Life In 19x19
http://lifein19x19.com/

Fair komi without ties
http://lifein19x19.com/viewtopic.php?f=45&t=12121
Page 3 of 5

Author:  moha [ Fri Oct 27, 2017 7:53 am ]
Post subject:  Re: Fair komi without ties

billyswong wrote:
So Luigi your first proposal "first passer wins ties" may be more likely to get considered seriously into common use.
Code:
komi 5.5               :  B wins ties
komi 7.5               :  W wins ties
button                 :  first passer / dame parity wins ties
somebody had a ko idea :  first passer in final two passes wins ties (may be more tricky)
or                     :  more time left wins ties
or even                :  lower rating wins ties
or politely            :  women or elder wins ties
or just                :  dicethrow wins ties
all to avoid reality   :  no winner in ties :)

BTW, even with the button isn't there cases where ko threats decide the winner, after competitive play is over? I'm not sure such behaviour is desirable.

Author:  luigi [ Fri Oct 27, 2017 9:01 am ]
Post subject:  Re: Fair komi without ties

Let A be a form of area scoring where, if White passes first, White gets an extra point. Komi is 6.5.

Let B be a form of territory scoring with pass stones where White must pass last if Black made the first pass and Black must pass last if White made the first pass. Komi is 6.5.

Let C be button Go (area scoring) with komi 7.

Bill, if B (a modification of AGA rules) is an implementation of double button Go, as you say, how can it also be equivalent to C?

I'm pretty sure A and B are fully equivalent, and I view them as the purest and best forms of territory scoring (much preferable to the convoluted Japanese rules, which I would replace with one of these right away). But you already convinced me that these weren't fully equivalent to C when you said:

Bill Spight wrote:
But there are one[-sided] dame, which do not have to be played before the button. For instance, suppose that Black has a one way dame and a territory score of 6. Because she does not have to play the one way dame before the button is taken, she will get an area score of 7.5. After the button is taken, she will have a score of 6.5, and then she takes the one way dame for one more point. Also, there are ko fights that may not be finished until after the button is taken, and they may be exceptions, as well. In addition, it is possible to have positions on the board that are equivalent to a button (I have constructed one), so that who gets the button does not matter.

I think this argument applies to Japanese rules as well as A and B. Under any of those rulesets, if Black passes before playing the one-sided dame, White will just end the game by passing in return. As a result, Black wins by 6 on the board and loses by 0.5 after applying 6.5 komi. Compare this to C, where Black wins by 0.5 by taking the button first and then playing the one-sided dame, as you pointed out.

Author:  Bill Spight [ Fri Oct 27, 2017 9:02 am ]
Post subject:  Re: Fair komi without ties

moha wrote:
BTW, even with the button isn't there cases where ko threats decide the winner, after competitive play is over? I'm not sure such behaviour is desirable.


Those situations can arise with a modification of AGA rules that attempts to make a form of territory scoring but does not dictate three passes to end the game under some circumstances. The modification of AGA rules that I set out above avoids those problems by dictating three passes under some circumstances. It just does not always force White to make the last pass.

The problematic situation arises this way. A pass costs one point by handing over a pass stone, and two passes in succession end the play. Player A passes. Player B does not want to pass last, so plays a ko threat, which Player A must answer. Now Player B passes, and if Player A makes the last pass he is penalized one point. So Player A makes a ko threat if possible, and we have a pass fight. We agree that that is undesirable. :)

Author:  Bill Spight [ Fri Oct 27, 2017 9:12 am ]
Post subject:  Re: Fair komi without ties

luigi wrote:
Let A be a form of area scoring where, if White passes first, White gets an extra point. Komi is 6.5.

Let B be a form of territory scoring with pass stones where White must pass last if Black made the first pass and Black must pass last if White made the first pass. Komi is 6.5.

Let C be button Go (area scoring) with komi 7.

Bill, if B (a modification of AGA rules) is an implementation of double button Go, as you say, how can it also be equivalent to C?

