Game 1, move 14, Herd to play 6-3

[Zorba]

The Herd entered with a 1-5 and slot the threepoint, B/24 8/3. dorbel rolled a 5-3, obviously makes the eightpoint and brings a builder down: 13/8 13/10.

Now we roll a 6-3!

XGID=---A-BC-B-B-bC-a-b-bbbbbB-:1:1:1:63:0:0:0:5:10
GNUBg Id: tm0mBgBkZ8YBMA:UYm5AAAAAAAA

[RickrInSF]

Spoiler

i think we can't hit, too much chance of gammon with the blot on the 3

[close]

[ah_clem]

Spoiler

These positions make my head hurt.  Hitting is soooooo tempting, but it risks putting a checker on the roof and a blot behind a 5 prime.  Extricating the second runner is going to be a problem, and if he enters with a 3, kiss the match goodbye as it's almost certain gammmon.  

OTOH, 3's are duplicated after the hit; if that was the end of the story, I'd be glad to risk a 14/36 shot at winning, but there's still the second checker to escape.  

On the gripping hand, he'll have only two checkers in the outfield aimed at the blot on the ace, and they're both outside of the direct hit zone.  I doubt he'll break his 5-prime to hit loose (although 1-1 and 2-2 will probably result in him switching points) , so the blot on the ace would most likely be safe for a move or two.

The safe thing to do is hunker down on the ace, hope for a shot, and be content to run home for a loss of two if the shot doesn't appear or bear fruit.  The alternative is to try to turn the game around now with a risky play that puts the gammons right back on the table.  

We've got a blockade of our own, albeit a porous one.  That's worth something. Saving gammon is not a sure thing if we hold the ace point anchor.   With his lack of builders in the zone, the hit seems to be a reasonable choice.  No idea if it's actually correct, but I'm gonna go for it.  24/15*.

[close]

[sixty_something]

Spoiler

i was sorely tempted to play 13/7 6/3, as our timing is good, but we may never get another shot as good as this one .. thus, i am going with 24/15* .. if dorbel hits our slotted 3 we may be sunk .. if he misses, i like that our other blot in the outfield duplicates the 3 needed coming in .. so, with a miss we may have a chance to close ranks and get all but the straggler at 24 to safety on our next roll .. we are fortunate that there are no extra builders with a direct shot at 24 on the next roll, but we are also very vulnerable to the joker 33 and even 11 and 22 giving dorbel opportunities for double hits on the next roll or the one(s) after .. this one could be for the whole ballgame as we may be giving away a gammon and a 0-4 deficit, but seems worth the risk .. if we win, our gammon chances are much lower, but that possibility certainly exists so we could be looking at a possible 4-0 lead .. thus, hit and hope

[close]

[ah_clem]

Rollout:

Spoiler

The good news is that the Herd's choice wins the most games by a fair amount.
The bad news is that it loses so many gammons that it's a double-whopper.  Oops.

What's the best move? Basically, anything other than the hitting play.



    1. Rollout          13/7 6/3                     Eq.:  -0.787
       0.190 0.012 0.002 - 0.810 0.197 0.011 CL  -0.878 CF  -0.787
      [0.002 0.001 0.001 - 0.002 0.002 0.001 CL   0.004 CF   0.006]
       
    2. Rollout          13/7 10/7                    Eq.:  -0.801 ( -0.014)
       0.190 0.012 0.001 - 0.810 0.208 0.012 CL  -0.895 CF  -0.801
      [0.002 0.001 0.000 - 0.002 0.002 0.001 CL   0.005 CF   0.006]
       
    3. Rollout          10/4 6/3                     Eq.:  -0.803 ( -0.016)
       0.186 0.011 0.001 - 0.814 0.201 0.012 CL  -0.893 CF  -0.803
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.006]
       
    4. Rollout          13/4                         Eq.:  -0.831 ( -0.044)
       0.178 0.011 0.001 - 0.822 0.204 0.012 CL  -0.917 CF  -0.831
      [0.002 0.000 0.001 - 0.002 0.002 0.001 CL   0.004 CF   0.006]
       
    5. Rollout          10/7 10/4                    Eq.:  -0.834 ( -0.047)
       0.177 0.009 0.000 - 0.823 0.206 0.011 CL  -0.921 CF  -0.834
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.006]
       
    6. Rollout          13/7 8/5                     Eq.:  -0.839 ( -0.052)
       0.173 0.010 0.001 - 0.827 0.201 0.012 CL  -0.925 CF  -0.839
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.005]
       
    7. Rollout          10/4 8/5                     Eq.:  -0.845 ( -0.058)
       0.167 0.007 0.000 - 0.833 0.191 0.011 CL  -0.927 CF  -0.845
      [0.002 0.001 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.005]
       
    8. Rollout          8/2 6/3                      Eq.:  -0.864 ( -0.077)
       0.168 0.007 0.000 - 0.832 0.205 0.011 CL  -0.944 CF  -0.864
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.006]
       
    9. Rollout          8/5 8/2                      Eq.:  -0.889 ( -0.102)
       0.159 0.006 0.000 - 0.841 0.209 0.012 CL  -0.971 CF  -0.889
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.005]
       
