Hint: I don't think Jacoby Beaver matters.

XGID=-a--BaC-AA--cD-a-c-dA-b-BA:0:0:1:52:0:0:3:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X O O | | O X O X |
| X O | | O O X |
| X O | | O |
| X | | O |
| | X | |
| |BAR| |
| | | |
| | | |
| O | | X |
| O | | X X |
| O X X | | X O X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 188 O: 147 X-O: 0-0
Cube: 1
X to play 52

--
Ah....Clem
The future is fun, the future is fair.

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

On 8/13/2022 12:57 PM, ah....Clem wrote:

Hint: I don't think Jacoby Beaver matters.

XGID=-a--BaC-AA--cD-a-c-dA-b-BA:0:0:1:52:0:0:3:0:10

X:Player 1   O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
 +13-14-15-16-17-18------19-20-21-22-23-24-+
 | X     O     O    |   | O  X     O     X |
 | X           O    |   | O        O     X |
 | X           O    |   | O                |
 | X                |   | O                |
 |                  | X |                  |
 |                  |BAR|                  |
 |                  |   |                  |
 |                  |   |                  |
 | O                |   | X                |
 | O                |   | X     X          |
 | O        X  X    |   | X  O  X        O |
 +12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count  X: 188  O: 147 X-O: 0-0
Cube: 1
X to play 52

The question isn't whether X wants to play a backgame but whether
X feels the need to make a second anchor. But after bar/20 there
is no good 2 to play. I would hit, bar/23 20/15*.

---
Tim Chow

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

As Tim observes, it's not a backgame yet even when X makes the second
anchor. My thought was

"We enter with the five and create a backgame structure, but without the
timing of a backgame. But even in a proto-backgame pure play and
mobility are usually the right idea. So, bring one down - either we get
hit and gain timing or we solidify our blockade. 9/7 also looks
playable, but 13/11 looks stronger."

XG rollout agrees with entering with the five and also that 13/10 is
stronger than 9/7. But it likes the passive play of 8/6 that minimizes
shots. Presumably, this is because we're not committed to a backgame yet.


Interestingly, XGR++ reverses the order of second and third liking the
hit better than bar/20 13/10.


XGID=-a--BaC-AA--cD-a-c-dA-b-BA:0:0:1:52:0:0:3:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X O O | | O X O X |
| X O | | O O X |
| X O | | O |
| X | | O |
| | X | |
| |BAR| |
| | | |
| | | |
| O | | X |
| O | | X X |
| O X X | | X O X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 188 O: 147 X-O: 0-0
Cube: 1
X to play 52

1. Rollout¹ Bar/20 8/6 eq:-0.267
Player: 45.22% (G:10.91% B:0.44%)
Opponent: 54.78% (G:19.28% B:1.50%)
Confidence: ±0.017 (-0.284..-0.250) - [81.4%]
Duration: 1 hour 21 minutes

2. Rollout¹ Bar/20 13/11 eq:-0.280 (-0.013)
Player: 46.24% (G:10.38% B:0.47%)
Opponent: 53.76% (G:22.74% B:2.11%)
Confidence: ±0.021 (-0.300..-0.259) - [16.8%]
Duration: 59 minutes 49 seconds

3. Rollout¹ Bar/23 20/15* eq:-0.292 (-0.025)
Player: 45.57% (G:11.74% B:0.58%)
Opponent: 54.43% (G:22.97% B:2.48%)
Confidence: ±0.019 (-0.311..-0.273) - [1.9%]
Duration: 1 hour 44 minutes

4. Rollout¹ Bar/20 9/7 eq:-0.315 (-0.048)
Player: 45.08% (G:9.54% B:0.40%)
Opponent: 54.92% (G:22.32% B:1.83%)
Confidence: ±0.020 (-0.335..-0.295) - [0.0%]
Duration: 1 hour 23 minutes




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


eXtreme Gammon Version: 2.10

--
Ah....Clem
The future is fun, the future is fair.

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

On 8/16/2022 9:25 AM, ah....Clem wrote:

Interestingly, XGR++ reverses the order of second and third liking the
hit better than bar/20 13/10.

There's a lot of variance left.

      Confidence: ±0.017 (-0.284..-0.250) - [81.4%]
      Confidence: ±0.021 (-0.300..-0.259) - [16.8%]
      Confidence: ±0.019 (-0.311..-0.273) - [1.9%]
      Confidence: ±0.020 (-0.335..-0.295) - [0.0%]

---
Tim Chow

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

On 8/16/2022 9:14 PM, Timothy Chow wrote:

On 8/16/2022 9:25 AM, ah....Clem wrote:

Interestingly, XGR++ reverses the order of second and third liking the
hit better than bar/20 13/10.

