62, play the obvious part and then what?

[BlueDice]

XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1

Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62
--
BD

[Tim Chow]

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62

I would play 23/21 13/7. Making the five-prime with 13/7 is the obvious
part. It is tempting to keep the straggler where it is to make it harder
for O to play safely. However, that would require advancing one of our
outfield checkers. While it would be nice to send a second checker back,
I think the priority is to contain the checker that is already back. So
we want to keep our outfield checkers far back, and 23/21 prepares to bring
a third checker into the outfield.

---
Tim Chow

[Paul Epstein]

Post by Tim Chow

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62

I would play 23/21 13/7. Making the five-prime with 13/7 is the obvious
part. It is tempting to keep the straggler where it is to make it harder
for O to play safely. However, that would require advancing one of our
outfield checkers. While it would be nice to send a second checker back,
I think the priority is to contain the checker that is already back. So
we want to keep our outfield checkers far back, and 23/21 prepares to bring
a third checker into the outfield.

I think Tim might not have come up with the best idea since the invention
of the sandwich. After Tim's play, O has a dream position with sixes
escaping and non-sixes pointing. He's diversifying the opponent.
I'm playing 13/7 13/11 to at least give my opponent a few bad rolls.
All the opponent's fives play badly after my play.

Paul

[BlueDice]

Post by Tim Chow

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62

I would play 23/21 13/7. Making the five-prime with 13/7 is the obvious
part. It is tempting to keep the straggler where it is to make it harder
for O to play safely. However, that would require advancing one of our
outfield checkers. While it would be nice to send a second checker back,
I think the priority is to contain the checker that is already back. So
we want to keep our outfield checkers far back, and 23/21 prepares to bring
a third checker into the outfield.
---
Tim Chow

I can see that containment is a priority...
Making the 5 prime is good for containment.
Staying back with the straggler is good for contact.
I'm struggling to see why those two factors are outweighed by leaving a direct shot and having an extra checker in the outfield?
--
BD

[Tim Chow]

Post by BlueDice
Staying back with the straggler is good for contact.

So I think this is the key point. Contact is a means toward an end. The
end, for X, would be to send another checker back. The point is that
O already has a checker back, which X is pretty well-positioned to contain,
so the incremental value of sending a second checker back is reduced. Also,
the contact doesn't unequivocally favor X. O's spares are flexibly positioned,
so depending on the dice, the contact might favor O (who could attack the
straggler).

I've encountered this theme rather frequently. The computer will often
focus its energies on containing a single checker by bringing the last
checker forward to assist, even though it might superficially seem beneficial
to keep the last checker back "for more contact."

---
Tim Chow

[BlueDice]

Post by Tim Chow

Post by BlueDice
Staying back with the straggler is good for contact.

So I think this is the key point. Contact is a means toward an end. The
end, for X, would be to send another checker back. The point is that
O already has a checker back, which X is pretty well-positioned to contain,
so the incremental value of sending a second checker back is reduced. Also,
the contact doesn't unequivocally favor X. O's spares are flexibly positioned,
so depending on the dice, the contact might favor O (who could attack the
straggler).
I've encountered this theme rather frequently. The computer will often
focus its energies on containing a single checker by bringing the last
checker forward to assist, even though it might superficially seem beneficial
to keep the last checker back "for more contact."
---
Tim Chow

OK, I follow your 'contact' argument.
Why is making the 5 prime vs leaving a direct shot best for containment?
--
BD

[Tim Chow]

Post by BlueDice
OK, I follow your 'contact' argument.
Why is making the 5 prime vs leaving a direct shot best for containment?

Well, this is the tough part to judge, but the main point is that if X plays
something like 13/7 13/11, most of O's sixes play *much* better than if X
plays 23/15. According to the computer, if X plays 23/15, then O isn't even
supposed to jump out with 61 62 63 65. The swing on 65 is especially dramatic.
Even though O can hit with the 5, the computer doesn't hit!

A smaller factor is that if X stays back, then O's rolls that make her 4pt
are better for her since they help block in X's straggler.

---
Tim Chow

[BlueDice]

Post by Tim Chow

Post by BlueDice
OK, I follow your 'contact' argument.
Why is making the 5 prime vs leaving a direct shot best for containment?

Well, this is the tough part to judge, but the main point is that if X plays
something like 13/7 13/11, most of O's sixes play *much* better than if X
plays 23/15. According to the computer, if X plays 23/15, then O isn't even
supposed to jump out with 61 62 63 65. The swing on 65 is especially dramatic.
Even though O can hit with the 5, the computer doesn't hit!
A smaller factor is that if X stays back, then O's rolls that make her 4pt
are better for her since they help block in X's straggler.
---
Tim Chow

Thanks for your insight with your detailed analysis, it really is very helpful.
--
BD

[b***@gmail.com]

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62
--
BD

After playing the obvious part you then pick up your dice.

