Removing Dame Filling from Area Scoring

[Mr. Mormon]

This thread is where to continue the "Removing the Obligation to Fill" discussion from http://senseis.xmp.net/?DameUnderAreaScoring.

[oren]

It seems easier under almost any rules to just fill in the dame. If they're dame, it takes no time anyway and makes counting easier.

[Dusk Eagle]

I don't really see the benefit to removing dame-filling, to be honest. Like oren said, even in territory scoring it is easier to fill in the dame before scoring when filling in real life.

On the sensei's page, you said:

Most would agree that the obligation to fill dame is an undesired side effect of area scoring. Simply disallowing dame plays would fix this, but is the loss of distinction between even and odd dame desirable?

First of all, in my understanding the original Chinese rules were entirely about who could get the most stones on the board. Players were theoretically suppose to fill in their entire territory with stones, leaving room for only two eyes (see http://senseis.xmp.net/?StoneScoring, which I know you already mentioned). So I dispute that filling in dame is an "undesired" or even "unintended" side effect of area scoring.

Second of all, and much more troubling to me, is where you suggest banning dame plays. Why would you ever want to ban them? How would that even be enforced? The white moves below are probably going to end up in dame, but you can't be sure. Should white be banned from playing his moves?

$$Bcm1
$$ | . . . O . O . O . . .
$$ | . . X X X X X X X . .
$$ | . . . . . . . . . . .
$$ | . . . , . . . . . , .
$$ | . . O O 2 4 6 . . . .
$$ | . O . X 1 3 5 7 . . .
$$ | . O . . . . . . . . .

Click Here To Show Diagram Code

[go]$$Bcm1
$$ | . . . O . O . O . . .
$$ | . . X X X X X X X . .
$$ | . . . . . . . . . . .
$$ | . . . , . . . . . , .
$$ | . . O O 2 4 6 . . . .
$$ | . O . X 1 3 5 7 . . .
$$ | . O . . . . . . . . .[/go]

Plus, Go would constantly devolve into fights about what is dame and what isn't. Take, for example, a dame move that carries a non-obvious threat behind it. One player would dispute the move, saying it is illegal, and the other player would be stuck between pointing out how his move threatened something or not being allowed to play his move.

Or, take this shape below:

$$Bcm1
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . O O O O O . .
$$ | . . . O O . . . O O .
$$ | O O O O X X X X X O O
$$ | X X X X T O O O O X .
$$ | . . . X X X X X X X .
$$ | . . . . . . . . . . .

Click Here To Show Diagram Code

[go]$$Bcm1
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . . O O O O O . .
$$ | . . . O O . . . O O .
$$ | O O O O X X X X X O O
$$ | X X X X T O O O O X .
$$ | . . . X X X X X X X .
$$ | . . . . . . . . . . .[/go]

White must fill in all of the three middle points in order to force black to capture his stones (and yes, this still makes a difference in area scoring, because as long as there are more dame elsewhere, black would not want to play at ). However, by your rule, white would be unable to force black to do capture the stones, as all three moves would be on dame points.

Overall, I think banning players from playing in dame points lacks justification, takes away from some of the elegance of Go (i.e. you can no longer play wherever you want), fixes something which isn't broken, and would have many undesirable consequences.

[RobertJasiek]

Teire is not a defined term yet. Judging if a dame threatens teire can be complex. Has anyone shown if it is in EXPTIME or something?:)

[HermanHiddema]

I do not think that people dislike dame filling so much as that they find it uninteresting. If there are 10 dame points, then there are 10! = 3,628,800 different alternating sequences for filling them, with 252 different resulting positions, all with the same resulting score. People play go for the mental challenge, they enjoy thinking about what the best move is, how to get the best result. Since there is no challenge in dame filling, there is no real enjoyment in it for them. Dame filling does, however, have value in preventing mistakes during counting, so I see no reason to disallow filling dame. There is an obvious value to filling them, but I think it is fine if it is quite an informal procedure. Once the last "real" point has been played, players will often "agree to fill the dame" verbally and do it quite quickly and informally anyway.

[HermanHiddema]

Teire is not a defined term yet. Judging if a dame threatens teire can be complex. Has anyone shown if it is in EXPTIME or something?:)

Since judging teire can require evaluating arbitrarily long sequences, including the option of ko and ko threats, it is not any less complex than judging go positions in general, so I would say it has the same computational complexity.

[Toge]

There are often latent threats that need to be taken into account when filling dame. When dame also give points, there's extra incentive to look for those threats. Sometimes it matters to fill dames in correct order, after all.

I think it's a good practice to always fill dame, especially for new players.

[Mr. Mormon]

Dusk Eagle, I think most of us would find filling in territory for ancient Chinese stone scoring boring since we don't have ancient Chinese philosophy. Is there a philosophy behind dame filling now?

Two points are clear to me now:

a) Everyone here so far thinks that the rules should be tailored to skill levels below those where counting is easy and does not need improving.

b) The formal/algorithmic definition of dame can only apply in the endgame, and therefore has to determine when both players agree there are no teire. Hopefully that does no require defining teire.

I have three counterarguments for a): scoring is automatic when playing by computer, by extension counting before actual territory is made should be made easier, and the current values of komi are not tailored to amateurs, which I agree with. Any ideas for b)?

[RobertJasiek]

Since judging teire can require evaluating arbitrarily long sequences, including the option of ko and ko threats, it is not any less complex than judging go positions in general, so I would say it has the same computational complexity.

This my guess, too, but I always want to see the proof;)

[RobertJasiek]

Any ideas for b)?

http://home.snafu.de/jasiek/j2003.html

[willemien]

b) The formal/algorithmic definition of dame can only apply in the endgame, and therefore has to determine when both players agree there are no teire. Hopefully that does no require defining teire.

There is no formal definition of dame (maybe because it does require defining teire)

I have three counterarguments for a): scoring is automatic when playing by computer, by extension counting before actual territory is made should be made easier, and the current values of komi are not tailored to amateurs, which I agree with.

It is bad to rely for scoring on a computer (and no reason to accept it as correct in every situation , and this refers back to the first point)

[HermanHiddema]

...the current values of komi are not tailored to amateurs...

Whuh? Wherever did you get that idea?

[RobertJasiek]

There is no formal definition of dame

Following my URL mentioned above, you find some formal definition of dame. The question is if you want to use that one or whether you are in need of a different appraoch.

[willemien]

There is no formal definition of dame

Following my URL mentioned above, you find some formal definition of dame. The question is if you want to use that one or whether you are in need of a different approach.

You are Right, I am wrong, and I am sorry (for being wrong )

[Mr. Mormon]

...the current values of komi are not tailored to amateurs...

Whuh? Wherever did you get that idea?

Komis usually come from statistics of professional games.