Whom belongs the territory to?

[skan]

Hello

I find not trivial to count the final area/territory score.

I know one can use different rules but all of them rely on the fact that you could recognize what areas belongs to each player. There are also some other difficulties, such as KOs...

I send you screenshot of a game against Fuego. GoGui shows me the territory that belongs to each player.



I understand that some areas are black or white because they are completly surrounded.
But, what happens with the red region (bottom-right)?
Shouldn't it be white?

I guess the reason is that if they continue playing white will die.
What happens with the area marked with green?. Why it's white?. I see white surrounds it but maybe black could win some territory if the game continues.

[Li Kao]

If you're area scoring it's simple: If you don't agree with your opponent on the status of some group you just continue playing, and the score won't change. With territory scoring it can be complicated since playing on can reduce the score of that player(playing inside ones own territory costs a point). So you need to agree on the status. Possibly by asking a stronger player for help.

In your examples the computer is right in both cases. On the left the black stones can't live no matter what black does. On the bottom the left white stones are in atari, and white can't prevent their capture. So they're dead, and the black stones are alive.

[Dusk Eagle]

I understand that some areas are black or white because they are completly surrounded.
But, what happens with the red region (bottom-right)?
Shouldn't it be white?

Stones 152 and 162 have only one liberty, whereas black 11 and 115 have two. This means that even if white plays first, no matter what he does black can just capture the two white stones. If stones are completely dead no matter who plays first, it is unnecessary to remove them from the board - they are simply marked as dead at the end of the game, and counted the same as captures.

What happens with the area marked with green?. Why it's white?. I see white surrounds it but maybe black could win some territory if the game continues.

That black group has zero eyes, and you need two to make your stones alive (or you can have a seki, but that's not happening here). Thus, since black passed, he has declared that he does not believe he can make two eyes even if he goes first (otherwise he wouldn't have passed). Since black is dead, his stones get removed from the board after both players pass and get counted as captures for white.

Notice that these same principles apply to Black 207 in the middle-right and and in the top right.

You may also want to check out http://senseis.xmp.net/?path=BeginnerEx ... rExercise1 and http://senseis.xmp.net/?path=BeginnerEx ... rExercise3 for a couple of good problems that will help you understand eyes better.

[skan]

All right, thanks.

What happens if the the territory is not completely surrounded?
Some time ago I was playing against the computer (low level, of course).
He resigned and when I checked the territories I noticed something strange, some territories were not completely surrounded by me but the computer stated that is was my territory. Why?

Maybe it thinks that if I can continue playing eventually I'll surround it.

[Li Kao]

You can only score a finished game. When a player resigns the game is usually unfinished, and thus the score can only be estimated. I assume the computer showed you such an estimate.

[RobertJasiek]

You can only score a finished game.

In principle, each position can be scored as if the game had ended. For typical territory scoring rules, the practical effort for determining the correct score can become arbitrarily great though. Hint for non-beginners, who imagine two passes ending the alternating-sequence:
http://home.snafu.de/jasiek/j2003.html

[Bill Spight]

In this diagram I have "scored" the game according to Ing's proximity scoring, assuming that all stones are alive. By proximity scoring an empty point belongs to the player with the closest stone by Manhattan distance (along the lines of the board). Doing so shows some interesting things.

First, the single Black stone on the right side commands no area. That does not prove that it is dead, but it is fairly obvious that it is.

The four White stones and two Black stones in the bottom right command no area, either. Again, none of this proves anything, but if one side can win the battle there it should be alive while the opponent's stones are dead. Inspection shows that two of the White stones are in atari. Black can win a fight, even if White plays first. The White stones are dead.

The two White stones in the top right command three points of area. However, even if White gets three moves in a row and completes the outline of the marked area, he has no eyes. White cannot make seki, and those stones are dead.

The Black stones in the top left command three points of area, and there are seven neutral points. However, even if Black gets to play on the edges of the Black area plus neutral points, as long as White prevents Black from capturing a stone, Black cannot form two eyes, and there is no seki. Those Black stones are dead.

With a little experience you will be able to distinguish dead stones from live stones with good accuracy.

[skan]

Hello

Bill Spight, I don't understand why you introduce the "proximity scoring" concept in order to explain it.

[Bill Spight]

Hello

Bill Spight, I don't understand why you introduce the "proximity scoring" concept in order to explain it.

As Robert Jasiek pointed out, you can score the game before it is over by assuming that all stones are alive. Here is the picture that you get if you do so by area scoring:



I think that the picture that you get by proximity scoring is more helpful in seeing which stones are alive or dead. Proximity scoring gives you information about who controls each point. Assuming that the boundaries of control are filled in is not unreasonable, if each opponent actually does control the points that would be scored as his. If you can't live even if you are allowed to do that, odds are that you are dead (late in the game).

[jts]

To put Bill's point in a different way (I think!) -

At the end of the game, beginners are often confused about dead stones that are still on the board. Couldn't they still make life?

People who have been playing for a while can just look and say "oh, they're dead. Game over." Less experienced players might not find it quite as obvious that the stones are dead, but they can look for eyes and say "oh, they can't make two eyes. They're dead. Game over." But a beginner might not even know where to start looking for the eyes!

One place to start is to remember that an eye is a point of territory. A point of territory is a point (or group of points) surrounded by friendly stones. If a point is touching an enemy stone it can't be territory (unless you capture that stone). If a point is closer to an enemy stone than to yours, it's unlikely that you'll be able to surround the point before the enemy plays next to it. So "proximity scoring" narrows down your options - it shows points that might conceivably become your territory.

"These are the points that might potentially become my territory - even if I wall them all in, I don't make two eyes - I can't make two eyes - the game is over." Bill is trying to use proximity scoring to help you figure out whether the groups can live at the most basic level.

[skan]

"These are the points that might potentially become my territory - even if I wall them all in, I don't make two eyes - I can't make two eyes - the game is over." Bill is trying to use proximity scoring to help you figure out whether the groups can live at the most basic level.

I see.

[RobertJasiek]

As Robert Jasiek pointed out, you can score the game before it is over by assuming that all stones are alive.

This implication for area scoring is a fraction of what I have said:) For territory scoring, life and death status needs to be determined.