Which rule set do you think is better, Chinese or Japanese?


#68

Just curious, is there a scenario that the computer can not score correctly under a ruleset?

I am referring to two high level players who agree to pass to end the game instead of filling the board completely.


#69

Not sure what you mean by THE computer, but

our auto-score can sometimes get confused by a complicated seki (well sometimes by an uncomplicated life&death also, but let’s not dig into that :smiley: ) regardless of ruleset. I have seen Leela misscore games as well in the past, did not see a problem with Lizzie yet, but I did not play with her too much.

Bots (as far as I know) score almost exclusively in Chinese, so it is hard to name actual examples


#70

Chinese rules are better because they are conceptually simpler. Japanese rules have special cases for bent-four-in-corner and for a specific kind of repetition (ko).

My personal preference is for New Zealand rules, because they have no special cases.

I suspect that, historically, the ko rule arose from a misunderstanding of the rule that a player may not play a move that results in a repetition of the position (with the same player to move). Perhaps long ago, a Chinese person explained this rule (which also exists in Chinese Chess, by the way) to a Japanese beginner, and the Japanese asked how it is possible to repeat a position when every move adds a stone to the board; and the Chinese showed the simplest example, a ko. The Japanese then mistakenly assumed that the ko situation was the rule, instead of merely one example of the rule’s operation; and this mistake became enshrined in official Japanese rules.

I also like playing with integer komi. If both players played perfectly, a draw would be the appropriate result, so I think the rules should allow that.


#71

I think that the scoring algorithm in OGS always gives the correct answer after it has been told the definitely life/death status of the groups.

It certainly cannot determine these correctly for itself in all cases, otherwise we would not have the stone removal phase. And this is the reason why the estimator is an estimator.

GaJ


#72

“I also like playing with integer komi. If both players played perfectly, a draw would be the appropriate result, so I think the rules should allow that.”

Strongly agree.


#73

I also agree. I’ve had a couple posts in the Forum about the modern prejudice against draws in competition, which I believe stems primarily from the commercialization of sports.


#74

And it’s not like a draw result is “wasted” (as annul is) because a draw is still a rateable result (right?).


#75

Unfortunately, it seems that OGS does not seem to recognize draws


#76

I don’t think that this is what happened.

First, in the practical application of both the modern Japanese and Chinese rules, some long cycles like triple ko result in annulled games. The official rules text of the Chinese rules seems to say that the rule is just superko, but apparently in many tournaments, games will just be annulled if certain types of long cycles develop and neither player is willing to relent.

Second, I think it’s unfair and frankly insulting to suggest that the practice was invented by the Japanese as a result of ignorance. While the Japanese rules are complex and imply some special rulings that some might dislike, I think their design has been deliberate and carefully considered in order to specifically realize those rulings for various cases influenced by historic precedent.


#77

The issue has been on GitHub for a long time. However, it is known that at least technically, draws are possible on OGS (by moderator decision): Thoughts Regarding Komi Tie-Breaker

That linked game was a ranked game too, so Glicko2 as implemented here seems to be able to handle draws in one way or another.

It would still be really nice to get some word from @anoek on this…


#78

I am trying to understand, but I cannot wrap my brain around this yet. Let me try:

Chinese, stone = +1 point.
Japanese, stone = 0 point.

Chinese, captured stone = 0 point (for either player, prisoners not counted during scoring)
Japanese, captured stone = -1 point (prisoners count as 1 point for opponent during scoring)

Chinese, dead stone = +1 point (stones on the board always count as 1 point)
Japanese, dead stone = -1 point (opponent gains 1 point during scoring)

I understand you are telling me that mathematically, both rulesets, when compared, balance out during scoring in an equal fashion. However, I still maintain that despite them being equal, how each ruleset calculates points differently can affect the way the game is perceived by the players, how they prioritize and respond to different situations, and, at times, the strategies and tactics they use.

Depending on the board setup and where you at in the progress of the game, playing in certain ways in one ruleset, would not be wise or might be too risky in another. The two easiest examples are territory re-enforcement and invasion attempts In high risk (to you) situations. Do they balance out mathematically? Yes. Can they affect the way I play or what actions I may/may-not take? Yes.

I do not agree that the result is the same, on account of human perception. Mathematically, they are the same. However, if a person is not versed in how the two rulesets are equal mathematically, then cosmetic difference literally affects the player’s understanding of what is possible, what is risky, what is wise, and in turn affects what they ultimately consider to be viable moves. The aesthetics they perceive and interpret with their eyes affects their focus, and their focus, in turn, determines their reality.

In this way, cosmetics likely affect the balance of reasoning in any player who is unaware of how the rules mathematically balance. Judging by how many of the Chinese vs Japanese threads there are throughout the internet, I’d say this is actually a significant number of Go players. If you shop around, a lot of these threads include high ranked players, who possess great skill but lack an understanding of how the rulesets mathematically balance out. However, I still maintain the belief that despite them balancing mathematically, how each ruleset handles scoring does in fact affect how games feel to players and how they behave.

