Ace point game bear-in 2-1 to play 1a4a cube at 2

Walt

unread,

May 17, 2013, 6:36:41 PM5/17/13

to

XGID=-c-BCBCBBA-----a--abbbbb--:1:1:1:21:4:1:0:5:10

X:You O:O
Score is X:4 O:1 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O O O O O |
| | | O O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X O | +---+
| X X | | X X X X O | | 2 |
| X X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 85 O: 129 X-O: 4-1/5
Cube: 2, X own cube
X to play 21

Bradley K. Sherman

unread,

May 17, 2013, 6:39:16 PM5/17/13

to

Walt <walt_...@yahoo.com> wrote:
>Score is X:4 O:1 5 pt.(s) match.
> +13-14-15-16-17-18------19-20-21-22-23-24-+
> | O O | | O O O O O |
> | | | O O O O O |
> | | | |
> | | | |
> | | | |
> | |BAR| |
> | | | |
> | | | |
> | | | X X O | +---+
> | X X | | X X X X O | | 2 |
> | X X X | | X X X X O | +---+
> +12-11-10--9--8--7-------6--5--4--3--2--1-+

> X:85 O:129, X to play 21

4/2,6/5

--bks

Tim Chow

unread,

May 17, 2013, 8:12:21 PM5/17/13

to

On May 17, 6:39 pm, b...@panix.com (Bradley K. Sherman) wrote:

That's what I'd play, too. If we don't fill in the 2pt then O is
playing a sort of phantom backgame. Slot the 2pt now when we still
have enough timing that we're highly likely to win the priming battle
even if we're hit.

---
Tim Chow

badgolferman

unread,

May 18, 2013, 12:18:49 AM5/18/13

to

I'll play 9/8, 4/2. Hoping to dance a few turns if hit.

peps...@gmail.com

unread,

May 18, 2013, 4:47:33 AM5/18/13

to

Yes, filling in the 2 point is obvious. "Even if we're hit" is probably misleading. Since X has a full prime, X wants to be hit which forces O to destroy O's prime in order to let X in. There's no downside at all to slotting the 2, which makes it obvious.

However, I didn't think the talk show guests meant to talk about 4/2. The debate is solely about how to play the ace. This is where I find myself as clueless as an ant in a bookshop, totally unable to discriminate between 9/8 and 6/5.
6/5 seems slightly more natural because X is spread on more points, leaving fewer possibilities for shot-leaving sequences and making it easier to fill in the 2 point. But I'm sure there's very little equity at stake.

Walt, I hope you don't do an XG rollout showing that 9/8 wins by 0.000001 with a confidence interval of 0.01, and then provide commentary about how amazing it is that XG "prefers" the counter-intuitive play.

Paul Epstein

Grunty

unread,

May 18, 2013, 1:20:26 PM5/18/13

to

I'd play 8/6 8/7.
This play looks good towards both creating builders for the 2-point
and killing sixes over the coming turns, while O crunches his board.

Tim Chow

unread,

May 18, 2013, 2:22:22 PM5/18/13

to

On May 18, 1:20 pm, Grunty <gruntingdw...@yahoo.com> wrote:
> I'd play 8/6 8/7.
> This play looks good towards both creating builders for the 2-point
> and killing sixes over the coming turns, while O crunches his board.

Though there's certainly some logic to this play, I'm fairly sure it's
not the right idea while X still has that gap on the 2pt. This
reminds me of a similar position in Chris Bray's book "Wind Assisted";
I'll see if I can dig it up.

Paul Epstein wrote:
> "Even if we're hit" is probably misleading. Since X has a full prime,
> X wants to be hit which forces O to destroy O's prime in order to let
> X in. There's no downside at all to slotting the 2, which makes it obvious.

This is also a good question...if X plays 4/2 and then O rolls an ace,
should O hit? I think she probably does want to. It looks to me that
her best hope is to counter-prime. Going further, if X knew ahead of
time that O were going to roll an ace, should X still play 4/2? My
guess is no.

---
Tim Chow

Tim Chow

unread,

May 18, 2013, 6:24:18 PM5/18/13

to

On May 18, 2:22 pm, I wrote:
> This reminds me of a similar position in Chris Bray's book "Wind Assisted";
> I'll see if I can dig it up.

I found the position in Bray's book, on page 102. See below. Bray's
position is not as similar to Walt's position as I had thought,
because X has two checkers trapped behind O's six-prime, but that's
just as well, because that means I can post it without necessarily
"spoiling" Walt's position. (I still haven't looked at the computer's
opinion of Walt's position, by the way.) Bray said that in an earlier
book of his, he stated that killing 6's with 12/7 11/7 was the correct
play.

