Some misconceptions about Go

Many Go players, even stronger ones, sometimes have some misconceptions about Go, specially when dealing with different rulesets than the ones they are used to. This page lists a few common ones (the list might grow larger with time, as I encounter more misconceptions).

I have written a separate page for some claims where my answers might be seen as disputed, a matter of opinion or even controversial.

Playing inside territory with area scoring rules

A common claim is that when using area scoring rules (eg. Chinese rules) you can play inside your own territory at no cost (territory point or own stone, it's one point in any case). Sometimes arguments pro area scoring say that this is good for beginners because they don't have to worry whether protecting a possibly weak point will cost them points or not.

This is highly misleading. The only case where playing inside your own territory does not cost points with area scoring rules is when the game has completely ended, that is, when there are no moves worth points left and all dames have been filled. If you play inside your territory while there are still valuable moves left on the board (including dames), it will cost you a point regardless of the ruleset used.

Usually the need for a teire (ie. a mandatory defensive endgame move inside your own territory) will happen before the last dame has been filled. It's extremely rare that the very last dame forces a teire.

Thus beginners will have to struggle with the problem of whether to defend or not regardless of the ruleset. It will cost a point.

(The reason why playing inside your own territory costs a point if there are still eg. dames left even when using area scoring is that the play inside your own territory does not expand your area. That is your area will be unmodified, while your opponent will then have another move to expand his area.)

Japanese rules and bent-four in the corner

A common claim is that the Japanese rules state that the so-called bent-four in the corner is dead.

This was true in the past. The 1989 Japanese rules used internationally nowadays do not say this at all (except perhaps in their commentary part, which is not really part of the ruleset).

What the 1989 Japanese rules do say is that at the confirmation stage (ie. if the status of the groups need to be demonstrated after the game has ended) passing is the only valid ko threat. That's it.

This rule has the consequence that the bent-four in the corner situation is (usually) dead. But that is only a consequence, not a rule.

In fact, there exist situations where the bent-four in the corner is not dead even under Japanese rules.

The pass-for-ko-threat might sound irrational at first, but it has actually quite valid reasons to exist: It brings Japanese rules closer to other rulesets which use area scoring. In the latter ko threats can usually be removed after the game has ended (with some exceptions) and then the bent-four group captured because there are no ko threats. Because with territory scoring it's not possible to explicitly remove all ko threats for free as in area scoring rules, the Japanese rules simply state that they are assumed to have been removed. This actually brings the Japanese rules closer to other rulesets, not farther as many seem to believe.

A remarkable difference between rulesets happens when there are unremovable ko threats on the board at the end of the game, but that's another story. See also the next item.

Bent-four in the corner with other rulesets

A very common claim is that in rulesets other than Japanese the bent-four in the corner situation has to be played out.

This is completely misleading and basically wrong. You don't have to play it out in every single case. More specifically, if there are no unremovable ko threats on the board at the end of the game (which is usually the case), the group is dead, period (regardless of the ruleset). It would be a complete waste of everyone's time to explicitly have to protect every single tiny ko threat on the board and then capture the corner group. The usual thing for the owner of the corner group to do is to agree that it is dead as it stands.

The only case where it may have to be played out explicitly is when there are unremovable ko threats at the end of the game (a seki is the most typical case). If that is the case then the invader may have to lose something in exchange for the corner group.

However, even in this case playing it is usually not forced in any ruleset: If the invader refuses to play it, it will usually be considered a seki (both the corner group and the invader will be alive).

So no, in most cases it does not have to be played out. In most cases the owner could just save everyone's time and agree that it is dead (when there are no unremovable ko threats). Also even if there are unremovable ko threats the invader could just leave it as a seki; he doesn't have to play it out if he doesn't want to. Only when there are unremovable ko threats and the invader has something to gain by playing it out, he may and should do so.

Triple ko

It is often said without too much thought that with Japanese rules triple-ko causes the game to be voided.

While what people are thinking when they say this is usually probably right, the sentence itself is a bit misleading. There are many things which are not accurate:

Firstly, one has to understand what a triple-ko is. A beginner might thing that it's enough to have three kos at the same time on the board, but that's not quite it. A triple-ko is a situation where there are three kos on the board so that each one serves as a big-enough ko-threat for the other two. This means that any of the kos can be taken as a ko-threat to any of the other two, and this threat is big enough that it cannot be ignored. There can perfectly be three or more kos on the board but all of them so small that they do not serve as ko-threats (this is actually rather common in the endgame) and it does not cause any problems. (In practice the kos in a triple-ko situation have to all be related to the same two groups fighting each other because if they are unrelated, one of the kos will be smaller or at most the same size as the other two and there's then no reason to answer it, that is, it's not a big-enough threat.)

Secondly, and relating to that, a triple-ko situation does not automatically result in game voiding as the usual claim seems to imply. There's no rule which says that. What causes the game voiding is when the same board position repeats itself indefinitely and neither player is willing to stop the loop. It is perfectly possible to have a triple-ko situation but one of the players to decide that he can tenuki, stopping the loop, and try to handle the loss (at least in theory). The game is not usually ended even when a situation repeats a few times, as long as it stops repeating.

Thirdly, it is possible for the loop to be longer than precisely three kos. There could be four or even more kos, all of them providing big-enough ko-threats for all the other kos. In that situation the term "triple-ko" is a bit limited.

And finally, and perhaps most importantly: Triple-ko is not the only endlessly-repeating board position which can cause game voiding. This is perhaps the most important thing which the claim is failing to express. There exist other situations, such as the so-called eternal life, where there are no kos involved and the same position repeats endlessly and causes game voiding if neither player is willing to stop it (it has happened in professional games). Other variants also exist.

It would be more accurate to say something along the lines of "if the same board position repeats indefinitely and neither player is willing to stop the loop, it causes the game to be voided".


Copyright 2005 Juha Nieminen


Valid HTML 4.0!