I hear what you are saying. I simply really appreciate how simple it is to handle under Chinese and how one never has to fear paying for an attempt to invade inside enemy territory. The method you describe for Japanese play makes it possible, yes, but it also makes gameplay strategy more complicated. I need to make sure I behave in a certain manner, handle my board moves in a specific order, and that I’m on top of all the potential dame positions (the bigger the board the harder this becomes). It also requires that I fill in all the dame first, before making my invasion attempt, which does limit my strategy options in some cases.

Sometimes you have your opponent running or feeling the pressure and you want to invade right now, striking while the mental stress your opponent is experiencing is at its greatest. If my attempt fails while they chose to fill in dame’s I was unable to get points for, then yes, it is a bad move. But I have the option to take this chance and my score will not be affected by as many points lost, since dead stones do not count during the scoring phase.

I like how I don’t even have to think twice about changing the way I play the game when I’m playing with Chinese. Chinese has a more natural feel to it. And by natural, I mean that I don’t have to add new layers of complexity into my strategy or increase the amount of information that I’m tracking on the board. This frees up brain power to concentrate on other aspects of the match, which I feel leads to less stress and better performance.

For example: I’m in a match that is seemingly entering the Endgame phase and my opponent begins filling dame’s while I decide to instead prioritize something else that potentially rewards me with greater returns or a stronger advantage; i.e. capturing stones, invasion attempt, creating eyes, resolving/baiting them with a Ko threat, or denying strategic/tactical gains they have temporarily overlooked. My point is, that there are actions you can take on a board that award more than the simple exchange of points. These actions can lead to greater gains down the road, providing more than the dame’s alone could have provided me.

Again, this is all based on my personal feelings, experience, and what I’ve learned by examining other players arguments in different Japanese vs Chinese threads. I acknowledge that other folks may not think twice about the points I’m mentioning here, but this is how I have come to see things. Please note that I would love to come to a point where I understood Chinese and Japanese to be truly identical, so I do strive to be objective and unbiased. So, if you have more to say, please do share it with me.

Yes, the bent four really makes my blood boil, as I have experience encountering it when it could have made a difference. As to other rules… my previous “kitchen sink” description was poorly composed. Generally, the term kitchen sink refers to an approach where every possible outcome has been accounted for, or an attempt has been made to accomplish that goal. What I intended to infer with that remark, is that the rules are worded in such a way that there are holes in their coverage. Which can create situations in a match that fall outside the purview of the rulings themselves. In turn requiring adjudication or interpretation by players to resolve, which should never be possible in a codified ruleset, especially considering that it is used for tournament level play.

I have been reading some older Go books and they speak of issues that would have existed during the author’s time (Japanese rules prior to 1949, after 1949 but before 1977, and between 1977 to 1988; examples: short, extended). This knowledge has colored my understanding and interpretation of Japanese rules a bit. Making my knowledge of them outdated and some of the problems I have harbored towards them superfluous. Specifically, because those problems have since been corrected. Please note that I only recently learned of the 1989 rules revision.

Since finding the 1989 iteration, I’ve been doing some searching to see, if what seems like crazy wording to me, makes sense to those who try to truly decipher the possible loopholes the wording in the 1989 rules might inadvertently create. The most complete analysis I have found so far is Robert Jasiek’s Commentary on the Japanese 1989 Rules. Jasiek’s analyzation of the rules is impressively thorough and he had this to say in his conclusion at the end of his commentary:

There is only one reason in favour of the Japanese 1989 Rules: They are the valid ruleset for the Japanese go associations. Reasons against the rules have been discussed elsewhere in great detail, so here is just a summary: Countless easy / easier alternatives of high design quality are available and some of them are even very close to the Japanese 1989 Rules’ strategic consequences. The rules contain many mistakes and gaps. They are too difficult for almost all players. The rules are inapplicable as rules because a correct application would demand arbitrarily much time for its execution.

Considering all the other better worded rulesets that are currently available, especially those that make it easier for a complete newcomer to Go to comprehend coherently, I cannot understand why Japanese rules are as popular as they are. That is, I understand how they historically rose to popularity and gained the following that they now hold. But I cannot understand, in spite of the competing rulesets available, how they continue to remain in such high standing. I do not say this in an insulting way. I simply find it curious =D.

Could you please explain your thoughts on this further? Why is it not a good thing? When you refer to the repercussions of playing nonsensical things under Chinese, are you saying that Japanese scoring punishes you for such behavior? I see many people speak to this perception, but the conversations are always old and over by the time I find them. I’d love to see someone elaborate on their personal thoughts on this particular topic.

How can you cheat in Chinese =/ ?

