Shenoute wrote:Hi Screature,
Just to supplement and illustrate what has already been said.
It is important to understand that, given optimal play by both players, filling the neutral points doesn't really change the score (even if to a beginner it may seem like they do). For instance, in the following diagram, it may seem that black has eleven points...
$$c
$$ ----------
$$ | . . . .
$$ | O O . .
$$ | . O O .
$$ | X X O .
$$ | . X O O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | X X X O
$$ | O O O O
- Click Here To Show Diagram Code
[go]$$c
$$ ----------
$$ | . . . .
$$ | O O . .
$$ | . O O .
$$ | X X O .
$$ | . X O O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | . . X O
$$ | X X X O
$$ | O O O O[/go]
[...] Now black's territory is ten points. But, it is important to understand that this does not really mean that black's territory has been reduced from eleven to ten points and that the score has changed. Both players knew (or should have known) beforehand that
b would have to be filled and that it was not a point to start with.
Actually, the official text of the japanese rules has always said that, left as it is, in this position, Black has not eleven points.
It's a shady part of the japanese rule.
From 1949 to 1989, the official territory points of Black were 10. But since 1989, they are 0 as long as the neutral point is not filled (and obviously 10 once the neutral point is filled and Black has connected).
Another extremely confusing point is whether there should be one point of territory inside a ko that is left open at the end. In short, the answer is no, but it is practically impossible to find a part of the rules, old or new, that clearly states so, at least in their available english translations.
These problems have given nightmares to programmers who want to develop go software. For this reason, AlphaGo plays in chinese rule only.
And when a question of this kind arises in kgs, an admin must connect and modify manually the final score according to the real japanese rule. CGoban can't count it right. It happened to me once in a position with a two-step mandatory connection at the end.
In conclusion, this is just a matter of convention. The japanese tradition has always considered mandatory to solve all positions of this kind before ending the game.
But writing a clear ruleset including this feature is so problematic that go federations worldwide seem to progressively abandon the japanese rule in favor of the AGA rule.
China first stopped using the japanese rule in 1975, and used a more complicated way of counting the score (stones + territory instead of territory minus prisoners), but with a much simpler rule, soon followed by New Zealand and the Ing foundation.
Later, in 1991, the american federation introduced pass stones with their AGA rule, so as to eliminate these problems without complicating the process of counting the score for the players.
It was followed by the french federation in 1993 and the british one in 2008.
In 2015 (if I'm not mistaken), the European Go Congress adopted the AGA rule too.
Another solution would have been to do just as you suggest : White didn't fill the neutral point, thus Black has got eleven points.
But it is not as easy as it seems, because in other kind of positions, a complex process is still needed to clearly define the score (look for the "simplified japanese rule", that does just that).
So-called "neutral points" may not actually be neutral. Your example of atari is to the point. Suppose that you make a play that puts some of my stones in atari, threatening to take them unless I connect, can I say you can't play there, it's a neutral point? Of course not. The rules of go allow you to play anywhere, except for the ko or superko rule or self-sacrifice (under most rules). "Neutral points" are not an exception. 