BGonline.org Forums

Variance Reduction Question

Posted By: Timothy Chow
Date: Saturday, 5 March 2011, at 4:10 p.m.

In Response To: Variance Reduction Question (Michael Depreli)

The main reason I can see for not using VR is if, despite its name, VR is not actually reducing the variance because the evaluations are too far off the truth. If you're confident that the bot is choosing the top play correctly but are worried about the evaluations of other plays, and if you're getting low enough variance anyway without VR because your computer is fast, then you could skip VR and eliminate a source of doubt.

Having said that, I can't say I've ever encountered a situation where I felt that VR would actually hurt.

Messages In This Thread

• Variance Reduction Question
  Michael Depreli -- Friday, 4 March 2011, at 11:00 a.m.
  • Variance Reduction Question
    Matt Cohn-Geier -- Friday, 4 March 2011, at 11:13 a.m.
    • Variance Reduction Question
      Casper van der Tak -- Friday, 4 March 2011, at 11:26 a.m.
      • Variance Reduction Question
        Phil Simborg -- Friday, 4 March 2011, at 3:28 p.m.
        • Variance Reduction Question
          LowRayTeeng -- Saturday, 5 March 2011, at 5:43 a.m.
          • Variance Reduction Question
            Phil Simborg -- Saturday, 5 March 2011, at 8:11 a.m.
            • Variance Reduction Question
              Neil Robins -- Saturday, 5 March 2011, at 1:08 p.m.
        • Variance Reduction Question
          Dmitriy Obukhov -- Saturday, 5 March 2011, at 7:25 p.m.
          • Variance Reduction Question
            Timothy Chow -- Sunday, 6 March 2011, at 7:52 p.m.
        • XGR+ vs. XG ROLLOUT results
          6stones -- Sunday, 6 March 2011, at 11:29 a.m.
  • Variance Reduction Question
    Timothy Chow -- Saturday, 5 March 2011, at 4:10 p.m.
  • Variance Reduction Question
    Stick -- Saturday, 5 March 2011, at 4:16 p.m.
  • Variance Reduction Question
    Maik Stiebler -- Saturday, 5 March 2011, at 9:13 p.m.