I felt this way when I first started to play Go. But after I became a little more skilled I realized that the player who plays Black begins with Sente, and that is worth something. On a 19x19 board this probably doesn’t matter as much, but on a 9x9 it can definitely influence how the game begins to unfold. Not a bunch, but at dan level play, every teeny tiny advantage could be a game changer. I’m curious how other folks here feel about the idea of a draw. I get super frustrated when I lose a game by 0.5 and feel bad for my opponent when I win by 0.5. So a part of me still sees the validity in allowing draws.

I recently learned about New Zealand rules. I really like the option of suicide. I don’t think it adds a lot to the game, but I don’t see a valid reason why it should be disallowed. Plus I’m a sucker for having Go rules be as simple as possible. New Zealand gets almost no love in the Go world though. At least, it is rare that I see mention of it. So I am hesitant to start using it. Is there anything further you could say, comparing Chinese to New Zealand? I’d love to hear what you think, since it is your favorite ruleset.

How can a draw be rated? An example: Instead of awarding the winner with 1 point and the loser with 0 points, both are awarded 0.5 points? I’m not sure how you could factor draws into a kyu rating system. I figure it is done mathematically. If both players are the same player rating, I’m not sure how a draw between two players lets you know anything useful. But if they were unequal in player rating, it should reflect well on the player with the lower rating, since they held their own against an opponent with greater skill. But if you follow that logic, it means you could go down in player rank by getting a draw, if you happened to be the higher skill player who reached a draw against a lower skilled player. That just seems weird. Maybe that is how you could handle it though.

@yebellz
There isn’t anything in particular I wanted to respond to, that you said. But I really enjoyed the way that you spoke about your viewpoints. You wrote eloquently and I really enjoyed reading what you had to say. Thank you for taking the time to post so many thorough responses.


#79

Except you’re wrong in thinking it makes a difference. I don’t use the term “mathematically” as meaning anything else than that it literally gives a completely equal score at the end of the game. It logically cannot be different. It’s provably impossible for them to be different. Therefore, there is no higher risk or lower risk with one of the rules compared to the other, since using the same strategy yields the same score. If you choose a different strategy, because you think it gives you an advantage, then it will also give that advantage in the other ruleset.

The only exception to this is the last dame point, passing or points in seki, none of which are part of typical early or midgame strategy, and seldomly part of endgame.

So can it affect the way you play? Yes. But does it affect the score? No. So if we assume one of the strategies is “stronger” than the other, then you shouldn’t let it affect the way you play, since it will make you weaker than you could be in one of the (equivalent) rulesets. It could be that perceptually you are inclined to play in a certain way, but this is nothing more than a handicap to yourself.

If you let your feeling guide you, you will be misguided.

Making a bad move that seemed to have potential is still an example of a bad move: had you read the position out beforehand and noticed it was not going to fly, then you wouldn’t have played it. It seems like you don’t lose points in the Chinese rulesets, since you felt like the potential was there, but this is just make belief. If you were a stronger player, you would have concluded there was no potential at all, and that the move was bad, and should have been played elsewhere.

The concept of “feeling the pressure” will shift as you get stronger. Weak players will protect against any threat, since they feel constantly pressured, while strong players will read out situations and be confident that they are not threatened. This has nothing to do with which ruleset is used, though. Stones get captured in the same way, territory gets surrounded in the same way, and in the end the score is the same.

Also note that dead stones do count: They count against you by the opponent’s live stones that he played when ignoring your moves while you played those dead stones. Each of your stones that your opponent ignores gives him a point (in both rulesets). If he has to answer, it doesn’t change the score (in both rulesets). If there are unfilled dame, then your opponent can gain points by filling them in Chinese rules, or he could gain points by passing in Japanese rules, so even in the latest stages of endgame the rulesets don’t differ.


#81

Doesn’t this mean that we would be well advised to look at each position both from Chinese and Japanese scoring perspective, and chose the best move only after assessing the risk that each viewpoint flushes out?


#82

That seems like a sound way to eliminate the psychological bias. Question remains to judge which of the two options is the best move, of course.


#83

My personal favourite is AGA rules because it’s concise and easy to apply.

Trump-Taylor and New Zealand are even simpler, which I think is better from and abstract / mathematical perspective. But, it seems to me (not from experience) that the differences between these and the more-usual rulesets would happen reasonably often in practice - most notably, some reasonably common positions such as https://online-go.com/puzzle/10604 have a ko threat in these rule sets which they wouldn’t have in Chinese or Japanese. That is, the difference between rulesets that allow or forbid suicide seems to be a practical one, while the differences between Chinese, AGA and Japanese are more theorical (i.e. two players who forgot to specify which ruleset they use are still unlikely to have any disagreements over the course of the game)

Isn’t that like saying that the Japanese rules complexity can be excused just because most of the time you can pretend you’d been using a different ruleset ? Which is true but just highlights the issue with Japanese rules IMO :slight_smile: