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

Long cycle repetition with continued clock
http://prod.lifein19x19.com/viewtopic.php?f=45&t=12285		 Page 1 of 1
Author:  RobertJasiek [ Tue Sep 22, 2015 8:12 am ]
Post subject:  Long cycle repetition with continued clock
Quotation reference: viewtopic.php?p=193464#p193464

HermanHiddema wrote:
Simple solution: Do not reset the clock.

Just clear the board and restart the game. This means the players may be in byoyomi immediately, if the triple ko occurs late in the game. This will probably add a little time to the round, as the players may play more byoyomi periods, but not 3 hours.

Effectively, with this rule, the players have to get a result within the time limits given for the game, even if it restarts.


Your proposal is not a general solution because the players can also create a long cycle in every subsequent replacement game by, e.g., applying the eternal life joseki variation. Eventually, in order to break no-results / ties, the game result needs to be decided by lot.
Author:  HermanHiddema [ Tue Sep 22, 2015 8:52 am ]
Post subject:  Re: Long cycle repetition with continued clock
RobertJasiek wrote:
Quotation reference: viewtopic.php?p=193464#p193464

HermanHiddema wrote:
Simple solution: Do not reset the clock.

Just clear the board and restart the game. This means the players may be in byoyomi immediately, if the triple ko occurs late in the game. This will probably add a little time to the round, as the players may play more byoyomi periods, but not 3 hours.

Effectively, with this rule, the players have to get a result within the time limits given for the game, even if it restarts.


Your proposal is not a general solution because the players can also create a long cycle in every subsequent replacement game by, e.g., applying the eternal life joseki variation. Eventually, in order to break no-results / ties, the game result needs to be decided by lot.


It is not supposed to be an solution to collusion by the players. If the players deliberately refuse to get a result, you can just disqualify them or record a loss for both players.
Author:  RobertJasiek [ Tue Sep 22, 2015 10:06 am ]
Post subject:  Re: Long cycle repetition with continued clock
If both players know the suitable joseki variant, their cooperation is not needed but they can generate the long cycle nevertheless, legally and in good sportsmanship. It is the tournament rules to be blamed, not the players.
Author:  skydyr [ Tue Sep 22, 2015 10:18 am ]
Post subject:  Re: Long cycle repetition with continued clock
RobertJasiek wrote:
If both players know the suitable joseki variant, their cooperation is not needed but they can generate the long cycle nevertheless, legally and in good sportsmanship. It is the tournament rules to be blamed, not the players.


Well, it takes two to tango. There are viable decision points at the start of every and any joseki.
Author:  RobertJasiek [ Tue Sep 22, 2015 10:28 am ]
Post subject:  Re: Long cycle repetition with continued clock
It is possible that both players find no-result a good result for themselves, e.g., because each player thinks that his winning chances against the particular opponent would be smaller than 50%. Then each player has a good reason to make strategic choices towards no-result in the creation of the joseki variant.

A better solution for the replacement game is, OC, to use (positional) superko. This combined with the continued clock works. (It remains unclear why superko would not be used from the beginning if avoided ties is that important in a particular tournament.)
Author:  oren [ Tue Sep 22, 2015 10:39 am ]
Post subject:  Re: Long cycle repetition with continued clock
Superko is not a good solution in my opinion. A tie or replay is fine.

I'm not a fan of having to pull back multiple moves to figure out the start of a mess.
Author:  Pio2001 [ Tue Sep 22, 2015 11:43 am ]
Post subject:  Re: Long cycle repetition with continued clock
I find it difficult to choose a universal solution. The superko rule leads to a disastrous result in case of a "molasses ko".
The molasses ko is known to slow down the game to 20 % of its speed, but it has also another effect under strict superko rulesets (be them positional, situational, or natural-situational) : the first player to pass looses the group !

Worst case scenario : a molasses ko appears around move 100. The game goes on with five stones used for each move (one for the move, four to avoid loosing the molasses ko). Let's say that the yose ends around move 250. It means that the last 150 moves were played with the molasses ko still pending, which makes 100 + 150 x 5 = 850 actual moves !
Now, the winner depends on who wins the molasses ko. The players are then forced to play inside their territories, or inside their opponents (superko usually go together with area counting) in order not to loose their group.

