Hello,

if newbies come to our chouette club, they often struggle with the
concept of wastage, because they do not understand that they cannot
control the pipcount, only the distribution.

While looking at EPCs in positions close to a pure n-roll position, I
found the following gem, which will teach the newbie a lesson or two if
played as a proposition ...

GNU Backgammon Position ID: /38AAAD3bgMAAA
Match ID : cAkAAAAAAAAE
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
| | | O | 0 points
| | | O |
| | | O |
| | | O |
| | | F |
v| |BAR| | (Cube: 1)
| | | |
| | | X |
| | | X X X |
| | | X X X X X | On roll
| | | X X X X X | X 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: axel
Pip counts: O 15, X 38

(The "F" in gnubg's position means 15 checkers on her ace point.)

You, X, are on roll against the newbie. Cube action?

Best regards

Axel

On 9/27/2022 7:20 PM, Axel Reichert wrote:
> Hello,
>
> if newbies come to our chouette club, they often struggle with the
> concept of wastage, because they do not understand that they cannot
> control the pipcount, only the distribution.
>
> While looking at EPCs in positions close to a pure n-roll position, I
> found the following gem, which will teach the newbie a lesson or two if
> played as a proposition ...
>
> GNU Backgammon Position ID: /38AAAD3bgMAAA
> Match ID : cAkAAAAAAAAE
> +13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
> | | | O | 0 points
> | | | O |
> | | | O |
> | | | O |
> | | | F |
> v| |BAR| | (Cube: 1)
> | | | |
> | | | X |
> | | | X X X |
> | | | X X X X X | On roll
> | | | X X X X X | X 0 points
> +12-11-10--9--8--7-------6--5--4--3--2--1-+ X: axel
> Pip counts: O 15, X 38
>
> (The "F" in gnubg's position means 15 checkers on her ace point.)
>
> You, X, are on roll against the newbie. Cube action?

Well, I've never had occasion to use hexadecimal in backgammon, but
there's a first time for everything.

The fact that O is a newbie is obvious because no seasoned player would
ever bury all 15 checkers. It's also irrelevant, since a shaved monkey
could play O's position as well as the highest settings on XG.

X is off in seven rolls. O needs 8. Is that enough? If O rolls
unanswered doubles, she has a nice recube. Also, there are few, but
non-negligible sequences where X misses. That's enough to make me take a
shake. Doubles lose X's market, but that's only 6 rolls, so not enough
market losers to cube now.

As Borat might say "This is a classic N-roll position with N==7. NOT!"

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

On 9/27/2022 8:34 PM, ah....Clem wrote:
> On 9/27/2022 7:20 PM, Axel Reichert wrote:
>> Hello,
>>
>> if newbies come to our chouette club, they often struggle with the
>> concept of wastage, because they do not understand that they cannot
>> control the pipcount, only the distribution.
>>
>> While looking at EPCs in positions close to a pure n-roll position, I
>> found the following gem, which will teach the newbie a lesson or two if
>> played as a proposition ...
>>
>>   GNU Backgammon  Position ID: /38AAAD3bgMAAA
>>                   Match ID   : cAkAAAAAAAAE
>>   +13-14-15-16-17-18------19-20-21-22-23-24-+     O: gnubg
>>   |                  |   |                O |     0 points
>>   |                  |   |                O |>   |                  |   |                O |
>>   |                  |   |                O |
>>   |                  |   |                F |
>> v|                  |BAR|                  |     (Cube: 1)
>>   |                  |   |                  |
>>   |                  |   |             X    |
>>   |                  |   |          X  X  X |
>>   |                  |   |    X  X  X  X  X |     On roll
>>   |                  |   |    X  X  X  X  X | X   0 points
>>   +12-11-10--9--8--7-------6--5--4--3--2--1-+     X: axel
>> Pip counts: O 15, X 38
>>
>> (The "F" in gnubg's position means 15 checkers on her ace point.)
>

Can we apply a race cube formula here? O has 13 spares on the ace, for a
penalty of 26, plus gaps on the 4 and 5 for another two. Add 28 to her
pipcount to get 43. Add four to X's pipcount for the spares on the ace
and deuce to get 42.

With an adjusted pipcount of 42 to 43, Isight says D/T, Trice says ND/T.

I have no confidence that this is correct, but it makes me less sure of
the ND verdict I came up with in the last post.

And I now recognize how O being a newbie affects the cube decision.
Despite the fact that O can make no checker play errors, she may well
make a cube error, so the practical approach against a human is to ship
it and make it their problem.

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

On 28/09/2022 10:47 am, ah....Clem wrote:

> Can we apply a race cube formula here? O has 13 spares on the ace, for a
> penalty of 26, plus gaps on the 4 and 5 for another two. Add 28 to her
> pipcount to get 43.  Add four to X's pipcount for the spares on the ace
> and deuce to get 42.
>
> With an adjusted pipcount of 42 to 43, Isight says D/T, Trice says ND/T.

The Isight count is actually 44 for O because X has a checker off.

XGID=-EEE---------------cdc----:0:0:1:00:0:0:0:0:10

Score is X:0 O:0. Unlimited Game
+13-14-15-16-17-18------19-20-21-22-23-24-+
| | | O O O |
| | | O O O |
| | | O O O |
| | | O |
| | | |
| |BAR| |
| | | X X X |
| | | X X X |
| | | X X X |
| | | X X X |
| | | X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 30 O: 50 X-O: 0-0
Cube: 1
X on roll, cube action

In Trice's "Backgammon Boot Camp," he writes, "A few years ago,
Position 3-15 was one of my Saturday Night Specials, to be served
up to tournament players looking for a little fast action at the
end of a long day of "serious" match play. "Which position do you
like here? Why don't we play a few games, rolling to see who goes
first? I'll take either side..." Of course, which side stands
better is not the point. The point is that if you understand EPCs
you can handle the cube much more accurately (and without any
apparent effort or thought) than someone who doesn't, even if you
have consumed an adult beverage or two."

---
Tim Chow

On Wednesday, September 28, 2022 at 12:20:24 AM UTC+1, Axel Reichert wrote:
> Hello,
>
> if newbies come to our chouette club, they often struggle with the
> concept of wastage, because they do not understand that they cannot
> control the pipcount, only the distribution.

Indeed, they don't. What is the German for "It doesn't matter where you place
the checkers -- think of the potential energy of the pieces."?
This is a direct quote from someone.

Paul

"ah....Clem" <ah_clem@ymail.com> writes:

> Can we apply a race cube formula here?

Sure. But as you noted correctly in your previous reply, this is not a
likely position to come up in real play rather than an "academic"
exercise, so your mileage may vary ...

Which reminds on some early feedback on my Isight method, when some guy
dismissed it immediately after noticing that it gives the wrong cube
decision for

GNU Backgammon Position ID: Pp8PAACfzwcAAA
Match ID : cAkAAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
| | | O O O | 0 points
| | | O O O |
| | | O O O |
| | | O O O |
| | | O O O |
v| |BAR| | (Cube: 1)
| | | X X X |
| | | X X X |
| | | X X X |
| | | X X X | On roll
| | | X X X | 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: axel
Pip counts: O 60, X 45

I asked him whether it does not make more sense to tune a method to
position that occur in real life ...

> With an adjusted pipcount of 42 to 43, Isight says D/T, Trice says ND/T.

44, as Simon noted, but the verdict remains D/T.

> I have no confidence that this is correct, but it makes me less sure
> of the ND verdict I came up with in the last post.

Big grin ...

> may well make a cube error, so the practical approach against a human
> is to ship it and make it their problem.

I like this kind of real world reasoning!

Axel

On 9/27/2022 10:03 PM, Simon Woodhead wrote:
> On 28/09/2022 10:47 am, ah....Clem wrote:
>
>> Can we apply a race cube formula here? O has 13 spares on the ace, for a
>> penalty of 26, plus gaps on the 4 and 5 for another two. Add 28 to her
>> pipcount to get 43.  Add four to X's pipcount for the spares on the ace
>> and deuce to get 42.
>>
>> With an adjusted pipcount of 42 to 43, Isight says D/T, Trice says ND/T.
>
> The Isight count is actually 44 for O because X has a checker off.
>

OOPs. Missed that. But you only add one for the extra checker on the board.

Trying it again: 13 spares on the ace for 26. One extra checker on the
board to make 27. Empty 4 and 5 points for another two to make 29.

29+ 15 == 44

Add 4 to X's pipcount for the spares on the ace and deuce to get 42.

42 to 44 is D/T according to both Isight and Trice. But I have little
confidence in the race formulas in such an unusual position.

