Banana split?

[Timothy Chow]

Paul recently remarked that the main challenge with banana splits
is seeing that they are on the menu. Once you see them, he says,
they are either obviously good or obviously bad.

Whether we should go bananas in the position below may be obvious
to Paul, but it wasn't obvious to me.

XGID=-BaBBBC--B--b-b--dBcab----:1:-1:1:21:0:0:0:0:10

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-+
| O O X | | O O O | +---+
| O O X | | O O | | 2 |
| O | | O | +---+
| O | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X | | X X X X X |
| O X | | X X X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 98 O: 134 X-O: 0-0
Cube: 2, O own cube
X to play 21

---
Tim Chow

[peps...@gmail.com]

My play is 9/6 although, with so many candidates, I think any choice I make is more likely than not to be suboptimal.
Correct banana splits are somewhat unusual so we should search for toxic-banana warnings and abstain if we see too many
of them. The warnings are as follows: 1) We have a huge pip lead and we don't want to jeopardise this. 2) A banana split is a super-bold play -- these are less effective without cube access since the opponent has residual chances even when the risk pays off. 3) Depending on how we play the banana split, the opponent may have many rolls that both hit and make the 5 point.
So I'm going to say that banana splitting is obviously bad here, but many things in bg that seem "obvious" to me are in fact false -- I'm not Stick.

Paul

[badgolferman]

I don't know what a banana split is but I'm not one for sweets and this
one doesn't look very appetizing either. Let's just go 6/3 and see
what happens.

[peps...@gmail.com]

Just cut two bananas in half and put some whipped cream with no
sugar added in the middle -- don't add ice cream.

However, in my experience, this doesn't work in Texas where dairy cream
is almost unobtainable. If you ask for "fresh cream", you get a disgusting
sugary mix.

Paul

[Timothy Chow]

On 3/7/2021 7:22 AM, badgolferman wrote:
> I don't know what a banana split is

Did you see my response to your recent post where you asked
for a list of backgammon terms?

---
Tim Chow

[peps...@gmail.com]

I interpreted him as saying that he's not familiar with the non-backgammon
culinary definition of "banana split". Since he says "this one doesn't look very
appetizing", he clearly understands it in the bg sense in regard to your position.

Paul

[Timothy Chow]

XGID=-BaBBBC--B--b-b--dBcab----:1:-1:1:21:0:0:0:0:10

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-+
| O O X | | O O O | +---+
| O O X | | O O | | 2 |
| O | | O | +---+
| O | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | X |
| O X | | X X X X X |
| O X | | X X X X O X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 98 O: 134 X-O: 0-0
Cube: 2, O own cube
X to play 21

Banana splits most commonly arise when you are in trouble and
desperately need a tempo hit. This position is an exception
because X is clearly better. The banana hit is justified
because X has a huge board advantage and can try for a blitz.
If it works, he has excellent gammon chances, and even if it
only sort of works, he gains time to bring his back checkers
around, which are currently somewhat awkwardly disconnected from
the rest of his army.

1. Rollout¹ 4/2* 3/2 eq:+0.425
Player: 63.90% (G:37.84% B:0.17%)
Opponent: 36.10% (G:9.22% B:0.46%)
Confidence: ±0.008 (+0.418..+0.433) - [100.0%]

2. Rollout¹ 6/4 3/2* eq:+0.389 (-0.037)
Player: 63.08% (G:36.12% B:0.32%)
Opponent: 36.92% (G:9.21% B:0.47%)
Confidence: ±0.009 (+0.380..+0.398) - [0.0%]

3. Rollout¹ 9/8 9/7 eq:+0.356 (-0.069)
Player: 63.19% (G:30.27% B:0.20%)
Opponent: 36.81% (G:6.52% B:0.30%)
Confidence: ±0.008 (+0.348..+0.364) - [0.0%]

4. Rollout¹ 9/8 6/4 eq:+0.349 (-0.076)
Player: 63.34% (G:28.57% B:0.21%)
Opponent: 36.66% (G:6.16% B:0.29%)
Confidence: ±0.008 (+0.341..+0.357) - [0.0%]

5. Rollout¹ 9/6 eq:+0.334 (-0.091)
Player: 63.50% (G:24.66% B:0.22%)
Opponent: 36.50% (G:4.77% B:0.20%)
Confidence: ±0.008 (+0.326..+0.342) - [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