BGonline.org Forums

Input g+bg for imitating G11 MET

Posted By: Ian Dunstan
Date: Saturday, 16 February 2019, at 7:33 a.m.

In Response To: Variable MET design (Ian Dunstan)

I wrote: "...so perhaps using 26.825% (or even 26.4%) as the VM input is reasonable".

I think I was on the right track. I searched the internet and found a Gnubg 0.14 build download and used it to do 5184 trials of 0-ply move 2-ply cube/25%, money, no Jacoby. This gave me a g+bg rate of ~26.57% which is in the right ballpark as far as a VM (Vers. 0.0) input for imitating G11 is concerned.

This produces discrepancies between [-0.19, +0.20] and a mean discrepancy of 0.06.

Here are the MET's:

In yellow are the actual G11 MET values and in blue is what the VM produces with 26.57% as the input. The MET's look similar IMHO.

Messages In This Thread

• Variable MET design
  Ian Dunstan -- Wednesday, 13 February 2019, at 11:03 a.m.
  • Variable MET design
    Albert Steg -- Wednesday, 13 February 2019, at 2:26 p.m.
    • Variable MET design
      roadkillbooks -- Thursday, 14 February 2019, at 2:01 a.m.
      • Variable MET design
        Florin Popa -- Thursday, 14 February 2019, at 11:26 a.m.
    • Variable MET design
      Ian Dunstan -- Thursday, 14 February 2019, at 10:20 a.m.
      • Variable MET design
        Albert Steg -- Thursday, 14 February 2019, at 4:06 p.m.
        • Variable MET design
          Ian Dunstan -- Friday, 15 February 2019, at 7:33 a.m.
          • Variable MET design
            Ian Dunstan -- Friday, 15 February 2019, at 7:39 a.m.
  • Variable MET design
    neilkaz -- Thursday, 14 February 2019, at 4:51 p.m.
    • Variable MET design
      Ian Dunstan -- Friday, 15 February 2019, at 6:27 a.m.
  • Input g+bg for imitating G11 MET
    Ian Dunstan -- Saturday, 16 February 2019, at 7:33 a.m.
  • Version 0.1
    Ian Dunstan -- Sunday, 17 February 2019, at 6:37 a.m.