XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63

---
Tim Chow

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

On Thursday, November 11, 2021 at 2:02:52 AM UTC, Tim Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63

I have four candidates:
1) Hit O
2) 8/5 8/2
3) 19/13 6/3
4) 19/13 5/2.

19/13 6/3 and 19/13 5/2 are very similar.
If the equities are different for these, it's likely to be because of the anti-jokers they leave us.
However, I can't see a difference so I'll just use my positional intuition.
I don't want to leave my six point stripped so I'll eliminate option 3).
(It's very possible that the equities for these are different by no more than 0.001 or something.)

If we generalise the DMP angle, and think of match scores which make the play clear, then
I think the simplest observation is that the gammon-go play is 1) Hit O. The main downside to this
play is that the return hits lose us a lot of gammons.

To me, the DMP play is completely unclear: Playing bold can decide the game immediately and
playing safe hopes that we can continue to avoid being hit and cash in our racing lead.
Any candidate might be the DMP play, based on my knowledge of the game.
So maybe we can generalise Stick's rule and say "If the play is unclear between several candidates
but one play stands out at a particular score, choose that one play."

My intuition is that bold is unnecessarily risky (both at DMP and for money).
I think cashing in the racing lead while never leaving a shot is also a bit much to ask for.
So I do want to leave the outfield covered and keep my 19 point checker fixed. Note also that the outfield blot is safer than it may seem because the hitting indirect 64 leaves us 55% favourites to hit back.
8/5 8/2 for me.
It's important to note here that our opponent can't play 5s or 6s from their inner board so we are likely to have good hitting opportunities
at our next shake. Keep the outfield covered.
8/5 8/2 (I know I'm repeating myself even though I didn't roll doubles).

Paul

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

On 11/10/2021 9:02 PM, Timothy Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1   O:Player 2
Score is X:0 O:0. Unlimited Game  +13-14-15-16-17-18------19-20-21-22-23-24-+
 | X                |   | X  O  O  O  O  O | +---+
 | X                |   |    O  O  O     O | | 2 |  |                  |   |    O     O       | +---+
 |                  |   |                  |
 |                  |   |                  |
 |                  |BAR|                  |
 |                  |   |                  |
 |                  |   |                  |
 |                  |   | X  X             |  |             X    |   | X  X     X  X  O |  |          O  X    |   | X  X  O  X  X  O |  +12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count  X: 104  O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63

I think the rule is "When in doubt, make the DMP play" which works fine
when you are good enough that most of the time you're only in doubt when
the difference is close. If you're in doubt a lot of the time (i.e.
like most of us) it might not be so helpful.

Anyway, hitting loose on the four point looks to be the DMP play (while
we're throwing adages around we can include "At DMP, hit first and ask questions later") Here, X is substantially ahead in the race so quietly clearing the eight looks best to me.

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

On Thursday, November 11, 2021 at 6:29:23 PM UTC, ah...Clem wrote:

On 11/10/2021 9:02 PM, Timothy Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63

I think the rule is "When in doubt, make the DMP play" which works fine
when you are good enough that most of the time you're only in doubt when
the difference is close. If you're in doubt a lot of the time (i.e.
like most of us) it might not be so helpful.

Anyway, hitting loose on the four point looks to be the DMP play (while
we're throwing adages around we can include "At DMP, hit first and ask questions later") Here, X is substantially ahead in the race so quietly clearing the eight looks best to me.

I strongly disagree with that play! I think it's much better to noisily clear the eight.
Clear the eight while singing the song "Tubthumping" (It's the one that goes "I get knocked down!..." )

Paul

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

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63

This is a position where one has to weigh wins against gammons.

First, which play wins the most? A lot of players overestimate the
opponent's gammons and underestimate their own, and might balk at
13/4* because they're afraid they would get gammoned if they get
hit back. Such players might be attracted to 13/4* at DMP since
they know they don't have to worry about getting gammoned.

However, when you have a solid race lead and have escaped your
checkers, and are just trying to come home safely against a deep
anchor game, often the DMP strategy is to play safe and aim to
win the race without getting hit. The way to parse this position
is that at DMP you should just play safe, but for money, you have
to consider 13/4* because it greatly increases your gammon *wins*.
You will lose some gammons after 13/4*, but not a lot, since O's
board isn't that strong and she still has a lot of work to do
even if she hits back.

So now you "just" have to assess whether 13/4* wins enough extra
gammons to compensate for the losses. O has enough blots lying
around that, according to the rollout, the decision isn't close.
If you're Stick, all this is obvious, so this position is not an
exception to his DMP rule.

