Go - Problem with current scoring of games Go forum

20 replies. Last post: 2021-02-23

Reply to this topic Return to forum

Go - Problem with current scoring of games
  • UnspeakableGamer at 2021-01-30

    Hello, I am currently involved in two 9x9 games of Go - and in both games it has scored them incorrectly.

    The first is game: #2211242

    I am black and have won the game, but it has just scored me as 0 points and says I have lost; despite there being no way for white's lone stone to capture any territory on my side before I can capture it.

    The second is game: #2211215

    In this game on move 47 I refused to pass and end the game because it was only scoring me as having 4 points (from the 2 tiny regions) despite white's stones being unable to capture my larger territory before I capture them.

    In the 1st game I have accepted the score - so will lose.

    In the 2nd game I played on to capture the stones - but think I will now lose too.

    Is there some way these games can be corrected? or at the very least the scoring system sorted out so that it scores future games correctly?

    It seems to be having trouble scoring these types of positions correctly.

  • Crelo at 2021-01-30

    After two passes you should mark dead stones by clicking on them.

  • UnspeakableGamer at 2021-01-30

    Perhaps that could be made more obvious then, because it was not clear to me that I had to do that (and I usually read things carefully).

  • Crelo at 2021-02-02

    True, it wouldn't hurt to have this information explicitly displayed. On the other hand is similar with other Go servers like KGS.

  • lazyplayer at 2021-02-03

    Hi all, can you explain me why not everyone is using area scoring? Is it due to “granularity” or it's just intellectual weakness? :)

  • lazyplayer at 2021-02-03

    I'm wondering how frequent are “sekis” in Go? If “sekis” are not rare then the granularity argument loses most if not all its value.

  • Crelo at 2021-02-10

    Even with area scoring you will still have to mark dead stones at the end. Territory scoring is older and because go was spread in the west by Japan mainly this is the default set of rules. There are countries like USA, UK, New Zealand, which are using area scoring rules, so I guess is more about how one learned the game than any other real advantage.

    Seki is rare, I would say under 5% of games, but I do not understand what is “granularity” in this context.

  • lazyplayer at 2021-02-10

    Crelo, it's explained here. Regarding default rules, yes, it's just tradition, as you've said. For me the fact that under Japanese rules you can't “play it out” is a big turn-off. We've this beautiful game with mathematically beautiful rules and then we can't play it out? It's just sad.

  • Crelo at 2021-02-10

    As I see it, playing it out is an issue only for complete beginners or for very high rank players. For beginners my advices is to always play in case of doubt. After all a game of go ends when the players agree there is nothing left to play. There is no other way to tell when is the end. Basically the moment the game ends depends on the skill of the players.

    Thinking that the score is modified because of playing it out happens only from the observer point of view, when the observer is more skilled than the players. If there is no skilled observer the players should sort it themselves, there is no 'correct' score. Why would be a mistake during the scoring phase more important than a mistake during the game? In my opinion they have equal weight, both change the score.

    These being said, if both players play alternatively and pass stones are given, playing it out will not change the score.

    As for the very high rank players, sometimes rules written to simplify the approach backfire, like the rule to always close a ko, I find these stories very entertaining. At their level the players can read it out so they don't need a special rule. For example I would like bent for in the corner to be played out even with territory rules, sometimes will remain as seki.

    I know there are a number of weird positions that have different outcomes depending on the scoring method, but those are oddities that appear very rarely, some of them are even artificially created, they are not very relevant in practice.

  • lazyplayer at 2021-02-10

    Crelo, it's not so simple because the fact that you lose points by playing useless stones in your own territory is a problem especially for beginners. I think territory scoring was invented to punish silly beginner endgame mistakes like useless play in your own territory but, as you've said, it has backfired by creating the problem that you can't play out the more unusual situations. And btw, even a beginner like me knows that If pass stones are given then it's area scoring in disguise plus a point for black in case white is the first player to pass. ;)

  • lazyplayer at 2021-02-10

    Actually, minus a point for black. Eheh. Well, hopefully Go community will one day agree on some international rules for everyone. I vote for New Zealand rules, let's have more fun.;)

  • lazyplayer at 2021-02-10

    Recently I've read about history of chess, and in chess there were disagreements on rules, for example, how to score a stalemate, but eventually it has been sorted out.

