1. Forum
  2. Usenet
  3. REC.GAMES.BACKGAMMON

  • All 33s are equal, but one 33 is more equal

    From Axel Reichert@21:1/5 to All on Thu Feb 9 07:32:40 2023
    Like in the opening, this 33 offers many choices.
    Like in the opening, these look far from obvious to me.
    In contrast to the opening the equity differences are much larger.
    So I could have labelled this quiz a "33 amplifier".

    GNU Backgammon Position ID: zE7wATCYZ/ABMA
    Match ID : cIltASAAGAAE
    +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
    | X O O O | | O X | 2 points
    | X O O O | | X |
    | O | | X |
    | | | X |
    | | | X |
    | |BAR| |v 11 point match (Cube: 1)
    | | | O |
    | X | | O |
    | X | | O |
    | O X X | | X O | Rolled 33
    | O X X | | X O | 3 points
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: axel
    Pip counts: O 155, X 161

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Timothy Chow@21:1/5 to Axel Reichert on Thu Feb 9 09:06:16 2023
    On 2/9/2023 1:32 AM, Axel Reichert wrote:
    Like in the opening, this 33 offers many choices.
    Like in the opening, these look far from obvious to me.
    In contrast to the opening the equity differences are much larger.
    So I could have labelled this quiz a "33 amplifier".

    GNU Backgammon Position ID: zE7wATCYZ/ABMA
    Match ID : cIltASAAGAAE
    +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
    | X O O O | | O X | 2 points
    | X O O O | | X |
    | O | | X |
    | | | X |
    | | | X |
    | |BAR| |v 11 point match (Cube: 1)
    | | | O |
    | X | | O |
    | X | | O |
    | O X X | | X O | Rolled 33
    | O X X | | X O | 3 points
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: axel
    Pip counts: O 155, X 161

    I don't think the score makes much difference here.

    O's prime is getting a bit dangerous, so I'd start with 24/21(2).
    After that, it's a choice between 13/10(2) and 8/5(2) for me.
    O's blot makes an argument for 8/5(2), strengthening our board,
    but it comes at the expense of our 8pt, so I would play 13/10(2).

    Looking again, I see that 8/5(2) 6/3(2) is a tempting alternative
    since it makes such a strong board. If O's position were weaker,
    this is probably what I'd play. But 24/21(2) seems necessary to me
    in this position.

    ---
    Tim Chow

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Stick Rice@21:1/5 to Axel Reichert on Thu Feb 9 11:47:39 2023
    On Thursday, February 9, 2023 at 1:32:42 AM UTC-5, Axel Reichert wrote:
    Like in the opening, this 33 offers many choices.
    Like in the opening, these look far from obvious to me.
    In contrast to the opening the equity differences are much larger.
    So I could have labelled this quiz a "33 amplifier".

    GNU Backgammon Position ID: zE7wATCYZ/ABMA
    Match ID : cIltASAAGAAE
    +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
    | X O O O | | O X | 2 points
    | X O O O | | X |
    | O | | X |
    | | | X |
    | | | X |
    | |BAR| |v 11 point match (Cube: 1)
    | | | O |
    | X | | O |
    | X | | O |
    | O X X | | X O | Rolled 33
    | O X X | | X O | 3 points
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: axel
    Pip counts: O 155, X 161

    When your opponent has made it so there's basically only one good anchor left to have and you roll a number that makes that anchor you should probably take it right? If not you risk your opponent making your 8pt and never finding the right dice to split
    safely or run.

    After making the anchor of course solidifying your board, points in order, is the other half of the equation. (5pt) It makes it so your opponent has a hard time splitting not only because he needs to find the numbers to do so but when he does you'll be
    able to attack thanks to that anchor on the other side of the board. By making the 5pt if he runs into the outfield trying to do something with his now awkwardly hemmed in back checkers and you hit that checker he'll not only be facing a strong board,
    but a bit of an overwhelming offensive position as he can only come in on the lower points. (and presumably you'll have multiple, like 3, builders in the outfield so he won't be able to hop again and they'll be poised to make the bar point or the 8pt)

    Stick

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Robert Zimmerman@21:1/5 to Stick Rice on Thu Feb 9 17:39:13 2023
    On Thursday, February 9, 2023 at 2:48:10 PM UTC-5, Stick Rice wrote:
    On Thursday, February 9, 2023 at 1:32:42 AM UTC-5, Axel Reichert wrote:
    Like in the opening, this 33 offers many choices.
    Like in the opening, these look far from obvious to me.
    In contrast to the opening the equity differences are much larger.
    So I could have labelled this quiz a "33 amplifier".

    GNU Backgammon Position ID: zE7wATCYZ/ABMA
    Match ID : cIltASAAGAAE
    +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
    | X O O O | | O X | 2 points
    | X O O O | | X |
    | O | | X |
    | | | X |
    | | | X |
    | |BAR| |v 11 point match (Cube: 1)
    | | | O |
    | X | | O |
    | X | | O |
    | O X X | | X O | Rolled 33
    | O X X | | X O | 3 points
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: axel
    Pip counts: O 155, X 161
    When your opponent has made it so there's basically only one good anchor left to have and you roll a number that makes that anchor you should probably take it right? If not you risk your opponent making your 8pt and never finding the right dice to
    split safely or run.