I'm pretty sure A and B are fully equivalent, and I view them as the purest and best forms of territory scoring (much preferable to the convoluted Japanese rules, which I would replace with one of these right away). But you already convinced me that these weren't fully equivalent to C when you said:

Bill Spight wrote:
But there are one[-sided] dame, which do not have to be played before the button. For instance, suppose that Black has a one way dame and a territory score of 6. Because she does not have to play the one way dame before the button is taken, she will get an area score of 7.5. After the button is taken, she will have a score of 6.5, and then she takes the one way dame for one more point. Also, there are ko fights that may not be finished until after the button is taken, and they may be exceptions, as well. In addition, it is possible to have positions on the board that are equivalent to a button (I have constructed one), so that who gets the button does not matter.

I think this argument applies to Japanese rules as well as A and B. Under any of those rulesets, if Black passes before playing the one-sided dame, White will just end the game by passing in return. As a result, Black wins by 6 on the board and loses by 0.5 after applying 6.5 komi. Compare this to C, where Black wins by 0.5 by taking the button first and then playing the one-sided dame, as you pointed out.


Yes, if taking the button counts as a pass for ending play, then there could be problems. That is why I do not advocate doing so. In fact, I have objected to ending play with two consecutive passes for over 20 years. Yasunaga, many years ago, proposed a three pass rule, and Ing has a four pass rule. But I think that the Mind Sports rules end play with two consecutive passes, and count "taking the button" as a pass. I am afraid that I am swimming against the tide.

Edit: Also, taking the button should lift any ko or superko ban.

Author:  luigi [ Fri Oct 27, 2017 9:33 am ]
Post subject:  Re: Fair komi without ties

Bill Spight wrote:
Yes, if taking the button counts as a pass for ending play, then there could be problems. That is why I do not advocate doing so. In fact, I have objected to ending play with two consecutive passes for over 20 years. Yasunaga, many years ago, proposed a three pass rule, and Ing has a four pass rule. But I think that the Mind Sports rules end play with two consecutive passes, and count "taking the button" as a pass. I am afraid that I am swimming against the tide.

Now I'm confused. I think the main reason why C is different from A and B is that taking the button doesn't count as a pass for ending play. (EDIT: Nevermind. I think this is what you meant in the first place.)

Also, the Mind Sports rules don't use the button at all. Right?

Author:  moha [ Fri Oct 27, 2017 10:44 am ]
Post subject:  Re: Fair komi without ties

Bill Spight wrote:
moha wrote:
BTW, even with the button isn't there cases where ko threats decide the winner, after competitive play is over? I'm not sure such behaviour is desirable.

Those situations can arise with a modification of AGA rules that attempts to make a form of territory scoring but does not dictate three passes to end the game under some circumstances. The modification of AGA rules that I set out above avoids those problems by dictating three passes under some circumstances. It just does not always force White to make the last pass.

The problematic situation arises this way. A pass costs one point by handing over a pass stone, and two passes in succession end the play. Player A passes. Player B does not want to pass last, so plays a ko threat, which Player A must answer. Now Player B passes, and if Player A makes the last pass he is penalized one point. So Player A makes a ko threat if possible, and we have a pass fight. We agree that that is undesirable. :)
I thought about simpler and more general cases, where there are only protective plays left. And W has to start making them (and lose) if he has less threats, but can postpone playing them and take the button first if he has more threats. So cases that are normally jigo, but with button the side with more threats wins.

Author:  Bill Spight [ Fri Oct 27, 2017 12:07 pm ]
Post subject:  Re: Fair komi without ties

moha wrote:
I thought about simpler and more general cases, where there are only protective plays left. And W has to start making them (and lose) if he has less threats, but can postpone playing them and take the button first if he has more threats. So cases that are normally jigo, but with button the side with more threats wins.


I don't see how that matters, since making the threat with sente does not alter the score. E. g., there are no dame left that are not threats and only one threat by Black, and White is to play:

1) White "takes the button", Black plays the threat, and White makes the protective play;

2) White makes the protective play, leaving a dame which is no longer a threat, Black plays the dame, and White "takes the button".

