If you have large areas like here in the white upper left corner,

Enter image description here

why can't black just put a stone in upper left corner? Then white would lose one point. Or white would have to capture it and get a point instead, but then there would be three white stones within the area and that would mean again for white to lose three points.

So is it just convention to end the game here, when black won't be able to create a living group within the area?

  • 1
    If black chooses to play anywhere in the upper left corner, white only needs to play the 1-2 point on the left, or answer 1-2 with 2-2. That will destroy any possibility for black to make two eyes and live. Commented Dec 30, 2020 at 22:14
  • Some beginners may think that if black plays a single stone into that empty corner, white would be obliged to capture it. That is not the rule. Consider the similar situation on the right side. Black has no obligation to capture that single white stone. It has no chance of survival. In that open top left corner white would defend only when they judge that black is actually threatening to make a living group (assuming Japanese rules). Commented Mar 2, 2021 at 12:28
  • This question is a great example of why Area Scoring is superior to Territory Scoring
    – Zags
    Commented Jun 29, 2021 at 17:19

6 Answers 6


Suppose there is a black stone in the corner but it is impossible for black to survive, no matter how well played.

By convention, both sides recognize this and the black stone is counted as a prisoner.

If either side doesn't accept the convention, both sides will alternately play until the situation is resolved (in favour of white).

At that point, black and white will have played the same number of stones each. White loses a point for each one it played (loss of territory), and white gains a point for each one that black played (taken as prisoner). There is no net change in the score as compared with accepting the initial situation and neither side playing any more.

Note that the above is based on territory scoring rules, where points are given only for unoccupied territory (and prisoners). Under area rules the non-prisoner stones on the board are also counted as part of the territory, so any stones played by white into its own territory would not cost white anything.

  • At that point, black and white will have played the same number of stones each. Does this mean Black cannot pass and let White play only? Also, can the two players disagree on whether a stone played by Black is within the "dead zone"?
    – Allure
    Commented Dec 11, 2020 at 19:10
  • 4
    @Allure, given a large area surrounded by white, with possibly a small number of black stones in it, the obvious default is that it all belongs to white unless black can do something to protect its few stones. It's up to black to decide whether to concede the territory or to play the next stone. Once black passes, white would pass too, with both of them understanding that it all belongs to white. Commented Dec 11, 2020 at 19:41

This depends on the rules used.

  • In rules using area scoring (Chinese style), a player's points include both the empty intersections they surround, and their stones. In this case, any extra moves made inside a player's own territory do not cost points, unless there are dame left, since the move could have been used to take one and with area scoring that would be worth points.

  • In some rules (AGA), a player passing must give their opponent a stone (a pass stone) which counts as a captured stone, and is worth one point if territory scoring is used. Therefore, for every move White makes to capture the Black stones inside White's territory, Black must also make a move (either inside their own territory, losing a point; or inside White's territory, eventually to be captured for a point) or pass (losing a point).

    Note that AGA rules also allow for area scoring, in which case pass stones are not necessary.


  • In traditional Japanese rules (which use territory scoring causing extra moves inside a player's own territory to cost points) there is a special confirmation phase after the game has (provisionally) ended. During this confirmation phase, a particular situation can be played out to determine if the contested stones are alive or dead, but the confirmation play is taken to take place "virtually", and does not actually modify the board. The phase of confirmation play also has special ko rules with somewhat odd results, like the infamous case of a bent-four in a corner being dead as a rule.



If Black plays in that area, White could ignore it or pass and gain a point - the black stone would be agreed as dead, and White still gets the point it's sitting on. Black can only force White to surround and capture the stone if they're threatening to capture some of White's stones - for example if they cut off part of the wall, and it creates a capturing race. If Black keeps playing in that area, White has to decide when to start answering - if they wait too long, Black will make a living group.

  • 2
    I'll note that this is specific to Japanese scoring. In other scoring systems, spurious moves in your own territory do not affect the final score. In high level play the distinction is typically moot (i.e., high-level players typically don't make spurious moves in their own territory). That said, it did lead to one game being judged as "White won by 1 or 2 points" because the judge wasn't sure how to score it.
    – Brian
    Commented Dec 11, 2020 at 15:32
  • "If Black plays in that area, White could ignore it or pass and gain a point - the black stone would be agreed as dead, and White still gets the point it's sitting on." This statement is confusing. If Black plays in that area, it's most likely because Black doesn't agree that their stone is dead.
    – Stef
    Commented Dec 11, 2020 at 18:44
  • @Stef: Not at all. All such Black stones without two eyes (or alive in seki) are presumed dead, regardless of location on the board, It is Black's responsibility to play so as to make that clear - or abandon the stones. Commented Dec 11, 2020 at 21:52
  • 1
    @ForgetIwaseverhere: Whether stones are presumed dead or must be proven dead also depends on the ruleset. As ikkachu mentions, "bent four in the corner" is a pathological example of this: the choice of ruleset may impact whether it is considered live or dead.
    – Brian
    Commented Dec 12, 2020 at 0:20
  • Dear @ForgetIwaseverhere, could you please include some reference for that claim? I have never encountered a ruleset where a player had to "prove that their stones were alive". It's always the responsibility of the player who wants them dead to prove that they are dead. In the 1949 Japanese rules, stones used to be declared dead or alive "by shape", in which case it was no player's responsibility, but this is no longer the case in the 1989 Japanese rules, and has never been the case in other rules.
    – Stef
    Commented Dec 12, 2020 at 11:28

Your final line has the correct question more or less answered. It is convention to end the game because black would not be able to create a living group. What is important is that both players agree to this. A lesser-skilled player (black) may not be convinced and request to play it out.

As the board sits, without visible captured (removed) stones, white wins by 0.5 komi (or higher) and would not want to play one extra stone. TimK's answer about waiting too long would apply and could turn the result.

It's extremely challenging (or impossible) to create life in a small corner as presented, hence the reference to the lesser-skilled player.


why can't black just put a stone in upper left corner? Then white would lose one point.

It does not work this way. In territory scoring, the captured stone counts for a point (which offsets the point for a stone played in one's own territory), and then the intersection is freed up (undoing the supposed damage to territory of placing it). In area scoring (which I strongly recommend for beginners - it's also faster to count on 9x9) it's simply obvious: the borders of the area haven't changed, therefore the covered area of the group hasn't changed.

Territory scoring (and counting) is in essence used as a shortcut. The idea is that territory represents points where you could later put a stone at your leisure and count it towards your area, but your opponent cannot - because such a stone is doomed to eventual capture. These facts are inherent to the definition of territory; if you can't put the stone there, or your opponent can, it is not your territory. (Notwithstanding, of course, the last two open spaces required to prevent a group from dying to self-atari. Some historical variants of go did not count points for that; this is called a "group tax". And yes, board positions involving a seki do make this more complicated, but not irreconcilable.)

So, in territory counting, you count a point for each intersection of your territory, because it represents potential area that can't be taken away from you. (In area counting, the players just fill it in during the counting phase.) You also count a point for each capture, because it represents area that your opponent had, but you then removed. (If you kill a group, for example, that changes the points involved from "will be covered by the opponent at the end" to "will be covered by you at the end"; so it effectively counts double.) By doing this, you get essentially the same result as with the area rules.

In some rare situations it can cause a point or so of difference in the overall score, but it does not create the kind of difference that beginners commonly imagine. AGA rules, with the use of "pass stones", are designed to ensure that territory counting gives the same result that the area scoring rule dictates.

Or white would have to capture it and get a point instead, but then there would be three white stones within the area and that would mean again for white to lose three points.

It does not work this way either. The idea is that the players are taking turns; while the opponent is capturing your stone, you would be filling your own territory. Or, more shrewdly, passing - but that is exactly why the AGA rules use pass stones: so that it costs you a point either way.

In beginner games, it's common that the players aren't sure whether some invasion could live. Please feel free to play it out. As you and your opponents get stronger, increasingly more complex situations will become increasingly more clear. You can worry about whether you are insulting your opponents by trying nonsense, when you reach a level where you know (and expect your opponents to know) that it's nonsense.

In some rare situations called semedori, it can happen that playing inside opponent's territory (especially at a cutting point) can end up being worth a point (or more) - because there is some tactical threat that eventually requires multiple plays inside the territory (while opponent applies pressure from outside). For this to make a difference, the initial play needs to force at least three eventual inside moves: one to balance out the thrown-in stone that will be captured, one to balance out the move that was already owed to protect against the semedori threat, and a third to make net profit above that breaking-even.


Simple answer:

If black did play in the upper-left, white does not necessarily have to capture that piece, as it's dead and captures no territory. So really, we are trying to capture as much territory, but that in-turn also means we don't waste pieces.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .