XG appears confused about the rules

[peps...@gmail.com]

Here, XG not only lists just one move.
But XG played what it seemed to think
was my "only move" for me.
However, 2/off is (of course) perfectly legal too.
Why was XG acting as if the hit is mandatory?
This sort of thing could really confuse a beginner.

Paul

XGID=-aA-----a---b--a---ecb----:1:-1:1:61:3:1:3:0:10
X:Daniel O:eXtremeGammon

Score is X:3 O:1. Unlimited Game, Jacoby Beaver
+13-14-15-16-17-18------19-20-21-22-23-24-+
| O | | O O O | +---+
| | | O O O | | 2 |
| | | O O | +---+
| | | O |
| | | O |
| |BAR| |
| | | |
| | | |
| | | |
| O | | |
| O O | | X O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 2 O: 130 X-O: 3-1
Cube: 2, O own cube
X to play 61

1. 3-ply 2/1* 1/Off eq:+3.000
Player: 100.00% (G:100.00% B:100.00%)
Opponent: 0.00% (G:0.00% B:0.00%)


eXtreme Gammon Version: 2.10

[MK]

On July 23, 2022 at 7:00:33 PM UTC-6, peps...@gmail.com wrote:

> Here, XG not only lists just one move.

> But .....

> XGID=-aA-----a---b--a---ecb----:1:-1:1:61:3:1:3:0:10

> X to play 61
> 1. 3-ply 2/1* 1/Off eq:+3.000

My XG doesn't do this.
You must have done a truncated rollout.

MK

[peps...@gmail.com]

I didn't do any rollout. This happened during a game.

Paul

[Timothy Chow]

I think you're the one who's confused about the rules. You have
to use both dice if you can. Therefore you must play the 1 first
and then the 6.

[Timothy Chow]

Here's a more consequential position where you presumably would
not make the same mistake of thinking that 6/off is legal.


XGID=-bbA-aA-------------------:1:-1:1:61:0:0:0:0:10

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-+
| | | | +---+
| | | | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | | |
| | | |
| | | |
| | | O O |
| | | X O X O O |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 9 O: 114 X-O: 0-0

Cube: 2, O own cube
X to play 61

1. 3-ply 6/5* 5/Off eq:+0.967
Player: 98.61% (G:0.00% B:0.00%)
Opponent: 1.39% (G:0.00% B:0.00%)

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

[peps...@gmail.com]

On Sunday, July 24, 2022 at 1:37:39 PM UTC+1, Tim Chow wrote:
> I think you're the one who's confused about the rules. You have
> to use both dice if you can. Therefore you must play the 1 first
> and then the 6.

You're exactly correct. My OTB play was 2/off which would, of course,
be accepted by a human opponent (although illegal) (if they hadn't already resigned).

I was then taken aback to have a different play made for me though,
of course, there is only one legal play.

At the highest levels of chess, there have been confusions over the basic rules
so I'm in good company.

Paul

[MK]

On July 24, 2022 at 2:49:59 AM UTC-6, peps...@gmail.com wrote:

> On July 24, 2022 at 8:32:03 AM UTC+1, MK wrote:

>> My XG doesn't do this.
>> You must have done a truncated rollout.

> I didn't do any rollout. This happened during a game.

It was a hiccup joke :) without malice.
I hope it didn't hurt any feelings...

MK

[MK]

On July 24, 2022 at 6:42:36 AM UTC-6, Tim Chow wrote:

> Here's a more consequential position where
> you presumably would not make the same
> mistake of thinking that 6/off is legal.
> XGID=-bbA-aA-------------------:1:-1:1:61:0:0:0:0:10

If you changed the dice from 61 (or also 51) to
non-hitting numbers like 62, 52 or 42, according
to (arbitrary) "rules" 6/off would also be illegal.

However, both Gnubg and XG play 62, 52 or 42
not as 6/5* 5/off but as 6/off at all ply levels.

Your venerated gamblegammon bots are buggy,
inconsistent, unpredictable, inaccurate pieces
of garbage and you folks are full of horseshit... :(

MK

[Timothy Chow]

On 7/24/2022 6:46 PM, MK wrote:
> However, both Gnubg and XG play 62, 52 or 42
> not as 6/5* 5/off but as 6/off at all ply levels.

With a roll of 62, 52, or 42, it is illegal to play 6/5* 5/off.

---
Tim Chow

[Timothy Chow]

I believe that what you meant to say is that you don't like it when
the bots say "6/off" when they mean "6/4 4/off." I personally don't
think there's anything wrong with omitting the intermediate number
when nothing is hit. Though I can see that it can be confusing to
some people in situations like the one below.

