BGonline.org Forums

The statsig police

Posted By: neilkaz
Date: Sunday, 18 October 2009, at 5:29 p.m.

In Response To: The statsig police (Timothy Chow)

Personally, I have no problem with the way the new set of Depreli positions are being rolled out. I'd state the same thing if I wasn't working doing rollouts for the project.

Positions are rolled out 1296 times with GNU and those candidates that are more than 2.33 sd (98% CI I believe) below the leader can stop after 720 or more games.

While it would be nice to have more rollouts in some cases for a more accurate final emg, most of us want to see this project finished at some reasonable point in the not too distant future. Once it is finished, interested parties certainly can do longer GNU rollouts or even use XG (it surely seems to be proving to be very strong) to add to the accuracy.

I must point out that there are about 4500 positions that will be rolled out and included in the bot comparision database. Statistically, while one would like each position's emg to be deadly accurate, I feel rather confident that the inaccuracies can be expected to even out somewhat over so many positions. If we were awarding $1,000,000 to the winner of this bot contest, we might want to roll everything out 12,960 times..but we aren't giving away megabucks.. we are just trying to get an idea of how the various bots and settings compare to each other, just as was done for the initial 626 positions.

Note that the current GNU was the best bot in the original 626, and that my understanding is that it also won long money sessions and match sessions vs Snowie 4...ie it is slightly better as was shown by the 626.

The current work (~285 pos) has allready shown a few important things. It has shown that BGB seems to be cubing too soon. This info should be valuable to Frank and shows him where he may be able to improve BGB strength. The current work is also implying that GNU 4-ply is really strong. I've been saying recently that those who study seriously with GNU and have fast comps would do well to run 4 ply analysis of their serious matches.

The current work is also implying that the new bot on the scene, XG is very strong as well. I do think that with the speed and ability of XG, all serious BG students would do well to get it. Note that work continues with frequent updates and consideration for what we high end bots users want.

Timothy, your post shows a deep understanding of statistics, however, I caution against reinventing the wheel and note that there are at least 4500 positions in this new set.

Messages In This Thread

• The statsig police
  Timothy Chow -- Saturday, 17 October 2009, at 9:00 p.m.
  • The statsig police
    Maik Stiebler -- Sunday, 18 October 2009, at 11:01 a.m.
    • The statsig police
      Timothy Chow -- Sunday, 18 October 2009, at 2:34 p.m.
      • The statsig police
        Maik Stiebler -- Sunday, 18 October 2009, at 2:56 p.m.
        • The statsig police
          Timothy Chow -- Monday, 19 October 2009, at 1:48 a.m.
          • The statsig police
            Maik Stiebler -- Monday, 19 October 2009, at 4:21 p.m.
            • The statsig police
              Timothy Chow -- Tuesday, 20 October 2009, at 2:19 a.m.
      • PS
        Maik Stiebler -- Sunday, 18 October 2009, at 3:10 p.m.
    • The statsig police
      Philippe Michel -- Sunday, 18 October 2009, at 3:57 p.m.
      • The statsig police
        Maik Stiebler -- Sunday, 18 October 2009, at 4:37 p.m.
        • The statsig police
          Philippe Michel -- Sunday, 18 October 2009, at 7:23 p.m.
  • The statsig police
    neilkaz -- Sunday, 18 October 2009, at 5:29 p.m.
    • The statsig police
      Timothy Chow -- Monday, 19 October 2009, at 1:54 a.m.
  • Another example, to clarify
    Timothy Chow -- Monday, 19 October 2009, at 2:15 a.m.
    • Another example, to clarify
      Timothy Chow -- Monday, 19 October 2009, at 2:21 a.m.
  • The statsig police
    Michael Depreli -- Monday, 19 October 2009, at 7:16 a.m.
    • The statsig police
      Timothy Chow -- Tuesday, 20 October 2009, at 2:10 a.m.
      • The statsig police
        MaX -- Thursday, 22 October 2009, at 12:19 p.m.
        • The statsig police
          Timothy Chow -- Friday, 23 October 2009, at 3:35 p.m.
          • The statsig police
            MaX -- Monday, 26 October 2009, at 4:32 p.m.
        • The statsig police
          Daniel Murphy -- Friday, 23 October 2009, at 3:50 p.m.