Idea. for go....

what if we had a tough programm that at the 110th move played both sides to see a win. if it couldnot see a win for one side the other player would simply chose score and game is over.

that could be good for chess, is so predictable...but for go such program dont exist

Nice idea in theory, but it just couldn't work.

Consider for example a complex tsumego problem where both players have read out that it can't be killed. If the program reads out the problem incorrectly it will get the wrong result.

Similarly, say black has a huge moyo and white has two invasion points. Since white has two invasion points, white treats the invasion points as miai and plays elsewhere. But if the program only finds one invasion point (perhaps the other is too complex) then it will get the wrong result.

As an alternative, how about permitting players to let a computer program take over from them. Gnugo should do. Their opponent will know that the computer has taken over from them so it shouldn't be too much of a problem for beginner vs beginner matches. This idea still has problems.

There isn't even a good program to decide the winner at the end of the game!

I also don't understand why you would like that? It's like a game of basket or soccer or football where at the halftime you decide to let the leader win. It is supposed to take a long time those go games...

I play so many games that are easy for my friends to see have gone south by my 50th stone.... it just seems like my openings would be stronger if i could see the logical out come as easily as my friends. i would at least like to appear conscious:))))

Games are won by the person who makes the second-to-last mistake.

In other words, you can perhaps write a program that estimates the probability that one side or the other will win. But you can never predict surprise moves, "brain freezes", and other human error conditions. I always try to play my games out for two reasons: 1) practice not making mistakes when I'm ahead, and 2) practice not making mistakes when I'm behind.

;)

Let's just add: Practice for utilizing opponnents mistake:)

I find it interesting how many people here are against this idea _on principal_(as opposed to being opposed to it because it is infeasable).

In countries that play go, it is considered rude to continue playing after you have lost. How rude depends on the culture -- Koreans tend to continue for longer than Chinese who continue for longer than Japanese. But even Korean's will only play to win -- they'll make a desperate attack hoping to come back, and if fails they will resign. American beginners, (perhaps from their chess background), will continue long after they've lost in the hopes of a 'mistake' -- aka falling asleep at the wheel.

This attitude that you should resign when you're too far behind differs fairly markedly from every western game I can think of. The key difference is perhaps that go is not so much about tricking the opponent into making a mistake as in playing better moves. I'm fairly happy to tell my opponent my intention when I play a move -- though I prefer them to be able to read it from the stone I played. The thinking goes somewhere along these lines:

"Well, I'm clearly behind now. If I continue playing then they might make a mistake which will mean I win. However, this person has already played better than me to get this lead. Why should they suddenly start playing worse than me. I won't waste their time".

Personally, I blacklist anybody who doesn't resign when they're behind. I'm playing a game against 'rotter' at the moment who appears to live up to the name. I got a commanding lead but the didn't resign, so I had no choice to continue. Since then I've killed all their stones and started passing but they _still_ won't resign!? What am I supposed to do now? I play to win by a handful of points, I take no pride in killing all of my opponent's stones.

For games like that, I'd love to be able to just have a computer take over so I never have to see the stupid game again. Especially if the computer is sufficiently good to make it almost impossible to swindle. That way maybe the person will learn to behave.

I guess it comes down to this: I play go to have fun, not waste my time fighting over whether I win by more or less than 20 points. Of course, I might go on to lose one of those games because I'm no longer interested in the game and no longer paying attention. To anybody who wins a game like that: You just played so badly that you put your opponent to sleep; I hope you're proud.

Sorry for the long rant.

some people failed to notice that the creator of this discussion is proposing that in the 110th move a way to consider to cancel a game...

i know that in an opening is posible that someone have an advantage but in a 110th move in go the game is on the middlegame...many things can happen.

for me, the average go game finish like in the 230th to 265th move. so is stupid to consider (there are exceptions) to end a game to early.

What if you're 40 pts behind? There is no way you're going to catch up 40 points in yose.

and if the other player has a build a moyo, letting the other player have the 40 points?

Corrin Lakeland you are right, I agree with that it is not fair to continue 'lost' game. But it is one thing more you shoud consider, some go players here are just beginers like me :). They can't see so much as strong players and they could not realize that they have lost. Maybe you sholud tell your opponent that the game is over and he should resing instead of adding to a blacklist ??

I'm sorry, my english is not good but i'm understood i hope :)

i dont think he is right.

there is lil possibility that a yose fase is in the 110 move, like the author of this discussion propose.

i understand that a lead of 40 points in a yose fase is hard (depends on the position) to beat, but not in a 110 move.

and by the way, i have seen many games of takemiya masaki where he let his oponent to have great quantity of secure territory and he only have influence...no secure territory. where all depends on that.

all the depends on the position and not just basic counting

bump