XGID=aBBBBBA-------------------:1:1:1:61:0:0:0:0:10

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-+
| | | |
| | | |
| | | |
| | | |
| | | |

| |BAR| |
| | O | |
| | | |
| | | | +---+
| | | X X X X X | | 2 |
| | | X X X X X X | +---+
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 36 O: 25 X-O: 0-0
Cube: 2, X own cube
X to play 61

1. 3-ply 6/Off eq:-0.615
Player: 16.40% (G:0.00% B:0.00%)
Opponent: 83.60% (G:0.00% B:0.00%)

2. 3-ply 6/Off 1/Off eq:-0.697 (-0.083)
Player: 12.89% (G:0.00% B:0.00%)
Opponent: 87.11% (G:0.00% B:0.00%)

[MK]

Sorry for the lazy copy/paste mistake. I'm sure
you are smart (ass) enough to understand that
I meant to write 6/4 4/off. So, what do you have
to say now?

MK

[MK]

On July 24, 2022 at 9:19:29 PM UTC-6, Tim Chow wrote:

> I believe that what you meant to say is that
> you don't like it when the bots say "6/off"
> when they mean "6/4 4/off."

It's not about what I like or not. If the rules is
that "You have to use both dice if you can.",
then the correct play *and* correct notation
of the correct play is "6/4 4/off".

> I personally don't think there's anything wrong
> with omitting the intermediate number when
> nothing is hit.

Did they consult your opinion when coding the
bots? and added an extra statement to say: IF
NOTHING IS HIT THEN OMIT THE INTERMEDIATE
NUMBER??

Why must you pile more bullshit upon bullshit?

To defend the garbage gamblegammon bots
that apparently (and very likely illegally) share
code (exposed by such finger-print evidence)??

There is no such thing as an intermediate number
(or intermediate numbers in case of doublets).

There are 2 dice numbers in the examples here
and according to what you claim to be the rule,
both numbers must be played if possible. End
of story. End of horseshit.

> Though I can see that it can be confusing to
> some people in situations like the one below.
> XGID=aBBBBBA-------------------:1:1:1:61:0:0:0:0:10

> X to play 61

> 1. 3-ply 6/Off eq:-0.615

> 2. 3-ply 6/Off 1/Off eq:-0.697 (-0.083)

After pasting the XGID into Gnubg, I get this:

1. Cubeful 0-ply 6/off Eq.: -0.615
2. Cubeful 0-ply 6/off 1/off Eq.: -0.656 (-0.041)

Are you, by chance, trying to prove that XG shares
uniquely confusing notation code with Gnubg by
providing additional evidence? If so, you are doing
a great job at it... ;)

MK

[Simon Woodhead]

On 25/07/2022 3:34 pm, MK wrote:

> It's not about what I like or not. If the rules is
> that "You have to use both dice if you can.",
> then the correct play *and* correct notation
> of the correct play is "6/4 4/off".

You're a one man conspiracy cult, Murat :-)

There are no notation rules in backgammon.
Common sense and practicality prevail.

[Timothy Chow]

The voice of reason!

Writing 6/off instead of 6/4 4/off is unambiguous. Obviously,
to anyone who knows the rules, 6/off is just short for 6/4 4/off.
There's nothing else it could mean.

---
Tim Chow

[Timothy Chow]

On 7/24/2022 2:59 PM, peps...@gmail.com wrote:
> On Sunday, July 24, 2022 at 1:37:39 PM UTC+1, Tim Chow wrote:
>> I think you're the one who's confused about the rules. You have
>> to use both dice if you can. Therefore you must play the 1 first
>> and then the 6.
>
> You're exactly correct. My OTB play was 2/off which would, of course,
> be accepted by a human opponent (although illegal) (if they hadn't already resigned).

It occurs to me now that the reasoning I gave elsewhere in this
thread---that if a notation is unambiguous then it is fine---would
imply that the notation "2/off" for the only legal play in your
position is also fine! Since hitting is forced, "2/off" can only
mean "2/1* 1/off."

I admit that it would definitely confuse a lot of people if the play
were notated "2/off" yet the opponent's checker ended up on the bar.
Still, I would say that it's not necessarily *wrong*, just confusing.

Below is an even more extreme example. Imagine using the notation
"5/off 4/2" instead of "5/2 4/off"! The final position is the same
either way, so what's the problem? I seem to recall that Tom Keith
gave an example like this one to explain why he changed the user
interface to one of his bots. In his original version, I think if
you were to play 5/off, and then pause to think, and then play 4/2,
the bot would accept your play. In the new version, the bot would
not let you do that; you would have to first undo 5/off, and then
play 5/2, and then it would let you take 4/off.