XGID=-b-BCBBB---AA-----cbbbbbB-:1:-1:1:45:0:0:0:0:10
X:Player 1 O:Player 2

Score is X:0 O:0. Money session
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O | | O O O O O X | +---+
| O | | O O O O O X | | 2 |
| O | | | +---+
| | | |
| | | |
| |BAR| |

| | | |
| | | |
| | | X |
| X | | X X X X O |

| X X X | | X X X X O |

+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pipcount X: 125 O: 109 X-O: 0-0
Cube : 2, O own cube
X to play 45

1. Rollout: 11/2 eq:+0.148
Player : 62.14% (G:15.81% B:0.88%)
Opponent: 37.86% (G:12.41% B:1.14%)
Confidence: ±0.011 (+0.137<E<+0.159)

2. Rollout: 12/8 11/6 eq:+0.048 (-0.101)
Player : 57.50% (G:14.42% B:0.82%)
Opponent: 42.50% (G:11.22% B:0.75%)
Confidence: ±0.010 (+0.038<E<+0.058)

3. Rollout: 12/3 eq:+0.027 (-0.121)
Player : 56.47% (G:14.25% B:0.80%)
Opponent: 43.53% (G:11.14% B:0.76%)
Confidence: ±0.011 (+0.016<E<+0.038)

4. Rollout: 12/7 11/7 eq:+0.024 (-0.124)
Player : 56.37% (G:14.05% B:0.77%)
Opponent: 43.63% (G:11.12% B:0.73%)
Confidence: ±0.011 (+0.013<E<+0.035)

Truncated at 7 moves using 3 ply.
648 Games rolled with Variance Reduction.
Dice Seed: 2
First 2 moves and cube decisions: 3 ply
Remaining moves: 2 ply, cube decisions: 3 ply

eXtreme Gammon Version: 1.21

---
Tim Chow

Walt

unread,

May 20, 2013, 10:13:56 AM5/20/13

to

JFTR, I'd like to reiterate that I don't do QF (at least not very often,
and when I do I'll try to warn folks). This problem was posed to
discuss the pros and cons of slotting the two point, which I didn't find
obvious.

X is just trying to bring this one home - gammons are useless, so
there's no need to try something fancy. Slotting the deuce risks being
trapped behind O's 5-prime and watching as X's board crunches. If O's
board wasn't so strong slotting would be obvious, but here there are
real risks to the slot.

I played 9/7 6/5 to aim four builders at the two point in hopes of
simply making it cleanly with the ~15 shakes that make it. But i was
overly cautious. After an exchange of hits, X has better timing and
only faces a 5-prime while holding a full prime himself, so X is still
in very good shape after the hit.

As for what to do with the ace after slotting the two, 9/8 brings
another checker in range and allows X to cover so that's why i think
it's slightly better. XG fails to distinguish between the two.

All that said, shuffling rather than slotting is not that big of an
error (.027). Variants follow below.

XGID=-c-BCBCBBA-----a--abbbbb--:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2

Score is X:4 O:1 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O O O O O |
| | | O O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X O | +---+
| X X | | X X X X O | | 2 |
| X X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 85 O: 129 X-O: 4-1/5
Cube: 2, X own cube
X to play 21

1. Rollout� 9/8 4/2 eq:+0.793
Player: 90.03% (G:37.87% B:9.82%)
Opponent: 9.97% (G:0.34% B:0.05%)
Confidence: �0.004 (+0.789..+0.797) - [57.3%]
Duration: 5 minutes 24 seconds

2. Rollout� 6/5 4/2 eq:+0.793 (-0.001)
Player: 89.95% (G:37.87% B:9.78%)
Opponent: 10.05% (G:0.29% B:0.05%)
Confidence: �0.004 (+0.789..+0.797) - [42.7%]
Duration: 4 minutes 37 seconds

3. Rollout� 9/7 6/5 eq:+0.766 (-0.027)
Player: 88.71% (G:34.54% B:7.58%)
Opponent: 11.29% (G:0.39% B:0.02%)
Confidence: �0.004 (+0.762..+0.770) - [0.0%]
Duration: 2 minutes 59 seconds

4. Rollout� 9/7 4/3 eq:+0.761 (-0.032)
Player: 88.49% (G:34.40% B:7.08%)
Opponent: 11.51% (G:0.40% B:0.01%)
Confidence: �0.004 (+0.758..+0.765) - [0.0%]
Duration: 2 minutes 23 seconds

5. Rollout� 9/6 eq:+0.754 (-0.039)
Player: 88.19% (G:33.40% B:7.66%)
Opponent: 11.81% (G:0.46% B:0.02%)
Confidence: �0.004 (+0.751..+0.758) - [0.0%]
Duration: 2 minutes 04 seconds

� 1296 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

////////////Variant 1//////////////////////////////

If we chow on the position a bit, giving O a full prime and some extra
timing by moving one of the trapped checkers out, the slot is too risky:

XGID=-b-BCBCBBAa-------bbbbbb--:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2

Score is X:4 O:1 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O | | O O O O O |
| O | | O O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |

| | | X X | +---+

| X X | | X X X X O | | 2 |

| O X X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 85 O: 117 X-O: 4-1/5

Cube: 2, X own cube
X to play 21

1. XG Roller++ 8/7 8/6 eq:+0.723
Player: 86.59% (G:11.50% B:0.33%)
Opponent: 13.41% (G:0.44% B:0.01%)

...