   10. Rollout          8/2 5/2                      Eq.:  -0.891 ( -0.104)
       0.161 0.006 0.000 - 0.839 0.212 0.012 CL  -0.972 CF  -0.891
      [0.001 0.000 0.000 - 0.001 0.002 0.001 CL   0.004 CF   0.005]
       
   11. Rollout          10/7 8/2                     Eq.:  -0.893 ( -0.106)
       0.158 0.005 0.000 - 0.842 0.209 0.011 CL  -0.971 CF  -0.893
      [0.002 0.000 0.000 - 0.002 0.002 0.001 CL   0.004 CF   0.005]
     
   12. Rollout          13/10 8/2                    Eq.:  -0.896 ( -0.109)
       0.157 0.006 0.000 - 0.843 0.210 0.011 CL  -0.976 CF  -0.896
      [0.001 0.000 0.000 - 0.001 0.002 0.001 CL   0.004 CF   0.005]
       
   13. Rollout          10/4 5/2                     Eq.:  -0.905 ( -0.118)
       0.154 0.004 0.000 - 0.846 0.209 0.011 CL  -0.982 CF  -0.905
      [0.001 0.000 0.000 - 0.001 0.002 0.001 CL   0.004 CF   0.005]
       
   14. Rollout          13/7 5/2                     Eq.:  -0.915 ( -0.128)
       0.147 0.003 0.000 - 0.853 0.204 0.011 CL  -0.990 CF  -0.915
      [0.001 0.001 0.000 - 0.001 0.002 0.001 CL   0.004 CF   0.005]
       
   15. Rollout          10/1                         Eq.:  -0.934 ( -0.147)
       0.143 0.004 0.000 - 0.857 0.214 0.014 CL  -1.010 CF  -0.934
      [0.001 0.000 0.000 - 0.001 0.002 0.001 CL   0.004 CF   0.005]
       
   16. Rollout          24/15*                       Eq.:  -1.031 ( -0.244)
       0.250 0.038 0.003 - 0.750 0.509 0.024 CL  -1.158 CF  -1.031
      [0.002 0.001 0.001 - 0.002 0.003 0.001 CL   0.007 CF   0.010]
        Full cubeful rollout with var.redn.
        1216 games, Mersenne Twister dice gen. with seed 889392929 and quasi-random dice
        Play: 0-ply cubeful prune [expert]
        Cube: 0-ply cubeful prune [expert]

[close]

[Zorba]

and here's the XG rollout:

Spoiler

I also made the double whopper hit... I had expected a bigger difference in wins and a smaller one in gammons. I got perfect hindsight though   I guess the thing here is that:

• We have a well-timed acepoint game, that will win about 20% of games at the cost of 20% gammon losses
• Running off that anchor, it's not likely that we'll make it again anytime, so we lose a big part of those wins
• We're not ready to make a checker dance often, or to contain it, so the hit is not at all decisive
• By trying to contain or slap around dorbel's backchecker anyhow, we nearly always run huge gammon risks against his 5pt board
• The 3pt blot is a huge liability right away, but the other two blots should be cleaned up ASAP too, which is hard to do all at once, while also trying to contain or attack dorbel
• Chances at picking up a second blot are very remote

   1. Rollout¹    10/4 6/3                     eq:-0,802
     Player  : 19,07% (G:1,21% B:0,09%)
     Opponent: 80,93% (G:21,17% B:1,53%)
     Confidence: ± 0,006 (-0,808<E<-0,796)
     Duration: 3 minutes 58 seconds

   2. Rollout¹    13/7 6/3                     eq:-0,805 (-0,003)
     Player  : 18,94% (G:1,19% B:0,09%)
     Opponent: 81,06% (G:21,03% B:1,44%)
     Confidence: ± 0,006 (-0,811<E<-0,799)
     Duration: 3 minutes 04 seconds

   3. Rollout¹    13/7 10/7                    eq:-0,815 (-0,014)
     Player  : 18,60% (G:1,17% B:0,09%)
     Opponent: 81,40% (G:21,37% B:1,44%)
     Confidence: ± 0,006 (-0,821<E<-0,809)
     Duration: 2 minutes 55 seconds

   4. Rollout¹    24/15*                       eq:-1,011 (-0,209)
     Player  : 25,45% (G:4,44% B:0,40%)
     Opponent: 74,55% (G:51,28% B:2,74%)
     Confidence: ± 0,008 (-1,019<E<-1,003)
     Duration: 8 minutes 37 seconds


¹ 5184 Games rolled with Variance Reduction.
 Moves: 1 ply, cube decisions: 3 ply Red


eXtreme Gammon Version: 1.14, MET: Rockwell-Kazaross

[close]

[sixty_something]

Spoiler

it is interesting to observe that none of us picked slotting the 4-point, the best move .. also of note is how bad a blunder this was .. i think this is a good example of some next stage of learning .. bots teach us aggression well, but the finesse of when to hold'em and when to fold'em was beyond me on this one

[close]