  • ypaul21 at 2021-02-10

    I think that this will be difficult because the largest Go community is in China/Korea/Japan, and they are quite happy with the current set of rules that they play with. There's nothing wrong with territory scoring in practice, even if you get weird side effects like how they deal with the bent four in the corner, etc. I definitely agree that it's much less beginner-friendly though.

  • Crelo at 2021-02-11

    lazyplayer, playing useless stones in your own territory is also a problem in area scoring. You lose the chance to make a more useful move, even taking a dame in yose. You are still losing at least one point.

    I know you are thinking about the counting phase, but the counting phase comes after the game ended - after two / three passes. Once the game ended the table is fixed regardless of the (missed) possibilities of the position. You can resume playing of course but then you are back in playing phase.

    For example it seems to me that recently the Koreans are filling all the dame even while they are using territory counting. Nobody is preventing players to play it out in case of unclear situations. It just happens that in most cases the players can read it out and just add necessary moves to defend. It is the same in both types of scoring.

    And territory counting was not invented, it is the original type of counting. The area scoring is the new one. Ancient Chinese games were using territory, they were even letting two eyes for each group, so in fact the more groups you had the less points you had.

  • Crelo at 2021-02-11

    Go is the surrounding game, territory, not the game where you try to put more stones on the board, area :)

    Besides I find very confusing putting the prisoners back into the bowl. How are the beginners counting during the game when they don't remember how many stones were captured?

    Anyway, the differences in the rules are subtle and they are relevant only in very rare, special cases. This should not stop anybody to enjoy the game.

  • lazyplayer at 2021-02-11

    Crelo, I'm very beginner and I don't count at all, I only take a look and I make a guess on who has a larger share of the board. Regarding the origin of the game, are you referring to " Stone scoring“? It seems to me stone scoring is closer to area than to territory. In fact I'm not even sure why we've abandoned stone counting, why not have a group tax? :)

  • lazyplayer at 2021-02-11

    Anyway I've to agree btw, probably group tax makes the game slightly less interesting because it forces players to play closer to their stones.

  • lazyplayer at 2021-02-11

    Frankly when I read sensei go and all the “corner cases” like “eternal life”, triple-ko, super-ko, ko during analysis under japanese rules, natural situational super-ko and so on, it's depressing. I think it's really like chess in a sense. It's a very pleasant game but when you look under microscope you see some “impracticalities” so to speak. In chess of course the 50-move rule is the ugliest feature of the game and yet it's absolutely necessary to ensure termination in a reasonable time. All the drawing opportunities are also very irrelevant for beginners but very relevant for top players. In summary, under the microscope, one can see that there are some definitive flaws.

  • lazyplayer at 2021-02-11

    I've found a great page on Go History. It seems that you are right that territory scoring plus group tax was the traditional way, But then they changed to stone scoring. Why this change? Probably because they ran into some disputes at high level. Eheh. When you want clarity of result you need to have a practical and well defined way to play it out.

    This is also moderately interesting: Compare Go to Chess/History, as well as this of course: Hex. For me in fact the similarities between Go and Hex are just very amusing.

  • Dvd Avins at 2021-02-23

    “And territory counting was not invented, it is the original type of counting. ”

    I doubt that, though I know it's true for as far back as we have records. I don't think anyone would have come up with it except that it is another representation of who controls more of the board. But it is easier to count on a physical board that's mostly filled in, so it makes sense that scoring would quickly evolve that way.

    As for hurting your score by playing in your own territory, it's also an issue when players disagree on the outcome and stones are not given for passing. Like, for instance, at Little Golem.

Return to forum

Reply to this topic