GNU Backgammon Position ID: /38AAAD3bgMAAA
Match ID : cAkAAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg
| | | O | 0 points
| | | O |
| | | O |
| | | O |
| | | F |
v| |BAR| | (Cube: 1)
| | | |
| | | X |
| | | X X X |
| | | X X X X X | On roll
| | | X X X X X | X 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: axel
Pip counts: O 15, X 38

Cube analysis
Rollout cubeless equity +0.576

Cubeful equities:
1. Double, pass +1.000
2. Double, take +1.068 (+0.068)
3. No double +0.867 (-0.133)
Proper cube action: Double, pass

Over the board, Isight gives D/T, so the newbie will get the tricky
take/pass decision. Most likely you will get a take ("I am more than 20
pips ahead!"). I would not have thought from gut feeling that this is a
pass. The EPC for X is 53.9 (my own estimation, see page 28 of my Isight
article, gives 55, O's EPC is of course 57). Trice's doubling criterion
gives 55 + 55/(8 + 1/6) - 3 = 58.7 as point of last take, with a 2-pip
doubling window this gives D/T as well. Even assuming you knew the exact
EPC you will get D/T, since the point of last take is then 57.5 > 57.

In contrast to Tim's position from "Boot Camp", my position is lopsided:
O, on roll has not even a double and only 53.9 % winning chances. This
leads to the idea of constructing a similar position that becomes a
beaver if reversed. Will be even more profitable ...

Rollout details:
Centered 1-cube:
0.788 0.000 0.000 - 0.212 0.000 0.000 CL +0.576 CF +0.867
[0.000 0.000 0.000 - 0.000 0.000 0.000 CL 0.000 CF 0.000]
Player gnubg owns 2-cube:
0.788 0.000 0.000 - 0.212 0.000 0.000 CL +1.153 CF +1.068
[0.000 0.000 0.000 - 0.000 0.000 0.000 CL 0.000 CF 0.000]
Full cubeful rollout with variance reduction
46656 games, Mersenne Twister dice gen. with seed 437730725
and quasi-random dice
Player 0:
Play: 0-ply cubeful [expert]
Cube: 3-ply cubeful prune [grandmaster]
Player 1:
Play: grandmaster 3-ply cubeful prune [grandmaster]
keep the first 0 0-ply moves and up to 16 more moves within equity 0.32
Skip pruning for 1-ply moves.
keep the first 0 2-ply moves and up to 4 more moves within equity 0.08
Cube: 3-ply cubeful prune [grandmaster]

Best regards

Axel

Hello,

so you have doubled, the newbie has wrongly taken, and you roll 21:

GNU Backgammon Position ID: /38AAAD3bgMAAA
Match ID : QQkFAAAAAAAA
+13-14-15-16-17-18------19-20-21-22-23-24-+ O: gnubg (Cube: 2)
| | | O | 0 points
| | | O |
| | | O |
| | | O |
| | | F |
v| |BAR| |
| | | |
| | | X |
| | | X X X |
| | | X X X X X | Rolled 12
| | | X X X X X | X 0 points
+12-11-10--9--8--7-------6--5--4--3--2--1-+ X: axel
Pip counts: O 15, X 38

Erroneously instead of 2/off 1/off you clear the 5-point with 5/4 5/3,
which is a quadruple whopper. Cheerfully the newbie redoubles:

GNU Backgammon Position ID: 994BAID/PwAAAA
Match ID : AQEAAAAAAAAA
+12-11-10--9--8--7-------6--5--4--3--2--1-+ O: gnubg (Cube: 2)
| | | O | 0 points
| | | O | On roll
| | | O |
| | | O |
| | | F |
^| |BAR| |
| | | |
| | | X X |
| | | X X X X |
| | | X X X X |
| | | X X X X | X 0 points
+13-14-15-16-17-18------19-20-21-22-23-24-+ X: axel
Pip counts: O 15, X 35

* gnubg doubles

Alert: wrong double (-0.540)!

Cube analysis
Rollout cubeless equity -0.023

Cubeful equities:
1. No double +0.153
2. Double, pass +1.000 (+0.847)
3. Double, take -0.387 (-0.540)
Proper cube action: No redouble, beaver (38.9%)

It turns out you have a beaver!

Axel

On 10/1/2022 10:23 AM, Axel Reichert wrote:

> Erroneously instead of 2/off 1/off you clear the 5-point with 5/4 5/3,
> which is a quadruple whopper. Cheerfully the newbie redoubles:

> It turns out you have a beaver!

So how high does the probability of the newbie redouble have to be to
justify 5/4 5/3?