And now, a battle begins (that doesn't exist in asian go !) where the players try to build shapes inside their territories that give their opponent the least possible legal moves, while leaving themselves the most possible legal moves. That is creating the most possible single eyes with the least possible stones.

That may lead the game to about 400 moves, not counting the molasses ko itself, thus 150 more after the yose, which gives 150 x 5 = 750 actual moves (the molasses ko still works !), for a grand total of 850 + 750 = 1600 moves !
Author:  RobertJasiek [ Tue Sep 22, 2015 11:58 am ]
Post subject:  Re: Long cycle repetition with continued clock
It depends on which superko rule is used whether molasses ko fights occur. In the history of go, there has been 1 game reported with molasses ko. 5 times the usual record of move numbers is quite an appropriate celebration for such an incident!
Author:  Matti [ Wed Sep 23, 2015 2:30 am ]
Post subject:  Re: Long cycle repetition with continued clock
Also if the molasses ko generates a no-pass-go endgame, then white cannot win by komi. To win the game he must win on the board.
Author:  asura [ Wed Sep 23, 2015 8:28 am ]
Post subject:  Re: Long cycle repetition with continued clock
Pio2001 wrote:
The molasses ko is known to slow down the game to 20 % of its speed, but it has also another effect under strict superko rulesets (be them positional, situational, or natural-situational) : the first player to pass looses the group !


With situation superko there is no pass fight and the molasses ko remains on the board.
Author:  Matti [ Thu Sep 24, 2015 4:34 am ]
Post subject:  Re: Long cycle repetition with continued clock
Anyway, if players happen to get two molasses kos a small and a big one. the game slows down to 1/25 speed. First play in the big molasses ko. When you are not allowed to play there play a move in the small one. Play another round in the big ko. Play another move in the small ko. when you cannot play in neither ko paly elsewhere. Repeat.
Author:  Bantari [ Thu Sep 24, 2015 9:59 am ]
Post subject:  Re: Long cycle repetition with continued clock
What's wrong with simply declaring a draw in any such cases?
Go players seem to be panickly afraid of draws. But a draw is actually a good solution, I think.
Author:  Kirby [ Thu Sep 24, 2015 10:27 am ]
Post subject:  Re: Long cycle repetition with continued clock
Bantari wrote:
What's wrong with simply declaring a draw in any such cases?
Go players seem to be panickly afraid of draws. But a draw is actually a good solution, I think.


Depends on what's on the line. If it's just a friendly game at your local go club, then yeah - just have a draw, laugh about it, and play another game.

But if it's in a big tournament, say with a lot of money on the line... Players care a little bit more about having a winner :-)
Author:  oren [ Thu Sep 24, 2015 10:34 am ]
Post subject:  Re: Long cycle repetition with continued clock
Bantari wrote:
What's wrong with simply declaring a draw in any such cases?
Go players seem to be panickly afraid of draws. But a draw is actually a good solution, I think.


I think most are fine with a draw. You find a few who can't stand it though. :)
Author:  HermanHiddema [ Thu Sep 24, 2015 10:39 am ]
Post subject:  Re: Long cycle repetition with continued clock
Bantari wrote:
What's wrong with simply declaring a draw in any such cases?
Go players seem to be panickly afraid of draws. But a draw is actually a good solution, I think.


The context of the original thread got lost here. I proposed this for those situations where a draw or other such solution is not possible, such as in a knock-out tournament. Nothing wrong with a draw otherwise.
Author:  Pio2001 [ Thu Sep 24, 2015 12:25 pm ]
Post subject:  Re: Long cycle repetition with continued clock
asura wrote:
With situation superko there is no pass fight and the molasses ko remains on the board.


That's true :shock:
Author:  RobertJasiek [ Thu Sep 24, 2015 10:28 pm ]
Post subject:  Re: Long cycle repetition with continued clock
Theory and application of molasses ko types are explained in http://home.snafu.de/jasiek/ko_types.pdf

Money is divisible in the case of a tied game. Single seeding places are not.
Author:  Pio2001 [ Fri Sep 25, 2015 3:59 am ]
Post subject:  Re: Long cycle repetition with continued clock
Thank you for the link (and the article), Robert.
Page 1 of 1 All times are UTC - 8 hours [ DST ]
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group
http://www.phpbb.com/