concurent easy rules idea

[willemien]

I see no problem. An open ko mouth can be a point, if the other player cannot force the first to connect. Of course, all the "dame" are potential ko threats for the attacker, so this is a rare situation.

Besides, this also reconciles territory with area scoring, where the ko master can delay filling until after all single points ("dame") have been taken.

The question is what happens after all dame have been taken. Is filling then prescribed for the komaster?

maybe i just add open ko to the last rule.

New completer easy rules overview

Rules:

- No suicide.
- A move is either a play or a pass.
- A succession of four passes ends the game.
- after a pass any previous position may reoccur (pass breaks ko- and superko bans)
- Only groups that are pass alive (Benson alive) are alive. (therefore suicide is not allowed)
- Natural komi (I like draws)
- Area scoring.

- seki, superko, open ko and all that lot -> loss for the stronger or Black player (depending if one player is supposed to be stronger)

addition for transforming them in to territory scoring:

- two button go (half point for the player who passes first and half a point for the player who passes last)

[Harleqin]

I see no problem. An open ko mouth can be a point, if the other player cannot force the first to connect. Of course, all the "dame" are potential ko threats for the attacker, so this is a rare situation.

Besides, this also reconciles territory with area scoring, where the ko master can delay filling until after all single points ("dame") have been taken.

The question is what happens after all dame have been taken. Is filling then prescribed for the komaster?

Under area scoring, it would not make any difference whether the ko mouth is filled or not (of course, since any move, while being worthless, would threaten to capture the ko, it is likely safer to fill it as soon as such a worthless move occurs). Under territory scoring, any (gote) play after all dame have been filled would threaten the ko capture, but since such a move also costs a point, the defender can then protect without a loss.

There is no need to prescribe anything for this; optimal play follows directly from the basic rules.

- seki, superko, open ko and all that lot -> loss for the stronger or Black player (depending if one player is supposed to be stronger)

I do not understand this.

[willemien]

There is no need to prescribe anything for this; optimal play follows directly from the basic rules.

We disagree here. to me you expect that each player will at the end play optimal, I lack this optimism (especially when i am one of of them

I think the rules should not expect optimal play in any way.

-- seki, superko, open ko and all that lot -> loss for the stronger or Black player (depending if one player is supposed to be stronger)


I do not understand this.

The background of my rules.

They are made for a (still only fictional) computer program that can reach a perfect strategy on a small board.

I was puzzeling what would be easy and fair rules for this program. and came to the following.

Territory scoring is more complicated than area scoring. (for area scoring only the end position is important, for territory scoring you also need the game history) so area scoring.


Testing of a group is alive is easiest done by testibng for being Benson-pass alive. therefore living means benson alive.

I guess this program can prevent seki, superko, and other problematic situations therefore it is no problem to just declare the program the loser , if they exsist at the end of the game.

(Seki's are for computer programs difficult to recognise, so again just treat them as losses.)


If normal players play eachother and one of the difficult situations is created, the easiest way out is just to declare one of the players as loser. and Black having the advantage of the first move is the unlucky one.
It is a clear and simple rule.

[Harleqin]

You are severely underestimating the state of computer go players. Seki have not been a real source of problems. Besides, optimal strategy on 4x4, for example, produces a whole-board seki.

[RobertJasiek]

willemien, how or when is history needed for territory scoring?

Harleqin, seki may not have been a problem for CG so far but seki could not be classified on the shape level yet. Presumably this will take some more decades.

[Harleqin]

As I see it, the classification of shapes is irrelevant for the rules. There is only 'alive' and 'dead'. 'Alive' is what will not be killed in the playout. The easiest way to get this is to make the "complete game" go until all dead stones have been captured, but allow the players to agree on some shortcut procedure that does not alter the score.

[Bill Spight]

With area scoring, the first button generally makes it irrelevant who gets the last Japanese dame. With territory scoring and pass stones, the second button makes it irrelevant who makes the last pass (with correct play).

But that is a problem you cannot expect correct play

The rules define correct play. A number of supposed rules anomalies show that some set of rules gives a strange result for some end position, but, according to those rules, somebody made a mistake and should not have passed or otherwise ended play. IMHO, altering the rules to yield somehow more reasonable results is not normally a good idea. Who knows what mistakes can give rise to other anomalies? You can't go chasing anomalies. Now, some people, like John Tromp, don't care about anomalies, but if you do, the only ones you have to own are those that result from correct play.

[RobertJasiek]

Bill, I do not quite agree that one could not chase all anomalies:) But... it is still a huge task to find all remaining, yet undiscovered anomalies. And, of course, rulesets designed around all anomalies would be EXTREMELY nasty and long.

[willemien]

willemien, how or when is history needed for territory scoring?

suppose you have the endposition in a game. and you know the komi.

For area scorinbg you can decide the outcome, nothing more is nescesary.

For territory scoring you need to know how many prisoners are taken and that is information of the game history.


Suppose a position can be reached via 2 different paths
in the first path White has captured a black stone.
In the second path black has made a pass.

The position is the same but (for terriotory scoring) the score/ outcome is different. (in path 1 there is one black prisoner more than in path 2)

for area scoring the score is the same so if the position is the same the score is the same.

