How do you know when not to invade?

[Sneegurd]

It depends on a ruleset a bit (with chinese rules you don't lose points by playing inside your own territory, so this aspect is easier to grasp for beginners). Here black has to be careful not to end in seki, but generally if white plays in his territory, black can say (to himself) "no, you cannot kill me or make seki" and pass. If white continues, he can ignore until it threatens his life and then capture white. Notice that even then he doesn't need to physically capture every white stone, if they don't have chance to live / kill black. If white doesn't agree with him, they can play the situation out.

OK, but if I wouldn't agree as white, I mean if I insist on "you cannot kill me" (even if I know better) we have to play it out. Then black will kill white, but to achieve this, he needs to play inside his territory. And with japanese rules, he loses points then! Is that right?

[uPWarrior]

Black will also capture a number of white stones similar to the number of black stones he had to play. The score will remain the same.

[hermitek]

OK, but if I wouldn't agree as white, I mean if I insist on "you cannot kill me" (even if I know better) we have to play it out. Then black will kill white, but to achieve this, he needs to play inside his territory. And with japanese rules, he loses points then! Is that right?

http://senseis.xmp.net/?JapaneseRules :
Resolving disputes about life and death or protective plays depends upon hypothetical play with special ko rules.
hypothetical play = you play it out and then go back to original situation.

[jts]

It depends on a ruleset a bit (with chinese rules you don't lose points by playing inside your own territory, so this aspect is easier to grasp for beginners). Here black has to be careful not to end in seki, but generally if white plays in his territory, black can say (to himself) "no, you cannot kill me or make seki" and pass. If white continues, he can ignore until it threatens his life and then capture white. Notice that even then he doesn't need to physically capture every white stone, if they don't have chance to live / kill black. If white doesn't agree with him, they can play the situation out.

OK, but if I wouldn't agree as white, I mean if I insist on "you cannot kill me" (even if I know better) we have to play it out. Then black will kill white, but to achieve this, he needs to play inside his territory. And with japanese rules, he loses points then! Is that right?

This worries beginners a lot. I think it helps to understand the original rules of the game and what the "Japanese" rules (i.e. the rules developed during the Tang) and the "Chinese" rules (i.e., the rules developed during the Qing) are trying to accomplish.

Originally the goal of go was to get more stones of your color on the final board. The board was completely filled, except (almost certainly) for the two eye spaces required for each group to live, and the stones of each color were counted up. Of course, for this stone-filling game to work, you need to have a capture rule; otherwise, as the players alternate placing tiles, the first mover will always win by one. But once you have a capture rule, the goal of the game becomes to have the stones you place now protect spaces that your opponent can't play on without being captured (or risking capture). That way you can fill in as many of the open, unprotected spaces as you can, and then still have more protected spaces inside your group to fill in at the end.

To beginners and GnuGo, this way of playing was probably very reassuring. To people who have been playing for a while, though, it is both boring and a bit inelegant. Once all of the spaces than can be protected from the opponent's stones are protected, then you split the remaining points that can't be protected evenly down the middle (the person who takes the first worthless middle space might sometimes get one more point here than the person who takes the second worthless space); then you have to completely fill in your own territory with your own stones. Boring, right? What's going to happen is a foregone conclusion.

So sometime around the Tang period, go players introduced an innovation. They just stopped filling in the almost-worthless points in the middle that didn't protect any extra spaces, and didn't count them for either side; and they stopped filling in the protected spaces, and just counted them for the side that surrounded them. And, making, things even easier, instead of having your stones that are still on the board count towards your score, they had your stones that were captured or doomed to capture count against your score. It comes out to the same thing (with one small difference), and allows you to dump the captures inside your own group to easily calculate the score at the end. --- As you have already guessed by now, this is the form in which Go went to Japan and thence to the West. We call the protected areas that used to get filled in at the end of the game "territory" and the least-valuable points which didn't affect territory or the life of groups "dame".

But you can see that this change wasn't really intended to affect anything other than the boring, inelegant stone filling stage at the end of the game. People were not supposed to have disputes about which stones were alive and dead once there were no more moves left that effect territory/captures. Indeed, experienced players know that if the life and death of the stones was in question, the player who claims they are alive would be desperate to keep playing, so as to give them two distinct eyes or capture enemy stones. With very few exceptions, when a player is claiming that his opponent has to make more moves to capture, but he refuses to play any more stones on the board himself, he is acting in bad faith.

However, there are a few exceptions that are genuinely confusing even to experienced beginners, and in the Japanese system they came up with a special set of rules to resolve them.

The Chinese rules came later. They're very similar to the Japanese rules, except that they do fill in the dame and do not keep track of captures. So, like the Japanese rules, the Chinese rules call for counting empty territory that was originally filled in, but instead of having captures count against territory, they stones on the board added to territory. It's the same thing, remember - each player has usually played the same number of stones, +/-1, so stones on the board plus dead stones for black equals on the board plus dead for white. Because the stones on the board count again, you get an advantage from filling in all the worthless spaces in the middle of the board. What's true under Japanese rules is true under Chinese: if your opponent insists that you need to captures his stones at the end and isn't willing to play more stones on the board himself, he's probably acting in bad faith. But because under Chinese it doesn't matter anyway, you can just keep playing and adding more stones inside your own territory to make the opponent happy.

Summary: The original rules involved filling in the entire board. Everyone agrees that's boring. We now have two different rulesets that implement the idea of not filling in the whole board. The result should be the same (+/- 1 point) in almost every single game of go. A good move in Japanese rules is equally good under Chinese, and vice versa. Their only disagreement is what to do once all the territory is settled. In Japanese rules you stop, and resolve disputes about which groups are dead by not being an cad. (Hehe, cad gets filtered to cad.) In Chinese rules you split the dame and resolve disputes by plunking stones down on the board until you've filled in your entire territory, if that's what it takes to convince you're opponent to be a good sport.

[illluck]

Just a random point: I'm not sure about the timing of the Chinese and Japanese rules. I guess you can claim that Chinese rules as we know it come later, but it (area scoring) really is the fundamental counting method (with group tax) much earlier than the Japanese rules (territory scoring). In other words, if there were ever a time when people played until the board was filled, it ended very early.

I think territory were more used as a quick count alternative to area scoring and was passed to Japan where it became more popular than area scoring.

Of course, I'm not an expert (or even reasonably knowledgeable) on this, but it seems weird to claim that Chinese rules came after Japanese rules.

[jts]

Just a random point: I'm not sure about the timing of the Chinese and Japanese rules. I guess you can claim that Chinese rules as we know it come later, but it (area scoring) really is the fundamental counting method (with group tax) much earlier than the Japanese rules (territory scoring). In other words, if there were ever a time when people played until the board was filled, it ended very early.

I think territory were more used as a quick count alternative to area scoring and was passed to Japan where it became more popular than area scoring.

Of course, I'm not an expert (or even reasonably knowledgeable) on this, but it seems weird to claim that Chinese rules came after Japanese rules.

This is actually what Chen Yaoye claimed in the article John F. translated and linked to a while ago. The same claim was on Sensei's even earlier, presumably based on the same textual evidence. The order doesn't really matter, of course - what I'm trying to communicate is that both Japanese and Chinese rules are developments from the same game, with slightly different ideas about how to skip filling the board "to overflowing," as the ancient manuals put it, at the end.* The strategy is the same in both games.** There is no way for a player to change the result in territory scoring to something other than what it would have been in area scoring by petulantly insisting that a player fill in his own territory, because either set of rules has been specifically designed to make the result the same as it would in the parent rule set, minus the laborious filling in of the entire board.***

* Except for the group tax, but this doesn't matter to beginners who are still confused about the rule sets!
** Except for 1/3 point kos! Thanks, Robert! Doesn't matter to beginners!
*** I'm sure you can guess what I'm going to say about unremovable ko threats.

[Bill Spight]

It depends on a ruleset a bit (with chinese rules you don't lose points by playing inside your own territory, so this aspect is easier to grasp for beginners). Here black has to be careful not to end in seki, but generally if white plays in his territory, black can say (to himself) "no, you cannot kill me or make seki" and pass. If white continues, he can ignore until it threatens his life and then capture white. Notice that even then he doesn't need to physically capture every white stone, if they don't have chance to live / kill black. If white doesn't agree with him, they can play the situation out.

OK, but if I wouldn't agree as white, I mean if I insist on "you cannot kill me" (even if I know better) we have to play it out. Then black will kill white, but to achieve this, he needs to play inside his territory. And with japanese rules, he loses points then! Is that right?

Under Japanese rules you do not have to play it out. Pros certainly would not. (The pros have their own rules, but they know what is going on.) Beginners might want to play it out to make sure. There are a couple of ways to play it out and keep the score the same. One is to play it out with captured stones. If there are not enough captured stones, then the players trade an equal number of stones. Another way is for each player to make the same number of plays, and to give up a stone as a prisoner if they pass.

Another way to handle this situation is to play by Chinese rules or AGA rules, in which case the capture will obviously not cost anything.

[Bill Spight]

Just a random point: I'm not sure about the timing of the Chinese and Japanese rules. I guess you can claim that Chinese rules as we know it come later, but it (area scoring) really is the fundamental counting method (with group tax) much earlier than the Japanese rules (territory scoring). In other words, if there were ever a time when people played until the board was filled, it ended very early.

I think territory were more used as a quick count alternative to area scoring and was passed to Japan where it became more popular than area scoring.

Of course, I'm not an expert (or even reasonably knowledgeable) on this, but it seems weird to claim that Chinese rules came after Japanese rules.

The earliest known scored game records (in China) apparently use territory scoring with a group tax. The earliest known rules for scoring apparently use area scoring with a group tax. Which came first? That is a matter of speculation.

Some people think that the original game was simply no pass go (at least up until the board was filled with live stones). But that game would not be the same as area scoring with a group tax. To get something like that you have to let the players return a prisoner instead of making a play on the board. And that, dear friends, is a form of territory scoring. (In area scoring prisoners don't count. )

[illluck]

Wow, this is very interesting - I've always been under the impression that area scoring came first and territory scoring was developed later for ease of estimating score. Thanks a lot to jts and Bill for the correction!

[Bill Spight]

Wow, this is very interesting - I've always been under the impression that area scoring came first and territory scoring was developed later for ease of estimating score. Thanks a lot to jts and Bill for the correction!

If, and it's a big if, go was originally a no pass game, then I think that territory scoring came first, since prisoners would count. But the converse does not hold true. If go started out as a game of capture, so that prisoners counted, it is not clear that territory would count, too. For instance, if Black played last and filled a dame, the prisoner count would reflect territory, but if she filled a point of territory, it would not.

See this New in Go entry: http://www.gogod.co.uk/NewInGo/ChenZuyuan_2.htm . The DunHuang classic states, "stones more is winner”. Chen interprets that to mean that the player with more stones on the board wins, but it could mean that the player who has captured more stones wins. I am inclined to agree with Chen, but the two games are approximately the same. Some time ago John Fairbairn mentioned the possibility that both territory and area scoring coexisted in China and then the territory form died out. That is an attractive idea.

[jts]

Ah yes, clearly I meant Chen Zuyuan, not Chen Yaoye. Thank you for posting the link, Bill.