Life In 19x19

I know that territory and area scoring might differ by one point (if there is only one pass for both players).

My question is whether AGA counting always provides the same result (difference in points between the players) as one of these two ways of counting or not (with the precondition that there is only one pass for both players in non-AGA games).

The question was a bit hard to formulate but I hope you can understand it.

AGA counting is the same as area scoring. Area and territory scoring differ when black and white play an unequal number of moves on the board, either because black makes the last play, or because one player passes and the other continues to play. The pass stones in AGA counting equalize those differences, so that players can use the territory scoring method, but get the area scoring result.

Follow-up questions -- Komi is 0.0 --

Click Here To Show Diagram Code

[go]$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------[/go]

Continued:

Click Here To Show Diagram Code

[go]$$Bm11 pass. pass. pass.
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------[/go]

Continued, showing passed stones by AGA rules:

Click Here To Show Diagram Code

[go]$$B ,
$$--------------
$$ | . X X O . |
$$ | . X X O . |
$$ | . X X O . |
$$ | . X O O . |
$$ | . X O O . |
$$--------------[/go]

Continued, with passed stones returned onto the board:

Click Here To Show Diagram Code

[go]$$B
$$--------------
$$ | B X X O W |
$$ | . X X O W |
$$ | . X X O . |
$$ | . X O O . |
$$ | . X O O . |
$$--------------[/go]

Under AGA rules, for the above game, does B win by 1 point ?
If not, then what is the correct result under AGA rules ?

If I've made any mistakes in the above sequences, please point them out --
I don't fully understand AGA scoring, so I'm asking these questions. Thanks.

See also Posts 67, 68, etc. here -- AGA Rules vs Japanese

In that situation black wins by 1 point, yes. AGA pass stones become prisoners, so you are correct to return them to the board as you would prisoners during territory scoring.

The same score would occur under Chinese area scoring rules, but the game would be a tie under Japanese rules. AGA rules effectively use area scoring, but you can count it like territory scoring thanks to the pass stones.

I have heard that there are some unique seki situations that can lead to different scores based on how you count. Seki where one group has an eye and the other group does not - but such situations are rare. I believe most rulesets just say that the eye doesn't count as any points, but if the eye does count for points when using territory scoring the result could be different.

A related question, where do the different scoring methods originate from / why did there ever end being two quite different ways of scoring being used? I ask, because after I got used to Japanese rules and later looking into AGA/area scoring, they seem unnecessary obscure. I mean, 3 (or sometimes more?) passes needed to end a game and some weird pass stones when you could just simply count your territory (thus possibly ending up with the above situation, which I can only see as a tie). Note: I'm only describing how I feel about area scoring from my perspective (as both system obviously work), not trying to start an argument about which is better

I always thought that AGA rules were contrived. I have never played under them: at least, I have not played a game where playing under them made any difference and I have never passed a stone. I seems to me the height of arrogance for an upstart neophyte go playing country like the USA to try to tell the rest of the world that they have it wrong. Japanese rules have worked since before the USA was even populated by westerners, let alone be a go asssociation.

People who want to apply western standards to the rest of the world have a problem. But then, they are only following their government, I guess.

walpurgis wrote:they seem unnecessary obscure. I mean, 3 (or sometimes more?) passes needed to end a game and some weird pass stones when you could just simply count your territory (thus possibly ending up with the above situation, which I can only see as a tie).

My perception would be that Japanese rules originally grew somewhat organically, but to the point that when you step back they're actually a bit rubbish. I'll stress, that's not in the sense that they don't work under normal play or anything like that, but they end up with weird arbitrary stuff about unusual situations that to my eyes is far more 'obscure' than the AGA rules way of solving it in one fell swoop with a single extra pass.

That is, Japanese rules mostly follow very naturally from the basic rules of the games. Except that eyes in seki aren't points for some reason. And bent four in the corner is just dead. And if you get a triple ko or some other wacky situation, the game is...void? Are these not equally 'obscure' and arbitrary as having an extra pass sometimes? Some of these may be out of date, I don't know the current state of codified Japanese rules, but this would be the kind of thing I expect in what I've heard called 'verbal' (or informal) Japanese rules which is what I've ever functionally actually played under. In contrast, merely having pass stones and picking a superko rule at least gives all of these situations a clear result that can be played out without penalty. Recognising that result may of course be difficult in an actual game (the old 'who repeated the board position first?' question) - I'm not claiming AGA rules are flawless (not that any ruleset is in this sense), but that can be resolved in other ways without introducing any more problems than Japanese rules already have.

On that side, I like that AGA rules neatly solve everything in a very well defined way that can easily be played out in a real game, with the only visible change that players really notice being the (apparently far more controversial than I expected) potential extra pass.

In the end, all of the debate is about stuff that rarely happens anyway. When it does happen, I think the Japanese way of handling it is understandable but has no particular merit other than it being a historic choice. I'm happy for people to consider that important, but I don't really understand how it can be considered natural or not 'obscure' compared to the AGA resolution.

when you could just simply count your territory

To condense my reply to address this particular bit, the problem is when the 'simply' bit isn't really the case. Bent four? Triple ko? Plus, the word 'simple' is only appropriate if you consider Japanese counting a foregone conclusion, which is itself an arbitrary choice that only makes sense because you happen to already have Japanese as the default in your head. I might equally say the way to count is to 'simply' count the area of the board controlled...oh, there's a spare space to control and it's black's turn, he plays there and the result is clearly a black win by 1 point.

In the end I don't really care, ultimately either set of rules produces the same result in the vast majority of situations, and clearly you can go hundreds of years without even defining things very well. But I personally think AGA rules are a very neat way to naturally and consistently resolve many problems at once when it comes to stuff that doesn't crop up very often, at essentially no cost to the players except the extra pass stones.

walpurgis wrote:A related question, where do the different scoring methods originate from / why did there ever end being two quite different ways of scoring being used? I ask, because after I got used to Japanese rules and later looking into AGA/area scoring, they seem unnecessary obscure. I mean, 3 (or sometimes more?) passes needed to end a game and some weird pass stones when you could just simply count your territory (thus possibly ending up with the above situation, which I can only see as a tie). Note: I'm only describing how I feel about area scoring from my perspective (as both system obviously work), not trying to start an argument about which is better

DrStraw wrote:I always thought that AGA rules were contrived. I have never played under them: at least, I have not played a game where playing under them made any difference and I have never passed a stone. I seems to me the height of arrogance for an upstart neophyte go playing country like the USA to try to tell the rest of the world that they have it wrong. Japanese rules have worked since before the USA was even populated by westerners, let alone be a go asssociation.

People who want to apply western standards to the rest of the world have a problem. But then, they are only following their government, I guess.

Historically, AGA rules derive from the Taiwan 1974 rules. They were, OC, Chinese rules, but with a superko rule and suicide (except suicide of a single stone). Ing later abandoned superko for his fighting-disturbing ko distinction. AGA did not include suicide and adopted a situational superko rule. The idea that the AGA was "telling the rest of the world that they had it wrong" is ridiculous. IMHO, in a country where beginners have no one else to play but beginners, Chinese style area scoring makes a lot of sense. It is hard to learn Japanese/Korean style territory scoring, which allows dead stones to be removed at the end of play without capture, unless one player is experienced enough to tell which stones are dead.

I wrote a brief article for the AGA journal about the Taiwan rules. Most players in the US played by Japanese style rules, so I suggested pass stones in order to make the number of stones for each player the same during counting, so that players could count territory to determine the score. (I was not the first to think of that. ) I was not directly involved with the later adoption of the AGA rules, but I suppose that the contrivance of pass stones can be traced to me, historically.

As for area scoring vs. territory scoring, both have roots in China. The oldest extant game records for which we have scores apparently used territory scoring, but the oldest known description of go scoring is apparently about area scoring. Both forms do not count the empty points required for life. (I. e., they have a group tax.)

Before the 20th century go rules were informal and unwritten. On rare occasions disputes occurred. It is important to note that all written rule sets are modern. As rules became formalized and written down, decisions had to be made concerning certain rare or contrived positions. Different rules making bodies made different decisions. There is nothing unusual about that.

Which is better, area scoring or territory scoring? Moi, I think that Button Go ( http://senseis.xmp.net/?ButtonGo ) is the wave of the future.

Hi Bill,

Glad you could join us in this thread. I'm still confused, sorry. Specifically, about:

go-master wrote:I know that territory and area scoring might differ by one point (if there is only one pass for both players).

emeraldemon wrote:Area and territory scoring differ when...

moyoaji wrote:In that situation black wins by 1 point, yes.
The same score would occur under Chinese area scoring rules, but the game would be a tie under Japanese rules.

