Hit a blot, any blot

[Tim Chow]

XGID=aaBBABC-A---d----dadCA----:1:-1:1:63:0:0:0:0:10
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 X X | +---+
| O | | O X | | 2 |
| O | | O X | +---+
| O | | O |
| | | |
| |BAR| |
| | O | |
| O | | |
| O | | X |
| O | | X X X X |
| O X | | X X X X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pipcount X: 131 O: 164 X-O: 0-0
Cube : 2, O own cube
X to play 63

---
Tim Chow

[Bradley K. Sherman]

Tim Chow <tchow...@yahoo.com> wrote:
>Score is X:0 O:0. Money session
> +13-14-15-16-17-18------19-20-21-22-23-24-+
> | O O | | O X X | +---+
> | O | | O X | | 2 |
> | O | | O X | +---+
> | O | | O |
> | | | |
> | |BAR| |
> | | O | |
> | O | | |
> | O | | X |
> | O | | X X X X |
> | O X | | X X X X X O |
> +12-11-10--9--8--7-------6--5--4--3--2--1-+

> X:131 O:164, X to play 63

Well I sure hope hitting on the ace point is wrong
because it looks all wrong.
21/18*,20/14

--bks

[badgolferman]

4/1*, 8/2

[crf]

Hit on the 1 or the 18?

Hit on the 1, O anchors only with 11 and 44. Hit on the 18, O anchors with any 1 and 44.

8/2 4/1*

Wait... do we really need to move to the 2 to cover the 1 so urgently? getting hit isn't so bad (wide open board), and we probably prefer making the 4 instead. Bring another stone around?

4/1* 20/14 -- more numbers to hit on the 18.

[peps...@gmail.com]

I adore 4/1* 8/2 though it looks most unnatural and I would never have considered this over the board, if I hadn't seen badgolferman's post.
An ace game is viable so let's prevent this. Even though badgolferman might not be the best golf instructor, he is definitely providing a master class in diversification here, with 2's 3's 1's 5's all playing like a winning lottery ticket.
Over the board, I would have made Bradley's play for sure, but I don't like it after further consideration.

Paul Epstein

[badgolferman]

I only chose that play because it seemed the most unnatural and
dangerous. Normally I would play farther back in the outfield. I am
biting on Tim's lure in this case to be taught a lesson.

[Tim Chow]

On May 6, 6:11 am, "badgolferman" <REMOVETHISbadgolfer...@gmail.com>
wrote:

> I only chose that play because it seemed the most unnatural and
> dangerous.  Normally I would play farther back in the outfield.

What exactly would you normally play? Would you still hit 4/1* and
move a back checker, or would you play 21/18*, or what?

> I am biting on Tim's lure in this case to be taught a lesson.

I don't quite follow this. Wouldn't stating your normal play be the
obvious way to be taught a lesson? I haven't been keeping track, but
my impression has been that when you state two different plays and
identify one as your own play and the other as what you think the
computer will play, your own play has been just as likely to be the
actual computer play as your guess.

---
Tim Chow

[badgolferman]

Tim Chow wrote:

>On May 6, 6:11�am, "badgolferman" <REMOVETHISbadgolfer...@gmail.com>
>wrote:
>> I only chose that play because it seemed the most unnatural and
>> dangerous. �Normally I would play farther back in the outfield.
>
>What exactly would you normally play? Would you still hit 4/1* and
>move a back checker, or would you play 21/18*, or what?

My normal play would be something like 21/18*, 20/14.

>> I am biting on Tim's lure in this case to be taught a lesson.
>
>I don't quite follow this. Wouldn't stating your normal play be the
>obvious way to be taught a lesson? I haven't been keeping track, but
>my impression has been that when you state two different plays and
>identify one as your own play and the other as what you think the
>computer will play, your own play has been just as likely to be the
>actual computer play as your guess.
>

In this case I chose something I wouldn't normally play because I am
more conservative in my inner board than most of you guys are and I
know you tend to present the type of plays that challenge my tendencies.

[Walt]

X has two builders for the 4 point for 22 covers vs O's 11 hits. Being
the 2-1 favorite makes hitting on the 18 attractive.

OTOH, hitting on the ace means O will have no anchor slotted and is thus
only 1/18 to anchor instead of 11/36.

Good problem. Tough choice.

I'm going to say that hitting on the ace means one less builder to cover
the homeboard blot, unless we play 8/2 which seems anti-positional. So
I'll go with hitting in the outfield 21/18* 20/14.

--
//Walt

[peps...@gmail.com]

Counting builders is too crude. After 4/1* 20/14 (for example) X has 19/36 cover shots for the acepoint. You might want to count the number of covering shots rather than just the number of builders. I'm confident that the golfer and I are right to hit on the ace to stop the anchoring. But I'm not at all sure that we're right to go with the anti-positional 8/2. (Not changing my mind, though).

[Tim Chow]

X is obviously blitzing here, but he doesn't have much ammunition
left. It's therefore tempting to switch to a more "positional" plan,
getting the back checkers moving, leaving the 4pt slotted in the hope
of making it, and resigning oneself to letting O anchor. What the
bots tell us, though, is that keeping the blitz going as long as
possible is usually the best game plan when we have such a huge board
advantage. And for blitzing, preventing O from anchoring with 4/1*
takes precedence over hitting a third blot with 21/18*. We might hit
the third blot anyway but "an anchor is forever."

I thought that the choice between 4/1* and 21/18* was the main
difficulty in this position, but after reading the responses I see
that some people find 8/2 to be an attractive 6. But 8/2 leaves that
checker with only one role left in life, to cover the 1pt if you roll
an ace. On the 8pt, it serves multiple functions: It prepares to hit
O if O enters on the 4pt; it provides an indirect cover for the 1pt;
and it is better placed positionally if the game takes a positional
turn. It's much more constructive to bring another checker closer to
the scene of action with 20/14.

=======
Rollout
=======

1. Rollout: 20/14 4/1* eq:+1.079
Player : 79.47% (G:59.02% B:2.40%)
Opponent: 20.53% (G:3.61% B:0.17%)
Confidence: ±0.012 (+1.067<E<+1.091)

2. Rollout: 21/15 4/1* eq:+1.003 (-0.076)
Player : 77.71% (G:56.18% B:2.22%)
Opponent: 22.29% (G:3.86% B:0.16%)
Confidence: ±0.012 (+0.991<E<+1.015)

3. Rollout: 21/18* 20/14 eq:+0.953 (-0.126)
Player : 77.06% (G:49.86% B:4.22%)
Opponent: 22.94% (G:3.07% B:0.12%)
Confidence: ±0.014 (+0.939<E<+0.967)

4. Rollout: 8/2 4/1* eq:+0.879 (-0.200)
Player : 74.69% (G:52.94% B:1.57%)
Opponent: 25.31% (G:4.89% B:0.23%)
Confidence: ±0.014 (+0.865<E<+0.893)

1296 Games rolled with Variance Reduction.
Dice Seed: 2
Moves and cube decisions: 3 ply

eXtreme Gammon Version: 1.21

---
Tim Chow

[peps...@gmail.com]

I think that the equity, when played correctly, is higher for the favourite than most non-experts would expect. As a thought experiment, suppose X offered to stop the game and take 2 points. I doubt that O would accept, even though XG thinks O would have a real bargain.

Paul Epstein