There's a lot of variance left.

       Confidence: ±0.017 (-0.284..-0.250) - [81.4%]
       Confidence: ±0.021 (-0.300..-0.259) - [16.8%]
       Confidence: ±0.019 (-0.311..-0.273) - [1.9%]
       Confidence: ±0.020 (-0.335..-0.295) - [0.0%]

Fair enough. My approach is that once I see a rollout that says the
plays are close I'm satisfied with that. My play is weak enough that
I'm not going to worry about differences of .02 or less. Unless it's
some regularly recurring position like openings or bearoffs.

Anyway, here's an extended rollout. Making the second anchor is the
right idea, and it's within the margin of error whether to bring one
down or safety the blot on the eight.

XGID=-a--BaC-AA--cD-a-c-dA-b-BA:0:0:1:52:0:0:3:0:10

X:Player 1 O:Player 2
Score is X:0 O:0. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| X O O | | O X O X |
| X O | | O O X |
| X O | | O |
| X | | O |
| | X | |
| |BAR| |
| | | |
| | | |
| O | | X |
| O | | X X |
| O X X | | X O X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 188 O: 147 X-O: 0-0
Cube: 1
X to play 52

1. Rollout¹ Bar/20 8/6 eq:-0.269
Player: 45.05% (G:11.04% B:0.43%)
Opponent: 54.95% (G:19.31% B:1.49%)
Confidence: ±0.013 (-0.281..-0.256) - [94.8%]
Duration: 1 hour 31 minutes

2. Rollout¹ Bar/20 13/11 eq:-0.285 (-0.016)
Player: 46.14% (G:10.39% B:0.45%)
Opponent: 53.86% (G:22.75% B:2.09%)
Confidence: ±0.015 (-0.299..-0.270) - [5.2%]
Duration: 1 hour 10 minutes

3. Rollout¹ Bar/23 20/15* eq:-0.303 (-0.035)
Player: 45.50% (G:11.63% B:0.60%)
Opponent: 54.50% (G:22.87% B:2.50%)
Confidence: ±0.015 (-0.318..-0.288) - [0.0%]
Duration: 1 hour 54 minutes

4. Rollout¹ Bar/20 9/7 eq:-0.313 (-0.044)
Player: 45.10% (G:9.61% B:0.41%)
Opponent: 54.90% (G:22.32% B:1.76%)
Confidence: ±0.014 (-0.326..-0.299) - [0.0%]
Duration: 1 hour 33 minutes

¹ 2592 Games rolled with Variance Reduction.
Moves: 3-ply, cube decisions: XG Roller


eXtreme Gammon Version: 2.10


--
Ah....Clem
The future is fun, the future is fair.

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

XGID=--BBaBC-A--bb----bccBC--b-:0:0:1:51:0:0:0:5:10

X:You O:O
Score is X:0 O:0 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O X X O |
| O O | | O X X O |
| O | | O X |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O O | | X X X X |
| O O X | | X X O X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 149 O: 132 X-O: 0-0/5
Cube: 1
X to play 51



--
Ah....Clem
The future is fun, the future is fair.

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

On 11/24/2022 3:56 PM, ah....Clem wrote:

XGID=--BBaBC-A--bb----bccBC--b-:0:0:1:51:0:0:0:5:10

X:You   O:O
Score is X:0 O:0 5 pt.(s) match.
 +13-14-15-16-17-18------19-20-21-22-23-24-+
 |             O  O |   | O  X  X        O |
 |             O  O |   | O  X  X        O |
 |                O |   | O     X          |
 |                  |   |                  |
 |                  |   |                  |
 |                  |BAR|                  |
 |                  |   |                  |
 |                  |   |                  |
 |                  |   | X                |
 | O  O             |   | X  X     X  X    |
 | O  O        X    |   | X  X  O  X  X    |
 +12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count  X: 149  O: 132 X-O: 0-0/5
Cube: 1
X to play 51

21/16 looks obvious, and then we can ask whether the banana split
5/4* makes any sense. I don't see what it buys us; O has no other
blots to pick up, and it doesn't seem that she has any threats
serious enough to warrant a tempo hit. I'd just continue on to
the 15pt.

---
Tim Chow

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

Like Tim, I didn't see the banana split play being applicable here.
There's not enough timing for a viable backgame, and I generally avoid
getting into them as the backgamer given the choice.

But XG likes the banana split. Not sure why. I thought it might have
something to do with O's made ace point, but (see variant below) that
doesn't seem to make much difference.

I played 8/7 (21/16 is obvious) and that's does much worse in the
rollout than the other non-hitting options. Not sure why that is either.

XGID=--BBaBC-A--bb----bccBC--b-:0:0:1:51:0:0:0:5:10

X:Player 1 O:Player 2
Score is X:0 O:0 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O X X O |
| O O | | O X X O |
| O | | O X |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O O | | X X X X |
| O O X | | X X O X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 149 O: 132 X-O: 0-0/5
Cube: 1
X to play 51

1. Rollout¹ 21/16 5/4* eq:-0.356
Player: 41.65% (G:13.30% B:0.41%)
Opponent: 58.35% (G:18.48% B:0.52%)
Confidence: ±0.017 (-0.373..-0.339) - [100.0%]
Duration: 6 minutes 19 seconds

2. Rollout¹ 21/15 eq:-0.396 (-0.040)
Player: 38.92% (G:10.98% B:0.36%)
Opponent: 61.08% (G:12.03% B:0.25%)
Confidence: ±0.014 (-0.410..-0.382) - [0.0%]
Duration: 6 minutes 22 seconds

3. Rollout¹ 21/16 6/5 eq:-0.412 (-0.056)
Player: 38.69% (G:10.46% B:0.35%)
Opponent: 61.31% (G:11.84% B:0.25%)
Confidence: ±0.014 (-0.426..-0.398) - [0.0%]
Duration: 6 minutes 15 seconds

4. Rollout¹ 21/16 8/7 eq:-0.450 (-0.093)
Player: 38.12% (G:10.66% B:0.38%)
Opponent: 61.88% (G:13.24% B:0.34%)
Confidence: ±0.014 (-0.463..-0.436) - [0.0%]
Duration: 5 minutes 51 seconds

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


eXtreme Gammon Version: 2.10, MET: Kazaross XG2



///////////////////////
// variant - move the ace point checkers to the three point //////////////////////

XGID=--BBaBC-A--bb----bccBCb---:0:0:1:51:0:0:0:5:10

X:Player 1 O:Player 2
Score is X:0 O:0 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O X X O |
| O O | | O X X O |
| O | | O X |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O O | | X X X X |
| O O X | | X X O X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 149 O: 136 X-O: 0-0/5
Cube: 1
X to play 51

1. XG Roller++ 21/16 5/4* eq:-0.430
Player: 38.82% (G:11.43% B:0.24%)
Opponent: 61.18% (G:17.22% B:0.53%)

2. XG Roller++ 21/15 eq:-0.453 (-0.023)
Player: 36.55% (G:9.30% B:0.26%)
Opponent: 63.45% (G:11.14% B:0.30%)

3. XG Roller++ 21/16 6/5 eq:-0.480 (-0.050)
Player: 35.88% (G:8.56% B:0.19%)
Opponent: 64.12% (G:10.97% B:0.27%)

4. XG Roller++ 21/16 8/7 eq:-0.494 (-0.064)
Player: 35.94% (G:9.38% B:0.25%)
Opponent: 64.06% (G:12.03% B:0.32%)



eXtreme Gammon Version: 2.10, MET: Kazaross XG2

--
Ah....Clem
The future is fun, the future is fair.

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

On 11/26/2022 11:49 AM, ah....Clem wrote:

But XG likes the banana split. Not sure why. I thought it might have something to do with O's made ace point, but (see variant below) that
doesn't seem to make much difference.

So I think that O's "threat" is simply to escape her straggler. The
banana split can be thought of as an attempt to contain the straggler
that gains a lot when it works and isn't a disaster when O hits back.
Note that if you shift X's 2pt checkers to his 1pt then XG doesn't
like the banana split any more, presumably because it becomes much
harder to contain the straggler.

Here's a potentially interesting variant.

XGID=--BBBaBA-A-bb----bccBC--b-:0:0:1:51:0:0:0:5:10

Score is X:0 O:0 5 pt.(s) match.
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O O | | O X X O |
| O O | | O X X O |
| O | | O X |
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | |
| O O | | X X X X |
| O O X X | | X O X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 149 O: 131 X-O: 0-0/5
Cube: 1
X to play 51

1. Rollout¹ 21/16 6/5* eq:-0.451
Player: 40.18% (G:12.30% B:0.38%)
Opponent: 59.82% (G:22.85% B:0.57%)
Confidence: ±0.017 (-0.468..-0.434) - [100.0%]

2. Rollout¹ 21/16 7/6 eq:-0.487 (-0.036)
Player: 36.70% (G:9.07% B:0.26%)
Opponent: 63.30% (G:11.94% B:0.26%)
Confidence: ±0.014 (-0.501..-0.473) - [0.0%]

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

eXtreme Gammon Version: 2.19.211.pre-release, MET: Kazaross XG2

I played 8/7 (21/16 is obvious) and that's does much worse in the
rollout than the other non-hitting options. Not sure why that is either.

I think duplication of O's hitting rolls plays a role here.

---
Tim Chow

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