Example (added later) (beccause you requested it

Code: Select all

+---------------+
| + + + X X O + |
| + + X X O + O |
| + + X O O O + |
| + + O X O + + |
| + X + X O + + |
| + + + X X O + |
| + + + X O O + |
+---------------+

Komi 9

For area scoring it is simple
Black
* 18 territory (empty points)
* 11 live stones
Total 29 points

White
* 9 territory (empty points)
* 11 live stones
* 9 komi
Total 29 points

Result Draw

for territory scoring:

How many prisoners have been taken?
It is not clear of the position, you need the game record.
But if that record is unavailable you cannot decide the result

And not suposing optimimal play you can construct game records for any result here...

[RobertJasiek]

For the purpose of result determination, it suffices to consider numbers of prisoners as being part of the current (final) game situation.

I cannot see your position; I see mostly just a grey box.

EDIT: Now I do not even see the box any longer.

[willemien]

I think we disagree here

For me rules decide about what is legal and illegal. Not about optimal and "normal".

correct play is a bit hazy concept. for me correct is all that is not illegal. pussing correct to the limit woulld mean (and i know i am exagerating here) , You have to play there, other moves allow your opponent kill your group and the rules do not allow you to let that happen. as a matter of fact you are not even allowed to lose. because losing is incorrect play.

With area scoring, the first button generally makes it irrelevant who gets the last Japanese dame. With territory scoring and pass stones, the second button makes it irrelevant who makes the last pass (with correct play).

But that is a problem you cannot expect correct play

The rules define correct play. A number of supposed rules anomalies show that some set of rules gives a strange result for some end position, but, according to those rules, somebody made a mistake and should not have passed or otherwise ended play. IMHO, altering the rules to yield somehow more reasonable results is not normally a good idea. Who knows what mistakes can give rise to other anomalies? You can't go chasing anomalies. Now, some people, like John Tromp, don't care about anomalies, but if you do, the only ones you have to own are those that result from correct play.

[willemien]

As I see it, the classification of shapes is irrelevant for the rules. There is only 'alive' and 'dead'. 'Alive' is what will not be killed in the playout. The easiest way to get this is to make the "complete game" go until all dead stones have been captured, but allow the players to agree on some shortcut procedure that does not alter the score.

It is not that easy:

What do you mean by "complete games"

Playout is a complicated thing . I suppose you mean playout is continuing the game by optimal play and the optimal result is the result of the game. (if it is not by optimal play there can be different outcomes)

A case in point the non filling of ko.
How to continue? May the ko be recaptured now? (so the koban is lifted)

[Harleqin]

As I see it, the classification of shapes is irrelevant for the rules. There is only 'alive' and 'dead'. 'Alive' is what will not be killed in the playout. The easiest way to get this is to make the "complete game" go until all dead stones have been captured, but allow the players to agree on some shortcut procedure that does not alter the score.

It is not that easy:

What do you mean by "complete games"?

A game without the players agreeing on a shortcut. A game is then complete when there are no dead stones on the board anymore, so that the players only need to agree that the game is ended, not on any removal procedure.

Playout is a complicated thing. I suppose you mean playout is continuing the game by optimal play and the optimal result is the result of the game. (If it is not by optimal play, there can be different outcomes.)

I do not prescribe optimal play. "Playout" is simply the part of the game that would usually not be played; i.e. the part of the game after the point where passing does not alter the score. Ikeda calls the last move before that point the "last competitive move".

A case in point: the non-filling of ko.
How to continue? May the ko be recaptured now? (So, the ko ban is lifted?)

I have written before that a ko ban should not be affected by passing. If an open ko mouth cannot be forced to be protected, it can be territory.

[willemien]

I do not prescribe optimal play. "Playout" is simply the part of the game that would usually not be played; i.e. the part of the game after the point where passing does not alter the score. Ikeda calls the last move before that point the "last competitive move".

I was planning a more comprehansive reply but i could not fine the "last competitive move" in my translation of the ing rules can you give me a link to which ing rules you mean? (Then at least we talk about teh same rules)

I have written before that a ko ban should not be affected by passing. If an open ko mouth cannot be forced to be protected, it can be territory.

Then in the situation below (where the opponent has no other option than to pass) it is always a point of territory.

situation:
Player A takes ko
Player B pass (B has no other reasonable options and all dame is filled)
Player A pass
2 passes game ends.

[Harleqin]

I do not prescribe optimal play. "Playout" is simply the part of the game that would usually not be played; i.e. the part of the game after the point where passing does not alter the score. Ikeda calls the last move before that point the "last competitive move".

I was planning a more comprehensive reply, but I could not find the "last competitive move" in my translation of the Ing rules. Can you give me a link to which Ing rules you mean? (Then at least we talk about the same rules.)

I was not talking about Ing but Ikeda (yes, both start with an 'I', but...). There is a translation of his series "On the Rules of Go" at gobase.org: http://gobase.org/studying/rules/ikeda/.

I have written before that a ko ban should not be affected by passing. If an open ko mouth cannot be forced to be protected, it can be territory.

Then in the situation below (where the opponent has no other option than to pass) it is always a point of territory.

situation:
Player A takes ko
Player B passes (B has no other reasonable options and all dame are filled)
Player A passes
2 passes, game ends.

In the result, yes (I cautiously said "can be territory" because I did not want to have a discussion about life and death of the surrounding groups).

I say "in the result" because I do not think that any number of passes should automatically end the game. I rather see a pass as a convenient point to allow the players to agree on the game end.