    After making the anchor of course solidifying your board, points in order, is the other half of the equation. (5pt) It makes it so your opponent has a hard time splitting not only because he needs to find the numbers to do so but when he does you'll be
    able to attack thanks to that anchor on the other side of the board. By making the 5pt if he runs into the outfield trying to do something with his now awkwardly hemmed in back checkers and you hit that checker he'll not only be facing a strong board,
    but a bit of an overwhelming offensive position as he can only come in on the lower points. (and presumably you'll have multiple, like 3, builders in the outfield so he won't be able to hop again and they'll be poised to make the bar point or the 8pt)

    Stick

    21/24 (2) is required. Last call at Club 21.
    8 points alone provide good service when they have their objective in sight, ie a split or anchor on the two point. Otherwise they're neither the rear guard nor the front line. They have to be prodded into action. 8/5 (2).

    21/24 (2) 8/5 (2).

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to Axel Reichert on Wed Feb 15 08:34:35 2023
    Axel Reichert <[email protected]> writes:

    Like in the opening, this 33 offers many choices.
    Like in the opening, these look far from obvious to me.
    In contrast to the opening the equity differences are much larger.
    So I could have labelled this quiz a "33 amplifier".

    GNU Backgammon Position ID: zE7wATCYZ/ABMA
    Match ID : cIltASAAGAAE
    +24-23-22-21-20-19------18-17-16-15-14-13-+ O: gnubg
    | X O O O | | O X | 2 points
    | X O O O | | X |
    | O | | X |
    | | | X |
    | | | X |
    | |BAR| |v 11 point match (Cube: 1)
    | | | O |
    | X | | O |
    | X | | O |
    | O X X | | X O | Rolled 33
    | O X X | | X O | 3 points
    +-1--2--3--4--5--6-------7--8--9-10-11-12-+ X: axel
    Pip counts: O 155, X 161

    Still amazed by the result that the innocently looking 13/10(2) 8/5(2)
    or 13/7(2) are quadruple whoppers.

    1. Rollout 24/21(2) 8/5(2) Eq.: +0.624
    0.594 0.202 0.008 - 0.406 0.092 0.003 CL +0.353 CF +0.624
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.002 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    565 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    2. Rollout 24/21(2) 6/3(2) Eq.: +0.451 (-0.173)
    0.567 0.170 0.007 - 0.433 0.097 0.003 CL +0.253 CF +0.451
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    509 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    3. Rollout 24/21(2) 13/10(2) Eq.: +0.446 (-0.177)
    0.568 0.167 0.006 - 0.432 0.092 0.003 CL +0.255 CF +0.446
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.002 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    514 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    4. Rollout 8/5(2) 6/3(2) Eq.: +0.438 (-0.185)
    0.566 0.211 0.013 - 0.434 0.148 0.007 CL +0.254 CF +0.438
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.002 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    982 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    5. Rollout 24/18(2) Eq.: +0.377 (-0.247)
    0.560 0.144 0.005 - 0.440 0.107 0.002 CL +0.195 CF +0.377
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    340 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    6. Rollout 24/21 13/10 8/5(2) Eq.: +0.268 (-0.355)
    0.522 0.196 0.013 - 0.478 0.167 0.009 CL +0.129 CF +0.268
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.002 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    598 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    7. Rollout 24/21(2) 13/7 Eq.: +0.244 (-0.380)
    0.519 0.145 0.005 - 0.481 0.109 0.004 CL +0.114 CF +0.244
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    528 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    8. Rollout 13/7(2) Eq.: +0.211 (-0.412)
    0.515 0.174 0.010 - 0.485 0.165 0.009 CL +0.085 CF +0.211
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    610 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    9. Rollout 13/10(2) 8/5(2) Eq.: +0.211 (-0.413)
    0.513 0.184 0.011 - 0.487 0.167 0.009 CL +0.093 CF +0.211
    [0.001 0.001 0.000 - 0.001 0.001 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    572 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune
    10. Rollout 24/18 8/5(2) Eq.: +0.182 (-0.442)
    0.514 0.175 0.012 - 0.486 0.182 0.008 CL +0.072 CF +0.182
    [0.001 0.001 0.000 - 0.001 0.002 0.000 CL 0.003 CF 0.005]
    Truncated cubeful rollout (depth 7) with variance reduction
    465 games, Mersenne Twister dice gen. with seed 449002682 and quasi-random dice
    Stop when std.errs. are small enough: limit 0.0050 (min. 36 games)
    Play: 1-ply cubeful prune
    keep the first 0 0-ply moves and up to 8 more moves within equity 0.16
    Cube: 1-ply cubeful prune
    Different evaluations after 2 plies:
    Play: 0-ply cubeful [expert]
    Cube: 1-ply cubeful prune

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to Stick Rice on Wed Feb 15 08:43:17 2023
    Stick Rice <[email protected]> writes:

    When your opponent has made it so there's basically only one good
    anchor left to have and you roll a number that makes that anchor you
    should probably take it right?

    That's a good one, thanks.

    After making the anchor of course solidifying your board, points in
    order, is the other half of the equation. (5pt) It makes it so your
    opponent has a hard time splitting not only because he needs to find
    the numbers to do so but when he does you'll be able to attack thanks
    to that anchor on the other side of the board. By making the 5pt if
    he runs into the outfield trying to do something with his now
    awkwardly hemmed in back checkers and you hit that checker he'll not
    only be facing a strong board, but a bit of an overwhelming offensive position as he can only come in on the lower points. (and presumably
    you'll have multiple, like 3, builders in the outfield so he won't be
    able to hop again and they'll be poised to make the bar point or the
    8pt)

    Probably the most extensive explanation from you for years, many
    thanks. This is far more helpful than your typical "twitter length"
    answers like "Isn't this obvious?" ... (-:

    No, it normally is not obvious for a weaker player like me, which is why
    I am asking. Of course I can ask the bot for perhaps an even better
    answer in terms of numbers, but the bot's explanation is even terser.

    I fully understand your being terse, which is why I am grateful for
    every bit of insight.

    Best regards

    Axel

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)