• race cube

    From rambiz@21:1/5 to All on Thu May 26 10:17:22 2022
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver
    +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    eXtreme Gammon Version: 2.10

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From pepstein5@gmail.com@21:1/5 to rambiz on Thu May 26 11:25:08 2022
    On Thursday, May 26, 2022 at 6:17:23 PM UTC+1, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    I don't see any reason not to Axelise this.
    It's been proven to be the best racing algo, after all.
    Our gap on the 3 point leaves us with the last banana in the fruit bowl
    so we should be prepared to hold if the Axelisation gives an ultra-marginal double.

    The raw pip counts are 51 to 57.
    For me, I add 4 extra for the acepoint stack, and 1 extra for the 2 point stack to get 56 points.
    For the opponent, we add 4 for the stacks and 2 for the crossovers to get 63 points.
    56 + 9 1/3 - 63 gives a very marginal pass.
    As promised, I'll adjust according to the unmeasured criterion to give a take. That's probably would Stick would do.
    In any case, it's my earlier plan and I don't see a reason not to Stick to it. D/T.


    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From pepstein5@gmail.com@21:1/5 to peps...@gmail.com on Thu May 26 12:32:13 2022
    On Thursday, May 26, 2022 at 7:25:10 PM UTC+1, peps...@gmail.com wrote:
    On Thursday, May 26, 2022 at 6:17:23 PM UTC+1, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action
    I don't see any reason not to Axelise this.
    It's been proven to be the best racing algo, after all.
    Our gap on the 3 point leaves us with the last banana in the fruit bowl
    so we should be prepared to hold if the Axelisation gives an ultra-marginal double.

    The raw pip counts are 51 to 57.
    For me, I add 4 extra for the acepoint stack, and 1 extra for the 2 point stack to get 56 points.
    For the opponent, we add 4 for the stacks and 2 for the crossovers to get 63 points.
    56 + 9 1/3 - 63 gives a very marginal pass.
    As promised, I'll adjust according to the unmeasured criterion to give a take. That's probably would Stick would do.
    In any case, it's my earlier plan and I don't see a reason not to Stick to it.
    D/T.

    Sorry 65 1/3 - 63 = 2 1/3 which is already a take even without any adjustment. It Axelises perfectly to a D/T.

    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to peps...@gmail.com on Thu May 26 23:02:32 2022
    "peps...@gmail.com" <pepstein5@gmail.com> writes:

    On Thursday, May 26, 2022 at 6:17:23 PM UTC+1, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10
    For the opponent, we add 4 for the stacks and 2 for the crossovers to
    get 63 points.

    The opponent also has a gap on point 6, so gets another penalty
    pip. Thus we have 56 versus 64.

    p = 80 - 56/3 + 2*(64 - 56)
    = 80 - 18 2/3 + 16
    = 77 1/3

    which is a double and a pass.

    Best regards

    Axel

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Stick Rice@21:1/5 to rambiz on Thu May 26 23:52:00 2022
    On Thursday, May 26, 2022 at 1:17:23 PM UTC-4, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    eXtreme Gammon Version: 2.10

    Monster pass and another EPC position. (one was recently posted on FB and it made me realize how few ppl know how to use it)

    Stick

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to Stick Rice on Fri May 27 10:16:42 2022
    Stick Rice <bananaboater315@gmail.com> writes:

    On Thursday, May 26, 2022 at 1:17:23 PM UTC-4, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    Monster pass and another EPC position. (one was recently posted on FB
    and it made me realize how few ppl know how to use it)

    While my EPC approximation gives 66 for X (GNU Backgammon reports 65.19)
    I still have some questions:

    1. What to do with O's two outfield men? Just use the straight pips for
    them? Then my approximation gives 73 for O (GNU Backgammon does not
    report an EPC directly with outfield checkers, but a one-sided rollout
    gives 72.26)

    2. How to proceed with 66 versus 73? Abusing Trice's doubling criterion
    (which was meant for pips-versus-rolls positions) gives

    66 + 66/(8 1/6) - 3 = 71.xxx

    as point of last take, so it also suggests a pass.

    3. What makes this position an "EPC position"?

    Best regards

    Axel

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From pepstein5@gmail.com@21:1/5 to Stick Rice on Fri May 27 07:22:37 2022
    On Friday, May 27, 2022 at 7:52:02 AM UTC+1, Stick Rice wrote:
    On Thursday, May 26, 2022 at 1:17:23 PM UTC-4, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    eXtreme Gammon Version: 2.10
    Monster pass and another EPC position. (one was recently posted on FB and it made me realize how few ppl know how to use it)

    Stick
    I like the fact that this position also Axelises correctly. Somehow, I managed to mess up the algo when I tried to apply it from memory.
    I forgot that missing high points are penalised even when they don't appear in gap formation.

    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to peps...@gmail.com on Fri May 27 17:02:33 2022
    "peps...@gmail.com" <pepstein5@gmail.com> writes:

    I forgot that missing high points are penalised even when they don't
    appear in gap formation.

    Yes. Like the Keith count, just penalize empty space. Unlike the Keith
    count, do this in a relative way. Easier. (-:

    Normally I use a further shortcut for points 4 to 6: Just count the
    TOTAL number of gaps there per player and penalize the player who has
    more with 1 pip per additional gap. So lump the gaps together instead of treating them individually. This loses about 4 per mille of accuracy,
    but gains about 4 per cent effort. I find this also less confusing.