---
Tim Chow

Timothy Chow <tchow12000@yahoo.com> writes:

> On 10/1/2022 10:23 AM, Axel Reichert wrote:
>
>> Erroneously instead of 2/off 1/off you clear the 5-point with 5/4 5/3,
>> which is a quadruple whopper. Cheerfully the newbie redoubles:
>
>> It turns out you have a beaver!
>
> So how high does the probability of the newbie redouble have to be to
> justify 5/4 5/3?

I assume we can safely leave this as an exercise for our resident cube
theory and programming expert. (-:

Best regards

Axel

Timothy Chow <tchow12000@yahoo.com> writes:

> On 10/1/2022 10:23 AM, Axel Reichert wrote:
>
>> Erroneously instead of 2/off 1/off you clear the 5-point with 5/4 5/3,
>> which is a quadruple whopper. Cheerfully the newbie redoubles:
>
>> It turns out you have a beaver!
>
> So how high does the probability of the newbie redouble have to be to
> justify 5/4 5/3?

I assume we can safely leave this as an exercise for our resident cube
theory and programming expert. (-:

And my main point was of course that X on roll is D/P while O on roll is
ND/B in this position:

GNU Backgammon Position ID: 994BAID/PwAAAA
Match ID : MAEAAAAAAAAA
+12-11-10--9--8--7-------6--5--4--3--2--1-+ O: gnubg
| | | O | 0 points
| | | O | On roll
| | | O |
| | | O |
| | | F |
^| |BAR| | (Cube: 1)
| | | |
| | | X X |
| | | X X X X |
| | | X X X X |
| | | X X X X | X 0 points
+13-14-15-16-17-18------19-20-21-22-23-24-+ X: axel
Pip counts: O 15, X 35

So as a proposition ("A coin toss decides who is on roll. But the cube
is active.") you will most likely get a wrong double or a wrong take.

Best regards

Axel

On October 1, 2022 at 2:57:35 PM UTC-6, Axel Reichert wrote:

> Timothy Chow <tchow...@yahoo.com> writes:

>> On 10/1/2022 10:23 AM, Axel Reichert wrote:

>>> Erroneously instead of 2/off 1/off you clear
>>> the 5-point with 5/4 5/3, which is a quadruple
>>> whopper. Cheerfully the newbie redoubles:

>>> It turns out you have a beaver!

>> So how high does the probability of the newbie
>> redouble have to be to justify 5/4 5/3?

> I assume we can safely leave this as an exercise for
> our resident cube theory and programming expert. (-:

Not only you can't figure out the answer, you are
talking from behind someone's back. Tsk, tsk! :(

I wonder if Tim asked you to bait you or because
he himself can't figure out his own question..? ;)

Even though I'm not the expert you are gossipping
about, I'll be glad to give my humble opinion if you
tell me the match length and the score, as well as
the newbie's and your ELO or PR ratings?

MK

On October 1, 2022 at 9:26:47 PM UTC-6, MK wrote:

> On October 1, 2022 at 2:57:35 PM UTC-6, Axel Reichert wrote:

>> Timothy Chow <tchow...@yahoo.com> writes:

>>> On 10/1/2022 10:23 AM, Axel Reichert wrote:

>>>> It turns out you have a beaver!

>>> So how high does the probability of the newbie
>>> redouble have to be to justify 5/4 5/3?

>> I assume we can safely leave this as an exercise for
>> our resident cube theory and programming expert. (-:

> Not only you can't figure out the answer, you are
> talking from behind someone's back. Tsk, tsk! :(

> I wonder if Tim asked you to bait you or because
> he himself can't figure out his own question..? ;)

> Even though I'm not the expert you are gossipping
> about, I'll be glad to give my humble opinion if you
> tell me the match length and the score, as well as
> the newbie's and your ELO or PR ratings?

I would have never guessed that I could own you
guys this badly. :)

Just keep jacking off with half-brained position
discussions. Don't push beyond your limits. ;)

MK

On 10/3/2022 2:46 PM, MK wrote:
> I would have never guessed that I could own you
> guys this badly. :)

Not very good at guessing, are you?

---
Tim Chow

On October 3, 2022 at 12:51:00 PM UTC-6, Tim Chow wrote:

> On 10/3/2022 2:46 PM, MK wrote:

>> I would have never guessed that I
>> could own you guys this badly. :)

> Not very good at guessing, are you?

I guess not... ;)

Stop shooting yourself in your dumb ass. :(

MK