday, 27 November 2012, at 5:34 p.m.
  • GNUbg vs XG as a learning tool update
    Timothy Chow -- Sunday, 25 November 2012, at 9:31 p.m.
    • Why we need actual statistics?
      Mr Majestyk -- Monday, 26 November 2012, at 1:16 a.m.
      • Why we need actual statistics?
        Timothy Chow -- Monday, 26 November 2012, at 4:18 a.m.
  • GNUbg vs XG as a learning tool update
    Chuck Bower -- Monday, 26 November 2012, at 1:19 a.m.
  • GNUbg vs XG as a learning tool update
    ah_clem -- Monday, 26 November 2012, at 3:37 p.m.
  • GNUbg vs XG as a learning tool update
    neilkaz -- Monday, 26 November 2012, at 4:40 p.m.
  • GNUbg vs XG as a learning tool update
    christian munk-christensen -- Monday, 26 November 2012, at 5:29 p.m.