All same same. :D

Author:  Bill Spight [ Fri Oct 27, 2017 12:17 pm ]
Post subject:  Re: Fair komi without ties

luigi wrote:
Bill Spight wrote:
Yes, if taking the button counts as a pass for ending play, then there could be problems. That is why I do not advocate doing so. In fact, I have objected to ending play with two consecutive passes for over 20 years. Yasunaga, many years ago, proposed a three pass rule, and Ing has a four pass rule. But I think that the Mind Sports rules end play with two consecutive passes, and count "taking the button" as a pass. I am afraid that I am swimming against the tide.

Now I'm confused. I think the main reason why C is different from A and B is that taking the button doesn't count as a pass for ending play. (EDIT: Nevermind. I think this is what you meant in the first place.)

Also, the Mind Sports rules don't use the button at all. Right?


I found my copy of the Mind Sports 2008 rules. They use the first pass to implement the button. If White makes the first pass, then Black subtracts one point from his score. Play ends with two consecutive passes. They do not allow resumption of play to fill one way dame. It seems that they do allow resumption to fight a ko, but any ko ban before the consecutive passes remains in effect. { :shock: shocked but not surprised ;)}

Author:  moha [ Fri Oct 27, 2017 12:27 pm ]
Post subject:  Re: Fair komi without ties

Bill Spight wrote:
moha wrote:
I thought about simpler and more general cases, where there are only protective plays left. And W has to start making them (and lose) if he has less threats, but can postpone playing them and take the button first if he has more threats. So cases that are normally jigo, but with button the side with more threats wins.
I meant something like this:
Click Here To Show Diagram Code
[go]$$c B connects
$$ ---------
$$ | . O . . . . . . . |
$$ | W X O O O O O . . |
$$ | . . X X X X O . . |
$$ | X X X . X O O . . |
$$ | . . . X X O . O . |
$$ | . . . X O O . . . |
$$ | . . . . X O O O O |
$$ | . . . X . X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]

(And if I count right this is a tie that B loses with the tiger at bottom but would win with a solid connection.)

Author:  luigi [ Fri Oct 27, 2017 12:49 pm ]
Post subject:  Re: Fair komi without ties

Bill Spight wrote:
I found my copy of the Mind Sports 2008 rules. They use the first pass to implement the button. If White makes the first pass, then Black subtracts one point from his score. Play ends with two consecutive passes.

My whole point is that this is neither Button Go (because there is no button) nor equivalent to Button Go, but just territory scoring done right.

It's not fully equivalent to Button Go because of one-sided seki and other stuff, as you pointed out.

Author:  pookpooi [ Fri Oct 27, 2017 4:36 pm ]
Post subject:  Re: Fair komi without ties

Any other fair komi without ties idea without involving who play last or who pass first or who pass last or add/substract point based on any additional condition other than existing rules?

Author:  Bill Spight [ Fri Oct 27, 2017 6:08 pm ]
Post subject:  Re: Fair komi without ties

moha wrote:
Bill Spight wrote:
moha wrote:
I thought about simpler and more general cases, where there are only protective plays left. And W has to start making them (and lose) if he has less threats, but can postpone playing them and take the button first if he has more threats. So cases that are normally jigo, but with button the side with more threats wins.
I meant something like this:
Click Here To Show Diagram Code
[go]$$c B connects
$$ ---------
$$ | . O . . . . . . . |
$$ | W X O O O O O . . |
$$ | . . X X X X O . . |
$$ | X X X . X O O . . |
$$ | . . . X X O . O . |
$$ | . . . X O O . . . |
$$ | . . . . X O O O O |
$$ | . . . X . X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]

(And if I count right this is a tie that B loses with the tiger at bottom but would win with a solid connection.)


Click Here To Show Diagram Code
[go]$$cm37 W is ko monster
$$ ---------
$$ | 5 O 4 . . . . . . |
$$ | O X O O O O O . . |
$$ | 3 1 X X X X O . . |
$$ | X X X . X O O . . |
$$ | . . . X X O . O . |
$$ | . . . X O O . . . |
$$ | . . 7 6 X O O O O |
$$ | . . . X 9 X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]

:w38: takes the button, :w44: takes ko, :w46: fills ko

:b37: leaves a ko fight which White can win by connecting at 40. However, White is komonster, which means that White can delay winning the ko, so :w38: takes the button. :b45: could be a pass, OC.

