BGonline.org Forums

43Z-64 and 43S-64

Posted By: Nack Ballard
Date: Thursday, 2 December 2010, at 9:55 p.m.

In Response To: 43Z-64 and 43S-64 (David Rockwell)

43Z is played more often than is justifed by the rollouts. Most of my club opponents open with it at all scores. The replies don't allow for many mistakes either. 64 is one of the few replies that I've seen played wrong. I don't play an opening 43Z very often.

On the other hand, 43S draws frequent errors in replies on 11, 41, 43, 61 and 64. I like to play 43S at scores that call for a 43Z move. I have NEVER seen the correct reply in an actual game with a 11 or 61 (5 point and split for 11 and hit and split for 61).

You make good points (as always), and you may well be right that S is the best practical choice, though it is worth measuring the size of the effects. First, here is some dmp and gammon save data for opening 43:

d[Z U2 D4 S5] /46
d[D S2 U2 Z4] <15

s[Z S3 U6 D10] /36**15
s[U Z6 D10 S11] <5
s[Z S1 U9 D16] ^46*10 5

It may depend on which bot you trust (and taking the number of trials of each rollout into account), but from those I would estimate aggs of

d[D=Z=U S1.5]
s[Z S2 U6 D12]

Thus, it is worthwhile to consider all four plays at dmp, and at least two plays at gammon save. To make proper adjustments, one would need to examine typical misreplies against all candidates and assign a percentage to each (the chance of the misreplies actually being made).

How much does the misreply of 43S-11N cost the opponent? At dmp, N is -.017, and at gammon save, N is about -.037 (the latter based on Snowie AtS data, lacking any other). The enhancement to playing 43S is about .0005 at dmp or .001 at gammon save if Opp wrongly plays N 100% of time.

You also mentioned the (presumably smaller-impact) rolls of 41, 43, 61 and 64. It will depend on which misreplies are made, and how often. Then, one has to subtract any (estimated) misreply possibilities to Z, D and U, accounting for all twenty-one rolls, to get a complete picture.

Granted, .0015 or .002 isn't much to make up, but in the case of dmp it has to be made up against all three rival plays, and the net effect of misreplies at non-money scores are smaller than one might imagine without doing calculations.

It is also worth considering the relative strength of the opponent and the types of games in which one might engage by playing D (Down) at dmp rather than one of the splitting plays.

Nack

Messages In This Thread

• 43Z-64^36 RO
  Claude Landry -- Thursday, 2 December 2010, at 3:21 p.m.
  • 43Z-64 and 43S-64
    Paul Weaver -- Thursday, 2 December 2010, at 3:52 p.m.
    • 43Z-64 and 43S-64
      David Rockwell -- Thursday, 2 December 2010, at 7:20 p.m.
      • 43Z-64 and 43S-64
        Nack Ballard -- Thursday, 2 December 2010, at 9:55 p.m.
        • 43Z-64 and 43S-64
          David Rockwell -- Thursday, 2 December 2010, at 10:51 p.m.
          • 32Z
            Nack Ballard -- Friday, 3 December 2010, at 1:04 a.m.
    • 43Z-64 and 43S-64
      Nack Ballard -- Thursday, 2 December 2010, at 10:18 p.m.
      • 43Z-64 and 43S-64
        David Rockwell -- Thursday, 2 December 2010, at 11:01 p.m.
        • 43Z-64 and 43S-64
          Stick -- Thursday, 2 December 2010, at 11:13 p.m.
          • 43Z-64 and 43S-64
            David Rockwell -- Thursday, 2 December 2010, at 11:44 p.m.
        • 54S-63 and 43Z-64
          Nack Ballard -- Friday, 3 December 2010, at 12:50 a.m.
          • 54S-63 and 43Z-64
            David Rockwell -- Friday, 3 December 2010, at 1:20 a.m.
            • 54S-63 and 43Z-64
              Nack Ballard -- Friday, 3 December 2010, at 1:46 a.m.
    • 43Z-64 and 43S-64
      Nack Ballard -- Thursday, 2 December 2010, at 10:36 p.m.
  • 43Z-64, data
    Nack Ballard -- Thursday, 2 December 2010, at 8:52 p.m.