I have a game, where I several times posted “Count score”, but my opponent always wants to count a different score.
What is the procedure at little golem to handle that kind of problem?

(see http://www.littlegolem.net/jsp/game/game.jsp?gid=1599252)

Well, I see you win even by his counting so you can just as well accept it. You will have a win and the same gain on rating.

If you realy want to make it straight you can continue playing. That's right, is possible to make a move after scoring, just click on the board. I suggest you capture something, making it clear for your weaker opponent.

I don't know if comments you type to the opponent are publicly visible; I don't see any comments, but did you try explaining to the opponent that those black stones are dead because they don't have 2 eyes and can be captured? Or did you both just keep silently keep changing the score status back and forth? :)

I tried to explain it by linking to Sensei's library (http://senseis.xmp.net/?TwoEyes), but he did not answer to that.

Russ: Comments in the game are not publicly visible.

Anyway, the best way to decide when is a dispute on life and death is to play it out. Many rule sets have this.

Playing it out can affect the score in the rule set we use here.

So maybe we should switch to chinese rules. It solve some other problems too like “bent four in the corner” or eternal ko/life. In other way it increases length of game because dame(neutral points at border) have to be played as well.

Dvd If “Playing it out can affect the score in the rule set we use here”, then this is does not look like a good rule set.

If the same number of black and white stones are played inside a territory the score doesn't change. Playing it out doesn't mean only one player has to play, it means normal, alternative moves.

It is true the Chinese rules solves this issue in a most elegant way. What I hate about them is way of counting at the end, destroying the position. But computers do not need this and they will count just fine.

The score will change if the defending player passes, which could affect the result if the score is close and several turns are required to prove that a group of stones is dead. Admittedly, this must be a very rare scenario.

I'm not sure I like the sound of the Chinese rules, the main impact would probably just be to make the games longer. I think the best solution would be to get a strong player to score the game in the event of a dispute, but I'm not sure how that would be implemented.

Isn't there any site where you can upload a SPF file and it will give current score pretty accurately?

If approximate score was displayed before each player passes, that would speed up games as most intermediate players are not sure they have counted well.

I agree that the rule set we use here is not a good choice for a turn-based site that includes players with varying speed, understanding of the rules, quality of play, etc. However, it is the only rule set available at this site and since all the arguments against it are old and Richard has shown no interest in changing it.

N. Pass
N+1. Pass
N+2. MyScore
N+3. HisScore
N+4. MyScore again
N+5. HisScore again
-> instantly game is frozen and submitted to an arbitration panel who decides the conclusion (which might include “play on, you're both wrong”).

@Caroll, while many computer programs calculate the score they also have a problem deciding what groups are alive and what are not. Usualy they ask the players to mark dead groups,just like LG. Moreover, displaying the score during the play will confuse and misdirect the players, because it will not be real,computers are not yet so good at GO (although in 5 to 10 years they will beat any human).