Being komonster means that White can delay winning the ko until Black passes (or plays inside territory), and can also take the button instead of Black. Without the button White could pass with :w38: instead, but the button is worth ½ pt. more than the pass.

Yes, the button affects the scoring when there is a ko fight that continues past the dame stage. :) That's not the same as just having protective plays left.

Author:  moha [ Fri Oct 27, 2017 7:21 pm ]
Post subject:  Re: Fair komi without ties

Bill Spight wrote:
Yes, the button affects the scoring when there is a ko fight that continues past the dame stage. :) That's not the same as just having protective plays left.
But there is no ko fight here normally, that only arises when W refuses to make his mandatory protective plays in time. And scoring an otherwise identical jigo position as W+ or B+ based on the type of an earlier connection B made at the bottom seems doubtful.

Author:  Bill Spight [ Fri Oct 27, 2017 9:07 pm ]
Post subject:  Re: Fair komi without ties

moha wrote:
Bill Spight wrote:
Yes, the button affects the scoring when there is a ko fight that continues past the dame stage. :) That's not the same as just having protective plays left.
But there is no ko fight here normally, that only arises when W refuses to make his mandatory protective plays in time. And scoring an otherwise identical jigo position as W+ or B+ based on the type of an earlier connection B made at the bottom seems doubtful.


Of course a ko fight is normal here. For instance:

Click Here To Show Diagram Code
[go]$$cm37 B connects
$$ ---------
$$ | 5 O 4 . . . . . . |
$$ | O X O O O O O . . |
$$ | 3 1 X X X X O . . |
$$ | X X X . 2 O O . . |
$$ | . . . X X O . O . |
$$ | . X . X O O . . . |
$$ | . . . . X O O O O |
$$ | . . . X . X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]


On this board :w38: is normal here, rather than winning the ko. It just happens on the previous board that White can delay winning the ko past the dame stage. The ability to do so is not the usual state of affairs. Winning the ko is not a protective play.

Click Here To Show Diagram Code
[go]$$cm39 Not jigo
$$ ---------
$$ | . O . . . . . . . |
$$ | O X O O O O O . . |
$$ | . X X X X X O . . |
$$ | X X X . X O O . . |
$$ | . . . X X O . O . |
$$ | . . . X O O . . . |
$$ | . . . . X O O O O |
$$ | . . . X . X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]


Let's be clear. This is not a jigo position. It is a ko position.

Click Here To Show Diagram Code
[go]$$cm39 Atari
$$ ---------
$$ | . O W . . . . . . |
$$ | O X O O O O O . . |
$$ | 1 X X X X X O . . |
$$ | X X X . X O O . . |
$$ | . . . X X O . O . |
$$ | . . . X O O . . . |
$$ | . . . . X O O O O |
$$ | . . . X . X X X X |
$$ | . . . . . . . . . |
$$ ---------[/go]


Here, connecting the ko is a normal protective play, but with button go Black can, and should, simply take the button instead of making the ko. Note that connecting the ko is a protective play only because White has already played :wc:. Playing :wc: on the previous board is not a protective play, it is winning the ko.

Author:  moha [ Sat Oct 28, 2017 6:27 am ]
Post subject:  Re: Fair komi without ties

I think on your first board the added empty points are basically points that can still be contested. That's why the behaviour changes.

The second board seems identical to the original, but you say "not jigo". I guess because W not yet made the first connection, which you consider a competitive play? It would be ko if it would be B's turn (or if W passes). But if W's turn it seems jigo both in Chinese and Japanese.
Bill Spight wrote:
Playing :wc: on the previous board is not a protective play, it is winning the ko.
Ok, I can see the logic in this. (although I think it prevents the ko, not wins it)

And still, I'm not sure the button behaviour here is desirable. Or if resolving a jigo by threats is ok, the button seems redundant and unnecessary. :)

Author:  Bill Spight [ Sat Oct 28, 2017 9:51 am ]
Post subject:  Re: Fair komi without ties

moha wrote:
I think on your first board the added empty points are basically points that can still be contested. That's why the behaviour changes.


