from your post #32



maybe the date was a typo?

"No result" is an outcome. I'm not sure I follow the logic here, so it might help to explain.

1) Under the Japanese 1989 Rules ("if the players agree"), the players are not required to agree on a "no result" but could continue forever. Then the outcome is undefined.

2) The result "no result" is uncomparable to number scores. Therefore in shapes like one play before either triple ko or double ko seki it is undecidable for the moving player whether creating a triple ko with the outcome "no result" or a double ko seki with a particular number score (like, e.g., zero for a tie in a no komi game) is better. The player cannot meaningfully decide his strategy while this is not his fault. Therefore, at such a moment of the game, there is no meaningful, unambiguous outcome. Instead the game should halt, the rules be corrected and the players should start a replacement game. Also see http://home.snafu.de/jasiek/mistakes.html

1) If the players agree that you have reached one of the conditions that make up a no result, then it is a no result. The cases are basic, so it seems a clear rule.

2) No results are not uncommon in other games and sports. A baseball game that has to end early due to weather is a no result and rescheduled. I prefer the no result to superko, but that is simply opinion. This opinion is one that Nihon Kiin has kept and some others do not.

Mathematically speaking, "no result" is equivalent to the game going on forever (compare with the bottom type in computer science). What would happen if the game were actually to go on forever? No result.

Luckily, it's not math.

It's not?

Oh, yes, of course.

Sorry.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

oren, IF(!) the players agree, then the outcome (but not its relative meaning to an outcome with a score) is clear. If the players do NOT agree, then there is not even an outcome.

The rule is NOT clear for in particular the two reasons I have stated earlier. (More reasons you can find on my webpage.)

It would be rather easy to replace the rule by something clear. Since the Japanese 1989 Rules does NOT replace its "no result" rule by something clear, it remains unclear though.

As I suggest on
http://home.snafu.de/jasiek/j1989c.html
the rule could be replaced by the following clear text:

"
§12 Long cycle

A "position" is the distribution of black or white stones on the board's specific intersections. Recreation of the position after three or more board-plays ends the alternation prematurely with an exceptional result. The result depends on the numbers of black and white stones that have been removed from the board since first leaving the position and until including recreating the position:

* If equally many black and white stones have been removed, then the default result is called "no result", which, only for the purpose of the players' strategic planning, is supposed to equal a tie.
* If fewer black than white stones have been removed, then the result is a win for Black.
* If fewer white than black stones have been removed, then the result is a win for White.

This rule applies only until the first game stop or in between a resumption and the next game stop following the resumption. After a resumption, recreation compares only those positions since the last resumption.
"

This clarifies the rule at least for rules of play in the context of the intention of the Japanese 1989 Rules. In a tournament and given the tournament system and prizes, a player still has to consider which of "no result" or tie is better and the decision might have to rely on mutually conflicting aims.

The usual translation "no result" is a bit misleading.

The term used in the Japanese Rules is 無勝負 and means "no victory nor defeat".

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Dear Robert, never forget the connection between the rules' clauses.

Article 12 about 無勝負 says that the players are entitled to end the game with the result of "no victory nor defeat" if there is a repetion of a whole-board position.

It's a privilege, no obligation.

If they are not willing to do so, they must play on according to the Rules.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Of course. So they continue to play the cycle "forever" and we do not get an outcome.

We do eventually.

Of course you can press common sense into rules, e. g.

"Identical board position and difference of captured stones as more than 4 moves before."

but will this be really necessary ?

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Since the original rule is incomplete, some additions are necessary. Different players might guess different additions, so specified additions are better. Some gaps may be obvious for expert players but not for "beginners"; there must be a solution also for them. Reasonable solutions for some gaps are by far not obvious also for most referees; there must be a solution also for those gaps. The missing strategic interpretation of a no result with the implied undecidable strategy must be clarified even before the start of a game; otherwise players can run into such situations. In summary, all this is necessary for always having an outcome and more specifically having a predictable outcome.

"Identical board position and difference of captured stones as more than 4 moves before." is arbitrary; there is no reason why exactly more than 4 moves. (Triple ko would be played only once while sending-2-returning-1 and eternal life would have to be played twice.)