Could we return to the above example:

Click Here To Show Diagram Code

[go]$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------[/go]

Click Here To Show Diagram Code

[go]$$Bm11 komi = 0.0
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------[/go]

Under Chinese area scoring, does B win by 1 point ?
Under Japanese territory scoring, is it a tie (5 points vs. 5 points) ?

EdLee wrote:Could we return to the above example:

$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------

Click Here To Show Diagram Code

[go]$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------[/go]

$$Bm11 komi = 0.0
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------

Click Here To Show Diagram Code

[go]$$Bm11 komi = 0.0
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------[/go]

Under Chinese area scoring, does B win by 1 point ?
Under Japanese territory scoring, is it a tie (5 points vs. 5 points) ?

Yes, and yes.

Bill Spight wrote:Yes, and yes.

Bill, is there a reference somewhere to explain why B wins by 1 point under Chinese area scoring ?

EdLee wrote:

Bill Spight wrote:Yes, and yes.

Bill, is there a reference somewhere to explain why B wins by 1 point under Chinese area scoring ?

Black wins because he will have 13 points (5 points of territory + 8 living stones) while white will only have 12 points (5 territory + 7 living stones).

From the Chinese Rules of Go:

I have no better reference than this: http://games.sports.cn/datebase/encyclo ... 90880.html

From Sensei's Library: http://senseis.xmp.net/?ChineseRules

DrStraw wrote:I seems to me the height of arrogance for an upstart neophyte go playing country like the USA to try to tell the rest of the world that they have it wrong. Japanese rules have worked since before the USA was even populated by westerners, let alone be a go asssociation.

Actually I think the best reason to use AGA rules (or Area scoring) is because we are a neophyte go playing country. Japanese rules requires judges or arbiters for decisions. Many tournaments in this country occur, and we do not have many strong players around in all parts of the country to make these decisions. AGA rules let people play and score the game as they learned with the benefit that any tournament director can just tell players to play out any situations and not change the score.

EdLee wrote:Hi Bill,

Glad you could join us in this thread. I'm still confused, sorry. Specifically, about:

go-master wrote:I know that territory and area scoring might differ by one point (if there is only one pass for both players).

emeraldemon wrote:Area and territory scoring differ when...

moyoaji wrote:In that situation black wins by 1 point, yes.
The same score would occur under Chinese area scoring rules, but the game would be a tie under Japanese rules.

Could we return to the above example:

$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------

Click Here To Show Diagram Code

[go]$$ Komi = 0.0
$$--------------
$$ | . 1 . 2 . |
$$ | . 3 . 4 . |
$$ | . 5 . 6 . |
$$ | . 7 . 8 . |
$$ | . 9 . 0 . |
$$--------------[/go]

$$Bm11 komi = 0.0
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------

Click Here To Show Diagram Code

[go]$$Bm11 komi = 0.0
$$--------------
$$ | . X 1 O . |
$$ | . X 3 O . |
$$ | . X 5 O . |
$$ | . X 4 O . |
$$ | . X 2 O . |
$$--------------[/go]

Under Chinese area scoring, does B win by 1 point ?
Under Japanese territory scoring, is it a tie (5 points vs. 5 points) ?

Under AGA rules it would be B+1 either way, because white's last pass would be a net prisoner in favor of black. If counting by area, black controls one more intersection, and if counting by territory, the territories are equal but black would have one more prisoner.

It's important to note that this game wouldn't normally be one of those examples where Chinese and Japanese scoring lead to a different result, because area scoring rulesets typically require odd komi because there are an odd number of intersections on a 19x19 (or 5x5) board. Combined with the fact that one typically disallows ties by adding a half point, this means that a komi of 1.5, 3.5, 5.5, etc. or -0.5, -2.5, -4.5, etc. would be used.

Let's try komi 1.5 on the game above:

komi = 1.5
Japanese score = W+1.5
Chinese score = W+0.5 (which would normally be reported as W+0.25 due to the use of half-counting), but still, white wins.

komi = -0.5
Japanese score = B+0.5
Chinese score = B+1.5 (normally reported as B+0.75 in half-counting)

We don't do komi = 0.5, which would change the winner.