That behavior change reveals that the top left is a ko position. :) That's why I say that White's play there is not a protective play. Locally it gains ⅓ pt. or 1⅓ pt., depending on the rules.

Quote:
The second board seems identical to the original, but you say "not jigo". I guess because W not yet made the first connection, which you consider a competitive play?


Well, that depends upon the exact definition of competitive play, but there is a ko fight in this position, even if we can predict who wins it. Chinese/AGA and Korean/Japanese rules treat the ko fight differently, because when White takes the ko back Black must pass (or play inside territory). Chinese/AGA rules treat the pass as the same as filling one's own territory (as long as it does not make it dead); Korean/Japanese rules treat the pass as the same as filling a dame, which is one point better than filling territory.

Quote:
It would be ko if it would be B's turn (or if W passes). But if W's turn it seems jigo both in Chinese and Japanese.


The result, with best play, would be jigo. Traditional baduk evaluation and characterization of positions does not depend upon who has the move. Unlike one form of game theory. (Combinatorial game theory, OTOH, agrees with baduk tradition.)

Quote:
And still, I'm not sure the button behaviour here is desirable. Or if resolving a jigo by threats is ok, the button seems redundant and unnecessary. :)


The basic idea of button go is to unite area and territory scoring by being a hybrid of the two. It adds a play that gains ½ pt. by area scoring or loses ½ pt. by territory scoring. (Button go uses area style scoring.) Under normal circumstances who gets the last dame matters to the area score, despite the fact that correct play by territory scoring is also correct play by area scoring. Adding the button in such circumstances makes who gets the last dame irrelevant. If Black gets it, she gains 1 pt., but then White gets the button, so the net gain to Black is only ½ pt. If White gets the last dame, Black gets the button, also for a gain of ½ pt. Under ordinary circumstances in button go, who gets the last dame does not matter, as with Korean/Japanese rules, but life and death questions can be resolved by play, as with Chinese/AGA rules. Button go has the best of both worlds. :)

Button go normally produces scores ½ pt. better for Black than Korean/Japanese scoring would, but there are some exceptions. In some very rare cases there are plays that, like the button, gain less than filling a dame. They have been constructed on the board, but none is know to have occurred in actual play. In rare cases there are one-way dame which are worth a point by Chinese/AGA scoring and button go, but not by Korean/Japanese scoring. And occasionally there are ko fights which are not resolved while there are dame left, which is the case in your example.

Black is an odd number of points ahead on the board by both area and territory rules, so normally Black would take the button. But because White can win the ko fight, White can take the button and then fight and win the ko. The result is then ½ pt. less for Black instead of ½ pt. more. One may consider this an undesirable side effect of Button go, but it is an exceptional case, not a common one.

Suppose, instead, that there was also a dame on the board besides the ko. If White had no ko threat, then White would resolve the ko and Black would take the dame. The area score would be 1 pt. better for Black than the territory score, and the button score would be ½ pt. better. The hybrid score lies in between the two other scores. If White had a ko threat, then White would take the dame and later win the ko fight. In button go Black would get the button. The area score would be 1 pt. worse for Black than the territory score, and the button score would be ½ pt. worse. The button score is not better for Black than the territory score but it would still lie in between the area score and territory score. One might regard the fact that the score by button go lies in between the area and territory scores as a desirable side effect, but such kos are still uncommon. :)

Anyway, button go is a hybrid of area and territory scoring, and there are pluses and minuses to that. :)

Author:  billyswong [ Sat Oct 28, 2017 8:19 pm ]
Post subject:  Re: Fair komi without ties

The actual "minuses" of button go / first passer win jigo now looks like the side-effect of players may in some rare cases, as illustrated above, want to pass or take button before everything has been played out, thus changed the nominal play sequence. The result looks fair but the gameplay looks 'abnormal'. In traditional scorings, when one pass and the opponent pass too, it should be that there will be no reason for one not to pass again, thus the 2-pass-end-a-game being implemented in most modern systems. Now with first passer win jigo, not necessarily true anymore.