1. Rollout¹ 13/4* eq:+0.803
Player: 76.70% (G:37.77% B:4.79%)
Opponent: 23.30% (G:6.67% B:0.21%)
Confidence: ±0.007 (+0.795..+0.810) - [100.0%]

2. Rollout¹ 19/13 6/3 eq:+0.718 (-0.085)
Player: 80.25% (G:18.88% B:1.15%)
Opponent: 19.75% (G:1.88% B:0.04%)
Confidence: ±0.006 (+0.712..+0.725) - [0.0%]

3. Rollout¹ 19/13 5/2 eq:+0.703 (-0.100)
Player: 79.88% (G:18.31% B:1.22%)
Opponent: 20.12% (G:1.82% B:0.03%)
Confidence: ±0.006 (+0.696..+0.709) - [0.0%]

4. Rollout¹ 8/5 8/2 eq:+0.658 (-0.144)
Player: 76.51% (G:23.15% B:1.76%)
Opponent: 23.49% (G:2.89% B:0.07%)
Confidence: ±0.007 (+0.651..+0.666) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

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

On Saturday, November 13, 2021 at 5:12:54 AM UTC, Tim Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63
This is a position where one has to weigh wins against gammons.

First, which play wins the most? A lot of players overestimate the opponent's gammons and underestimate their own, and might balk at
13/4* because they're afraid they would get gammoned if they get
hit back. Such players might be attracted to 13/4* at DMP since
they know they don't have to worry about getting gammoned.

However, when you have a solid race lead and have escaped your
checkers, and are just trying to come home safely against a deep
anchor game, often the DMP strategy is to play safe and aim to
win the race without getting hit. The way to parse this position
is that at DMP you should just play safe, but for money, you have
to consider 13/4* because it greatly increases your gammon *wins*.
You will lose some gammons after 13/4*, but not a lot, since O's
board isn't that strong and she still has a lot of work to do
even if she hits back.

So now you "just" have to assess whether 13/4* wins enough extra
gammons to compensate for the losses. O has enough blots lying
around that, according to the rollout, the decision isn't close.
If you're Stick, all this is obvious, so this position is not an
exception to his DMP rule.

1. Rollout¹ 13/4* eq:+0.803
Player: 76.70% (G:37.77% B:4.79%)
Opponent: 23.30% (G:6.67% B:0.21%)
Confidence: ±0.007 (+0.795..+0.810) - [100.0%]

2. Rollout¹ 19/13 6/3 eq:+0.718 (-0.085)
Player: 80.25% (G:18.88% B:1.15%)
Opponent: 19.75% (G:1.88% B:0.04%)
Confidence: ±0.006 (+0.712..+0.725) - [0.0%]

3. Rollout¹ 19/13 5/2 eq:+0.703 (-0.100)
Player: 79.88% (G:18.31% B:1.22%)
Opponent: 20.12% (G:1.82% B:0.03%)
Confidence: ±0.006 (+0.696..+0.709) - [0.0%]

4. Rollout¹ 8/5 8/2 eq:+0.658 (-0.144)
Player: 76.51% (G:23.15% B:1.76%)
Opponent: 23.49% (G:2.89% B:0.07%)
Confidence: ±0.007 (+0.651..+0.666) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

Thanks.
Do you know approx how much MWC 13/4* loses at DMP?

Paul

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

On Saturday, November 13, 2021 at 5:17:44 AM UTC-5, peps...@gmail.com wrote:

On Saturday, November 13, 2021 at 5:12:54 AM UTC, Tim Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game +13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63
This is a position where one has to weigh wins against gammons.

First, which play wins the most? A lot of players overestimate the opponent's gammons and underestimate their own, and might balk at
13/4* because they're afraid they would get gammoned if they get
hit back. Such players might be attracted to 13/4* at DMP since
they know they don't have to worry about getting gammoned.

However, when you have a solid race lead and have escaped your
checkers, and are just trying to come home safely against a deep
anchor game, often the DMP strategy is to play safe and aim to
win the race without getting hit. The way to parse this position
is that at DMP you should just play safe, but for money, you have
to consider 13/4* because it greatly increases your gammon *wins*.
You will lose some gammons after 13/4*, but not a lot, since O's
board isn't that strong and she still has a lot of work to do
even if she hits back.

So now you "just" have to assess whether 13/4* wins enough extra
gammons to compensate for the losses. O has enough blots lying
around that, according to the rollout, the decision isn't close.
If you're Stick, all this is obvious, so this position is not an
exception to his DMP rule.

1. Rollout¹ 13/4* eq:+0.803
Player: 76.70% (G:37.77% B:4.79%)
Opponent: 23.30% (G:6.67% B:0.21%)
Confidence: ±0.007 (+0.795..+0.810) - [100.0%]

2. Rollout¹ 19/13 6/3 eq:+0.718 (-0.085)
Player: 80.25% (G:18.88% B:1.15%)
Opponent: 19.75% (G:1.88% B:0.04%)
Confidence: ±0.006 (+0.712..+0.725) - [0.0%]

3. Rollout¹ 19/13 5/2 eq:+0.703 (-0.100)
Player: 79.88% (G:18.31% B:1.22%)
Opponent: 20.12% (G:1.82% B:0.03%)
Confidence: ±0.006 (+0.696..+0.709) - [0.0%]

4. Rollout¹ 8/5 8/2 eq:+0.658 (-0.144)
Player: 76.51% (G:23.15% B:1.76%)
Opponent: 23.49% (G:2.89% B:0.07%)
Confidence: ±0.007 (+0.651..+0.666) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

Thanks.
Do you know approx how much MWC 13/4* loses at DMP?

Paul

3.5%.

Stick

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

On Saturday, November 13, 2021 at 6:46:32 PM UTC, J R wrote:

On Saturday, November 13, 2021 at 5:17:44 AM UTC-5, peps...@gmail.com wrote:

On Saturday, November 13, 2021 at 5:12:54 AM UTC, Tim Chow wrote:

XGID=-bBBaCC-Ba---B-----Acbcab-:1:-1:1:63:0:0:0:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game +13-14-15-16-17-18------19-20-21-22-23-24-+
| X | | X O O O O O | +---+
| X | | O O O O | | 2 |
| | | O O | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X X |
| X | | X X X X O |
| O X | | X X O X X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 104 O: 121 X-O: 0-0
Cube: 2, O own cube
X to play 63
This is a position where one has to weigh wins against gammons.

First, which play wins the most? A lot of players overestimate the opponent's gammons and underestimate their own, and might balk at
13/4* because they're afraid they would get gammoned if they get
hit back. Such players might be attracted to 13/4* at DMP since
they know they don't have to worry about getting gammoned.

However, when you have a solid race lead and have escaped your
checkers, and are just trying to come home safely against a deep
anchor game, often the DMP strategy is to play safe and aim to
win the race without getting hit. The way to parse this position
is that at DMP you should just play safe, but for money, you have
to consider 13/4* because it greatly increases your gammon *wins*.
You will lose some gammons after 13/4*, but not a lot, since O's
board isn't that strong and she still has a lot of work to do
even if she hits back.

So now you "just" have to assess whether 13/4* wins enough extra
gammons to compensate for the losses. O has enough blots lying
around that, according to the rollout, the decision isn't close.
If you're Stick, all this is obvious, so this position is not an exception to his DMP rule.

1. Rollout¹ 13/4* eq:+0.803
Player: 76.70% (G:37.77% B:4.79%)
Opponent: 23.30% (G:6.67% B:0.21%)
Confidence: ±0.007 (+0.795..+0.810) - [100.0%]

2. Rollout¹ 19/13 6/3 eq:+0.718 (-0.085)
Player: 80.25% (G:18.88% B:1.15%)
Opponent: 19.75% (G:1.88% B:0.04%)
Confidence: ±0.006 (+0.712..+0.725) - [0.0%]

3. Rollout¹ 19/13 5/2 eq:+0.703 (-0.100)
Player: 79.88% (G:18.31% B:1.22%)
Opponent: 20.12% (G:1.82% B:0.03%)
Confidence: ±0.006 (+0.696..+0.709) - [0.0%]

4. Rollout¹ 8/5 8/2 eq:+0.658 (-0.144)
Player: 76.51% (G:23.15% B:1.76%)
Opponent: 23.49% (G:2.89% B:0.07%)
Confidence: ±0.007 (+0.651..+0.666) - [0.0%]

¹ 1296 Games rolled with Variance Reduction.
Dice Seed: 271828
Moves: 3-ply, cube decisions: XG Roller

eXtreme Gammon Version: 2.19.207.pre-release

Thanks.
Do you know approx how much MWC 13/4* loses at DMP?

Paul

3.5%.

Thanks. That's huge.

Paul

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