If the players continue to play a cycle "forever", we do not get an outcome or a clear outcome because these things might happen and it unclear which of these will happen: 1) One player dies before an outcome is reached. 2) Both players die simultaneously before an outcome is reached:) 3) At some time, one player violates the spirit of the aim of the game to win, gives in and likely loses the game, 4) the players change their mind and come to the necessary agreement, 5) the difference of removed stones per cycle does not equal zero and one player is cute enough to realize finally that he can win by departing from the cycle before he will be buried under prisoners, 6) the referee, if any, interferes and makes an unpreditable decision, 7) several cycles could be played and the players play through them in different orders (while some of the cycles might have different differences of removed stones and some of these types could be mixed by strange move orders) so that the right outcome is unclearer than ever, 8) before(!) reaching an outcome, the players are stoned by other tournament participans:)

Your hint

exactly gives the reasons for my choice of "4". There are some mistakes you have to do twice before realising it.

By the way: "sending-2-returning-1" will not repeat identical "situations" in the game ! This is no usecase of "no victory nor defeat".


1), 2), 3), 4), 5), 6), 7), and 8) are true at any moment of the game before it has reached it's "end". They have nothing particular with "no victory nor defeat".

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Playing a cycle is not necessarily a mistake. Not even if the player captures fewer stones during it than the opponent. If the player is in a lost position anyway, he may as well lose by playing through a cycle many times.

For the J1989 rule, it does not matter much that sending-2-returning-1 after 3 moves that are all plays does not repeat the situation: The rule refers to the position - not the situation.

***

(1) and (2) are related to the no result with respect to the likelihood. Without a long cycle, a game consists of fewer than 500 moves and it is not particularly likely for a player to die during them. Contrarily a long cycle game can consist of potentially arbitrarily many moves and in principle the game can continue for years. The makes it much more likely.

(3). Without a long cycle, a player can and should follow the aim of the game to win and, since he has agreed to play the game, there should be no time problem. With a long cycle, the game can suddenly become indefinitely longer and a player might at some time feel a necessity to use (3).

(4), (5). Read the points again. You will notice that it refers to a long cycle specifically and does not refer to other situations of the game.

(6) That other parts of the rules are also flawed does not justify the flaws of the no result result. Here I am referring to these when speaking of unpredictable referee decisions. Let me be more precise: Case 1: The players recur a cycle with an equal number of removed stones. The referee might, e.g., require the players to make a specific agreement or allow the players to continue forever (i.e., he does not intervene). Case 2: The players recur a cycle with an unequal number of removed stones for the, say, 500th time. The referee could let the players continue playing on their own or he could intervene and declare Dagobert Duck (the player with the greater heap of opposing stones) the winner. Case 3: A player points out before or during a game that a strategically undecidable situation could occur or has occurred. The referee could declare the value of such strategic decisions in terms of scores or the referee could do nothing and let the players make meaningless, arbitrary pseudo-strategic decisions. In all the cases, the referee decision is unpredictable and related specifically to long cycles and is not related to other (potential) situations of the game.

(7) Read it again and you will notice that it refers to long cycles and does not refer to ordinary game situations without long cycles.

(8) Well, too much of a joke indeed.

***

You have not discussed my clear rule replacement for §12 yet, as defined in http://home.snafu.de/jasiek/j1989c.html

In particular, it must be discussed why the original no result rule fails to consider its relation to the game phases, why it hides its two major cases (equal versus unequal number of prisoners per cycle) behind the option of a social action ("if the players agree"), why it fails to assess a relation between no result and all the possible scores for purposes of strategic decision making.

Also I would like to know from those liking the original rule why they like also all these gaps and flaws in it. It requires A LOT of commentary to understand the implications on just the rules level of the original rule. This makes it a very nasty rule, IMO. Besides, it is already the second rule and because of it the first ko rule is longer than necessary ("Repetition of a position is prohibited after a succession of exactly 2 moves." instead of "Repetition of a position is prohibited.").

Robert, since this has nothing to do with suicide rule, you may want to create a new topic.

When a thread develops itself into other topics like here, I don't really mind whether discussion proceeds inside or outside the thread. My decision for either is subtle then. Here I have chosen the inside because I replied to earlier remarks by others. When a topic is more independent because there is less relation to making replies, I tend to open new threads. I do not fight for starting above 50% of all threads of a forum though:)

Returning to the original topic, I think that suicide is outlawed in most parts of this world. And banned by many religions.

So introducing something like this into the game of Go makes no sense.

Cutting your last connection to life by yourself is nothing you can be proud of. And I suppose that in Japanese eyes this is absolutely no honourable way to die.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)