XGID=a-A-AA--------------------:1:-1:1:63:0:0:3:0:10

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-+

| | | | +---+
| | | | | 2 |
| | | | +---+
| | | |
| | | |
| |BAR| |
| | O | |
| | | |
| | | |
| | | |

| | | X X X |
+12-11-10--9--8--7-------6--5--4--3--2--1-+
Pip count X: 11 O: 25 X-O: 0-0

Cube: 2, O own cube

X to play 63

1. 4-ply 5/2 4/Off eq:+0.930
Player: 96.65% (G:0.00% B:0.00%)
Opponent: 3.35% (G:0.00% B:0.00%)

2. 4-ply 5/Off 4/1 eq:+0.120 (-0.811)
Player: 62.13% (G:0.00% B:0.00%)
Opponent: 37.87% (G:0.00% B:0.00%)

eXtreme Gammon Version: 2.19.207.pre-release

---
Tim Chow

[peps...@gmail.com]

Actually, now that I think about it, I think that I had
a genuine confusion about the rules until I started this thread.

While I have no doubt that your and XG's interpretation of the rules
is correct, I would still maintain that, if you look at many rulesets literally,
it's a perfectly valid interpretation of the rules to say that 6/off without hitting
is perfectly legal.
In fact, I did think a non-hitting 6/off was legal here.
If you wonder how a strongish player like me could be so confused, please bear in mind
that this situation really is extremely rare, so my confusion was able to last for so many years.

All rulesets say (correctly) that you win when you remove all your checkers.
However, the rulesets don't usually say that the objective can't be maintained (and the game
therefore won) mid-move.
I used to think that I could play 6/off and then say "I know there's normally a 1 remaining but
I've obtained my objective mid-move and I've won the game so the task of playing the 1 doesn't remain."

In case you think that this interpretation is ridiculous, a contrast with chess is useful.
In competitive chess, a move consists of moving a piece (or pieces) and then pressing the clock.
The move is not completed until you press your clock.
However, checkmate ends the game mid-move.
Once you have checkmated, you have already won the game, even if you have ignored the clock and not
done that part of the move.

I'm sure XG has got this right, so I'm not doubting you.
But I'd be interested to see where it says in the rules that you can't achieve your objective (and therefore win the game)
in the middle of your move, as you can in chess.

Thank you.

Paul

[peps...@gmail.com]

Sorry, this post doesn't make sense because 6/off doesn't remove all checkers.
In the original position, I still don't see why 2/off (without hitting) is clearly illegal because a player could
argue that they had achieved their objective mid-move and therefore, because the objective is
attained, the 1 need not be played.
This is analogous to checkmating the opponent and then claiming (correctly) that you don't need
to press the clock, even though that is normally an essential component of a move, because you've achieved
your objective mid-move.
I don't think my interpretation of the rules makes any difference in practice from the XG-and-Tim interpretation,
but it doesn't seem clearly wrong.

Could you tell me where it says, in any official set of rules, that a win based on achieving the objective of removing all your checkers
can only be claimed at the end of a move, and not after playing (for example) one of the two dice?

Paul

[Timothy Chow]

On 7/25/2022 7:17 PM, peps...@gmail.com wrote:
> In the original position, I still don't see why 2/off (without hitting) is clearly illegal because a player could
> argue that they had achieved their objective mid-move and therefore, because the objective is
> attained, the 1 need not be played.

Clever! I don't think that backgammon rules of most federations spell
out this sort of thing. In practice, nobody is going to care, so I
doubt that anybody will bother to spell it out in the future.

But if you push the analogy with chess a little further, I think it's
not immediately clear what the conclusion should be. The FIDE laws of
chess specify that checkmate ends the game, *provided* that the move
that delivers checkmate is legal. This is to prevent someone from
making an illegal move, declaring checkmate, and then shutting down
all protests about the illegal move by pointing out that protests
about illegal moves must be lodged before the game ends.

Claiming that the game ends with 2/off is analogous to making an illegal
move to end the game. Legally, you can't take your checker off the
board until you hit your opponent's checker. That is, until the
opponent's checker is hit, your checker is legally *still on the board*
and so the game has not yet ended. So if you want to claim that 2/off
ends the game then you have to accept that it's okay to end the game
with an illegal move. And the chess analogy would suggest that that's
not okay.

---
Tim Chow