5. XG Roller++ 9/8 4/2 eq:+0.677 (-0.045)
Player: 85.01% (G:15.32% B:0.71%)
Opponent: 14.99% (G:1.09% B:0.05%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

/////////////////////variant 2/////////////////////

Just giving O some timing without the full prime is enough to swing it
to no slot.

XGID=-b-BCBCBBAa----a--abbbbb--:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2

| | | X X | +---+

| X X | | X X X X O | | 2 |

| O X X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 85 O: 120 X-O: 4-1/5

Cube: 2, X own cube
X to play 21

1. XG Roller++ 9/7 6/5 eq:+0.693
Player: 85.30% (G:11.70% B:0.38%)
Opponent: 14.70% (G:0.63% B:0.02%)

2. XG Roller++ 9/8 4/2 eq:+0.665 (-0.028)
Player: 84.43% (G:15.61% B:0.77%)
Opponent: 15.57% (G:1.13% B:0.05%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

//////////variant 3////////////////

the crucial factor appears to be the fact that O has three checkers
back, not the timing in and of itself. with 3 checkers back and lots of
timing the slot is still correct.

XGID=-c-BCBCBCb---------bbbbb--:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2

| |BAR| |

| | | |
| | | |
| X | | X X O | +---+

| O X X | | X X X X O | | 2 |
| O X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 84 O: 144 X-O: 4-1/5

Cube: 2, X own cube
X to play 21

1. XG Roller++ 8/7 4/2 eq:+0.690
Player: 85.19% (G:34.09% B:2.94%)
Opponent: 14.81% (G:0.67% B:0.03%)

2. XG Roller++ 8/5 eq:+0.639 (-0.050)
Player: 82.59% (G:32.33% B:2.65%)
Opponent: 17.41% (G:0.59% B:0.02%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

--
//Walt

Tim Chow

unread,

May 20, 2013, 11:18:34 AM5/20/13

to

Nice variants! Very instructive.

On May 20, 10:13 am, Walt <walt_ask...@yahoo.com> wrote:
> If O's board wasn't so strong slotting would be obvious, but here
> there are real risks to the slot.

This might seem like a quibble, but what's important is that O's
*prime* is strong, not that O's *board* is strong. If you weaken O's
board by shifting her prime backwards, but improve it to a six-prime,
then I believe that O's position gets better. It'd be interesting to
see what happens in a position like this one, for example:

+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O O O O |
| O O | | O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |

| X | | X X O | +---+

| X X | | X X X X O | | 2 |

| X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+

> As for what to do with the ace after slotting the two, 9/8 brings
> another checker in range and allows X to cover so that's why i think
> it's slightly better. XG fails to distinguish between the two.

Playing 8/2 with a 6 maintains a full six-prime so I don't think that
9/8 "brings another checker in range and allows X to cover." If
anything, leaving it on the 9pt allows you to cover with 52.

---
Tim Chow

Walt

unread,

May 20, 2013, 12:08:39 PM5/20/13

to

On 5/20/2013 11:18 AM, Tim Chow wrote:
> Nice variants! Very instructive.
>
> On May 20, 10:13 am, Walt <walt_ask...@yahoo.com> wrote:
>> If O's board wasn't so strong slotting would be obvious, but here
>> there are real risks to the slot.
>
> This might seem like a quibble, but what's important is that O's
> *prime* is strong, not that O's *board* is strong.

I was being somewhat imprecise in my usage here. By "board' I meant the
totality of O's structure on her side of the board not just how many
points she held in her homeboard. Agree that it's the prime more than
the board that X should be afraid of.

>If you weaken O's
> board by shifting her prime backwards, but improve it to a six-prime,
> then I believe that O's position gets better. It'd be interesting to
> see what happens in a position like this one, for example:
>
> +13-14-15-16-17-18------19-20-21-22-23-24-+
> | O O | | O O O O |
> | O O | | O O O O |
> | | | |
> | | | |
> | | | |
> | |BAR| |
> | | | |
> | | | |
> | X | | X X O | +---+
> | X X | | X X X X O | | 2 |
> | X X | | X X X X O | +---+
> +12-11-10--9--8--7-------6--5--4--3--2--1-+

This looks even more dangerous for X - almost twice as many numbers
enter, and O has a bit more timing in addition to the full prime. I
wouldn't slot here.

(sets up the position in XG....)