Assuming we don't have "broken stones" on hand, how should players implement the 1/2 point gain by pass/take-button using only traditional go equipment?

EDIT:
For the board below, does the two circled points counted as black's area in area scoring if the game end as-is? If they are counted, then there will be less concern of "pass/take button first then has to play move again even if opponent passed too"
Click Here To Show Diagram Code
[go]$$
$$-----------
$$|...XCBO..|
$$|....XOO..|
$$|....XO...|
$$|....XO...|
$$|....XO...|
$$|...XXO...|
$$|...XOOO..|
$$|...XO....|
$$|...XXO...|
$$-----------[/go]


EDIT2:I constucted a case where play-after-pass/button will always happen.
Click Here To Show Diagram Code
[go]$$ W first
$$-----------
$$|.OX..X...|
$$|.OXXXX...|
$$|.OOOOX...|
$$|....OX...|
$$|OOOOOXXXX|
$$|XXXXXOOOO|
$$|XOO.XO...|
$$|COO.XO.O.|
$$|XXXXXO...|
$$-----------[/go]

Author:  Bill Spight [ Sat Oct 28, 2017 9:09 pm ]
Post subject:  Re: Fair komi without ties

billyswong wrote:
The actual "minuses" of button go / first passer win jigo now looks like the side-effect


Sorry if I have made things confusing. By area scoring, the player to take the button gains ½ pt., and therefore wins what would otherwise have been jigo if he had to pass instead. That is not a side effect, it is the intended effect.

In the example that moha gave, that is still the case. White takes the button and wins what otherwise would have been jigo, with a komi of 7. However, the number of stones on the board is even, which means that Black would usually win with a komi of 7, because White would fill the last dame and Black would take the button.

Quote:
In traditional scorings, when one pass and the opponent pass too, it should be that there will be no reason for one not to pass again, thus the 2-pass-end-a-game being implemented in most modern systems.


But that is not always so, as I have been saying for over 20 years, and that is why Ing rules require four passes, with fighting ko restrictions lifted after two passes. However, as Terry Benson said to me once, people would rather have simple rules than more complicated rules to handle very rare cases.

Quote:
Now with first passer win jigo, not necessarily true anymore.


You don't have to implement taking the button as a pass. Best not to, IMO. And such a rule is not complicated or difficult to understand.

Quote:
Assuming we don't have "broken stones" on hand, how should players implement the 1/2 point gain by pass/take-button using only traditional go equipment?


Except for regarding taking the button as a pass, the Mind Sports rules get it right. Psychologically, using a card or token that someone takes may be best, but you can simply have a player announce that they are taking the button instead of playing a stone on the board. If that person is White, then Black subtracts one point from his final score. Instead of a komi of 7, have Black give a komi of 6½. Giving the lower komi compensates for subtracting one point from Black's score about half the time. :)

Author:  billyswong [ Sat Oct 28, 2017 9:39 pm ]
Post subject:  Re: Fair komi without ties

I said "side-effect" because the gameplay deviate from the 'optimal jigo sequence', telling God to insert a premature pass / take button before all stones played out in order to score a win. The intended effect, I assume, is to avoid the situation of mainstream fractional komi that make all 'optimal jigo sequence' unilaterally assigned as a win of one color.

Author:  Bill Spight [ Sat Oct 28, 2017 10:07 pm ]
Post subject:  Re: Fair komi without ties

billyswong wrote:
I said "side-effect" because the gameplay deviate from the 'optimal jigo sequence', telling God to insert a premature pass / take button before all stones played out in order to score a win. The intended effect, I assume, is to avoid the situation of mainstream fractional komi that make all 'optimal jigo sequence' unilaterally assigned as a win of one color.


The overall intended effect of the button is to combine area and territory rules. The specific intended effect is to make who gets the last dame irrelevant. And yes, by area scoring with an integer komi, each player will win a tie on the board half the time. The intention has nothing to do with fractional komi. Mind Sports implements a kind of button with a fractional komi.

Page 3 of 5 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/