    Of course I tested "true" gaps and was surprised to find that the Isight
    method then performs worse than the Keith count. So "empty space" is an important feature and used by all variants in table 4 on page 17 of my
    article.

    Best regards

    Axel

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From pepstein5@gmail.com@21:1/5 to rambiz on Sat May 28 04:33:03 2022
    On Saturday, May 28, 2022 at 12:16:41 PM UTC+1, rambiz wrote:
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    eXtreme Gammon Version: 2.10
    Hi All,
    Thanks to all the contributors.
    I posted this position because not only Trice and Keith get it wrong but also because the taker has only %17.
    I hoped to find out why the two formulae are so far off.

    Stick said it is an EPC position. Well, fine but what to do about it, please? And as Axel already asked what makes it an EPC position in the first place, please?

    Now to the iSight method... I normally don't use it, but to make sure we are on the same page let me apply iSight and please follow my calculations and tell me specifically if I am right to penalise the on roll player by another one pip for having a
    gap on his three point.

    On roll players count=51 (Raw pip count)+4 (2 extra checkers on the ace)+1 (1 extra checker on the deuce)+1(gap on three)=57
    Takers Count=57(raw pip count)+2(1 extra checker on the ace)+2(2 extra checkers on the deuce )+1(gap on 6)+2(two extra cross-overs)=64

    Now the winning chance for the player on roll according to iSight=80-(57/3)+2(64-57)=80-19+14=75%
    So it should be a take according to the iSight article page 23.

    The gap on the three is not penalized. You say "Tell me specifically if I am right to penalise"
    the gap on the three point. But these things don't need to be said explicitly. They are stated by omission.
    Since the doc never mentions a gap on the three point, the gap can't be penalised.
    Axel also didn't say whether there is a penalty if your raw pip count is a multiple of 29 or if the day of the month
    in which you're playing matches your raw pip count, or whether you deserve a fashion penalty for having a 1970's
    hairstyle.
    When you correct this error, you get the correct drop.
    It's a drop according to the article, as has already been pointed out on this thread.

    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From rambiz@21:1/5 to All on Sat May 28 04:40:04 2022
    Thank you Paul for clarification.
    As mentioned I am really new to iSight and just overflew the article.

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From pepstein5@gmail.com@21:1/5 to rambiz on Sat May 28 04:46:17 2022
    On Saturday, May 28, 2022 at 12:40:06 PM UTC+1, rambiz wrote:
    Thank you Paul for clarification.
    As mentioned I am really new to iSight and just overflew the article.

    You're welcome.
    I also messed up the algo to get an incorrect take, because I forgot to count the gap on the 6 point.
    It's great that this algo works here when the others don't.

    I'd encourage a bit of common-sense adjustment, particularly if the verdict is close.
    In other words, if you are somewhat experienced, I would give at least a little bit of weight
    to your prior intuition.
    So, if your initial reaction is "This really looks like a drop" and the Axelisation computes to 75.8
    instead of 76, count it as a drop anyway. Fortunately, we don't need to use such intuition here.

    Paul

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From rambiz@21:1/5 to All on Sat May 28 04:16:40 2022
    XGID=-CDBBB----A-A------ccb-cd-:0:0:-1:00:0:0:3:0:10

    X:Player 2 O:Player 1
    Score is X:0 O:0. Unlimited Game, Jacoby Beaver +13-14-15-16-17-18------19-20-21-22-23-24-+
    | O O | | O O O O O |
    | | | O O O O O |
    | | | O O |
    | | | O |
    | | | |
    | |BAR| |
    | | | |
    | | | X |
    | | | X X X X |
    | | | X X X X X |
    | | | X X X X X |
    +12-11-10--9--8--7-------6--5--4--3--2--1-+
    Pip count X: 51 O: 57 X-O: 0-0
    Cube: 1
    X on roll, cube action

    eXtreme Gammon Version: 2.10

    Hi All,
    Thanks to all the contributors.
    I posted this position because not only Trice and Keith get it wrong but also because the taker has only %17.
    I hoped to find out why the two formulae are so far off.

    Stick said it is an EPC position. Well, fine but what to do about it, please? And as Axel already asked what makes it an EPC position in the first place, please?

    Now to the iSight method... I normally don't use it, but to make sure we are on the same page let me apply iSight and please follow my calculations and tell me specifically if I am right to penalise the on roll player by another one pip for having a gap
    on his three point.

    On roll players count=51 (Raw pip count)+4 (2 extra checkers on the ace)+1 (1 extra checker on the deuce)+1(gap on three)=57
    Takers Count=57(raw pip count)+2(1 extra checker on the ace)+2(2 extra checkers on the deuce )+1(gap on 6)+2(two extra cross-overs)=64

    Now the winning chance for the player on roll according to iSight=80-(57/3)+2(64-57)=80-19+14=75%
    So it should be a take according to the iSight article page 23.

    Regrads

    --- SoupGate-Win32 v1.05
    * Origin: fsxNet Usenet Gateway (21:1/5)
  • From Axel Reichert@21:1/5 to rambiz on Sat May 28 21:25:53 2022
    rambiz <rambiz@gmail.com> writes:

    As mentioned I am really new to iSight and just overflew the article.

    It is "Isight" (the name of the optimization software I used to find
    it), not "iSight" (Apple cameras).

    Best regards

    Axel

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