Stick

[Paul Epstein]

Post by b***@gmail.com

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62
--
BD

After playing the obvious part you then pick up your dice.

Not if it's a clocked game.

[BlueDice]

Post by BlueDice
XGID=--aBBBBA----AB--bdceB--A--:1:-1:1:62:0:0:0:0:10
X:Player 2 O:Player 1
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 144 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62
--
BD

The obvious play (according to XG) by a significant margin is to hop out with the straggler and refuse to make the 5 prime and 'volunteer' a direct shot!
From Stick's answer it was also obvious to him thus partially reconfirming his genius.
As for the explanation, we may need to wait for the next coming of the BG Messiah..


1. Rollout¹ 23/15 eq:+0.164
Player: 62.46% (G:12.64% B:0.24%)
Opponent: 37.54% (G:6.73% B:0.19%)
Confidence: ±0.008 (+0.156..+0.172) - [100.0%]
Duration: 3 minutes 35 seconds

2. Rollout¹ 23/21 13/7 eq:+0.053 (-0.111)
Player: 58.06% (G:13.24% B:0.29%)
Opponent: 41.94% (G:7.10% B:0.24%)
Confidence: ±0.008 (+0.045..+0.062) - [0.0%]
Duration: 3 minutes 10 seconds

3. XG Roller++ 13/7 12/10 eq:+0.013 (-0.151)
Player: 56.36% (G:13.33% B:0.28%)
Opponent: 43.64% (G:7.45% B:0.29%)

4. XG Roller++ 13/11 13/7 eq:-0.005 (-0.169)
Player: 55.78% (G:13.14% B:0.27%)
Opponent: 44.22% (G:7.40% B:0.28%)

¹ 1296 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.10
--
BD

[Tim Chow]

Post by BlueDice
As for the explanation, we may need to wait for the next coming of the BG Messiah..

Well, I think my explanation was on the right track; it's just that
I failed to consider 23/15. X's game plan is to contain O's straggler,
and 23/15 gains more outfield coverage. It also takes away some of
O's counterplay (attacking X's straggler). The cost is a direct shot,
but everything else has downsides as well.

Here's a potentially interesting variant.

XGID=--aBBBBA-----C--bdceB--A--:1:-1:1:62:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
[phone]------19-20-21-22-23-24-+
| X O O O | | O X X | +---+
| X O O O | | O X | | 2 |
| X O O | | O | +---+
| O | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| | | X X X X |
| X | | X X X X O |
[phone]-------6--5--4--3--2--1-+
Pip count X: 145 O: 124 X-O: 0-0
Cube: 2, O own cube
X to play 62

1. XG Roller++ 23/21 13/7 eq:+0.049
Player: 58.00% (G:11.32% B:0.19%)
Opponent: 42.00% (G:5.79% B:0.21%)

2. XG Roller++ 23/15 eq:+0.046 (-0.004)
Player: 57.83% (G:11.44% B:0.20%)
Opponent: 42.17% (G:6.04% B:0.19%)

3. XG Roller++ 13/11 13/7 eq:+0.012 (-0.037)
Player: 56.45% (G:13.16% B:0.25%)
Opponent: 43.55% (G:7.45% B:0.28%)

eXtreme Gammon Version: 2.19.207.pre-release


---
Tim Chow

[Paul Epstein]

On Saturday, June 20, 2020 at 4:45:06 PM UTC+1, Tim Chow wrote:
...

Post by Tim Chow
Well, I think my explanation was on the right track;

...

No, it was way off. Not even close to reality.

[Tim Chow]

Post by Paul Epstein
...

Post by Tim Chow
Well, I think my explanation was on the right track;

...
No, it was way off. Not even close to reality.

Why do you say that? The main point is to try to contain the straggler.
So we don't want to advance our outfield checkers and we want to bring
the straggler into a position where it can help with outfield control.

Certainly, I did get tricked by the subject line into thinking that making
the bar point was necessary, so that part was off. But even in the original
position, my play was XG's 2nd choice, and in the variant, the two plays
(running versus making the bar pt and stepping up) are close.

---
Tim Chow

[Paul Epstein]

Post by Tim Chow

Post by Paul Epstein
...

Post by Tim Chow
Well, I think my explanation was on the right track;

...
No, it was way off. Not even close to reality.

Why do you say that?

Just felt like winding you up. Kind of intended as a teasing comment.

Paul