Still correct to slot, but not by much. (and possibly the shuffling play
would come out on top in a rollout.

XGID=-c-BCBCBC--------bbbbbb---:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2
Score is X:4 O:1 5 pt.(s) match.

+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O O O O |
| O O | | O O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| X | | X X O | +---+
| X X | | X X X X O | | 2 |
| X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+

Pip count X: 84 O: 138 X-O: 4-1/5

Cube: 2, X own cube
X to play 21

1. XG Roller++ 8/7 4/2 eq:+0.702
Player: 85.97% (G:34.50% B:3.07%)
Opponent: 14.03% (G:0.80% B:0.04%)

2. XG Roller++ 8/5 eq:+0.697 (-0.006)
Player: 85.29% (G:31.71% B:2.44%)
Opponent: 14.71% (G:0.44% B:0.02%)

eXtreme Gammon Version: 2.10, MET: Kazaross XG2

How about if we move the prime back another point? As expected, with X
75% to enter and O having a bit more timing, slotting is a clear error.

XGID=-c-BCBCBC-------bbbbbb----:1:1:1:21:4:1:0:5:10

X:Player 1 O:Player 2
Score is X:4 O:1 5 pt.(s) match.

+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O O | | O O O |
| O O O | | O O O |
| | | |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| X | | X X O | +---+
| X X | | X X X X O | | 2 |
| X X | | X X X X O | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+

Pip count X: 84 O: 150 X-O: 4-1/5

Cube: 2, X own cube
X to play 21

1. XG Roller++ 8/5 eq:+0.604
Player: 80.83% (G:33.97% B:3.22%)
Opponent: 19.17% (G:0.60% B:0.03%)

2. XG Roller++ 8/7 4/2 eq:+0.556 (-0.048)
Player: 79.48% (G:32.93% B:3.14%)
Opponent: 20.52% (G:1.61% B:0.08%)

3. XG Roller++ 7/6 7/5 eq:+0.554 (-0.050)
Player: 78.42% (G:31.96% B:3.14%)
Opponent: 21.58% (G:0.70% B:0.03%)

4. XG Roller++ 8/6 4/3 eq:+0.550 (-0.054)
Player: 78.63% (G:32.66% B:3.04%)
Opponent: 21.37% (G:1.08% B:0.06%)

5. XG Roller++ 6/5 4/2 eq:+0.528 (-0.076)
Player: 78.18% (G:32.50% B:3.15%)
Opponent: 21.82% (G:1.71% B:0.09%)

Freeven

unread,

May 21, 2013, 3:08:31 PM5/21/13

to

I'm late to the party, and I've already looked at the rollout, but here
are some additional thoughts.

On 5/20/2013 10:13 AM, Walt wrote:
> After an exchange of hits, X has better timing and
> only faces a 5-prime while holding a full prime himself, so X is still
> in very good shape after the hit.

Also of significance is that O isn't at the edge of X's prime, so X's
position will remain quite strong even after he breaks his 8pt. This
adds a lot to his timing advantage.

> As for what to do with the ace after slotting the two, 9/8 brings
> another checker in range and allows X to cover so that's why i think
> it's slightly better. XG fails to distinguish between the two.

The big advantage I see for 9/8 is that it doesn't tear up X's position
on a subsequent 44. If X plays 6/5, he'll be forced to break his prime
on 44, meaning that a 6 now helps O's timing rather than hurting it. I'd
have missed this over the board and played 6/5. (I likely would have
played 6/5 even if I'd seen it.)

> XGID=-c-BCBCBBA-----a--abbbbb--:1:1:1:21:4:1:0:5:10
>
> X:Player 1 O:Player 2
> Score is X:4 O:1 5 pt.(s) match.
> +13-14-15-16-17-18------19-20-21-22-23-24-+
> | O O | | O O O O O |
> | | | O O O O O |
> | | | |
> | | | |
> | | | |
> ||BAR| |
> | | | |
> | | | |
> | | | X X O | +---+
> | X X | | X X X X O | | 2 |
> | X X X | | X X X X O | +---+
> +12-11-10--9--8--7-------6--5--4--3--2--1-+
> Pip count X: 85 O: 129 X-O: 4-1/5
> Cube: 2, X own cube
> X to play 21
>

> 1. Rolloutš 9/8 4/2 eq:+0.793

> Player: 90.03% (G:37.87% B:9.82%)
> Opponent: 9.97% (G:0.34% B:0.05%)

> Confidence: ą0.004 (+0.789..+0.797) - [57.3%]

> Duration: 5 minutes 24 seconds
>

> 2. Rolloutš 6/5 4/2 eq:+0.793 (-0.001)

> Player: 89.95% (G:37.87% B:9.78%)
> Opponent: 10.05% (G:0.29% B:0.05%)

> Confidence: ą0.004 (+0.789..+0.797) - [42.7%]