Korean Indirect Kos

23 Aufrufe
Direkt zur ersten ungelesenen Nachricht

Robert Jasiek

ungelesen,
12.07.2004, 11:18:4712.07.04
an
The KBA 1992 Rules have the following precedental rule:

>>
# O a . # O . . .
. # # # # O . . .
# # O O O O . . .
O O O . . . . . .
. . . . . . . . .

In [the figure], if Black clearly has [more ko-threats], Black a is not
needed.
<<

There are further precedents, but before we can understand all of them,
we might as well first try to understand the above one.

***

Apparently, the ruling refers to the alternating-sequence.

"clearly" is, OC, equivocal. However, in some good sense it means
something like: "Just so many [more that the ko is won.]"

"ko-threat" requires a definition. This will not be trivial.

Where there are ko-threats, one has to define "ko" as well. Matters are
much worse than that because obviously the precedent is for
"indirect-kos" only. They require a definition, too.

"ko" is defined in J2003. Before we can define "indirect-ko", it is
convenient to define "direct-ko".

***

Definition: A _move-sequence_ is a sequence of moves that starts from a
position by one of the players, lets the players alternate moves, ends
in a position, and follows the basic-ko-rule.

Notes: This is a move-sequence of Korean type. This term is valid only
for modern Korean rules. The basic-ko-rule is defined in J2003.

Definition: A _direct-ko_ is either a visible-direct-ko or a
hidden-direct-ko.

Definition: A _visible-direct-ko_ is a ko so that a player can force
that he controls the ko and the opponent can force that the opponent
controls the ko.

Notes: "force" is defined as in J2003. "controls" is defined as in NAJ.
However, these terms refer to the Korean type of move-sequence.

Definition: A _hidden-direct-ko_ consists of two adjacent intersections
so that at least one of the players can force a visible-direct-ko on
these intersections.

Note: Bent-4-in-the-corner has a hidden-direct-ko.

***

Definition: An _indirect-move-sequence_ for a player is a sequence of
moves that starts from a position by the player, lets the players
alternate moves, lets the opponent pass in between the player's plays at
least once, ends in a position, and follows the basic-ko-rule.

Definition: [_indirect-force_: like "force" but using
"indirect-move-sequence"]

Definition: An _indirect-ko_ is either a visible-indirect-ko or a
hidden-indirect-ko.

Definition: A _visible-indirect-ko_ is a ko so that a player can
indirect-force that he controls the ko and the opponent can force or
indirect-force that the opponent controls the ko.

Definition: A _hidden-indirect-ko_ consists of two adjacent
intersections so that at least one of the players can force a
visible-indirect-ko on these intersections.

***

In the figure, the thing is a visible-indirect-ko.

The definitions above are drafts that need testing in many shapes.

Next we need to specify where a ko-threat might be played.

***

Definition: For a ko, _ko-threat-local_ is all intersections not in the
ko.

Note: This definition will be useful for a rule.

***

What is a "ko-threat"? Apparently, a purpose of the precedental rule is
to refer only to such ko-threat-local plays that are "big" enough in
some sense. "ko-threat" should be defined accordingly.

However, what is "big"? Which size is measured and compared with what?

There will be further problems: Suppose we have a local-score-region for
a given indirect-ko and the rest of the board. Is the searched size
well-defined if the rest of the board contains further indirect-kos (or
other precedental shapes)?

It is possible to define something for "big" and let it be well-defined.
However, there are several design possibilities and I would like to hear
your opinion, especially if you are familiar with Korean rules in
practice.

***

As can be seen, the horror of J1989 pales in comparison with K1992... It
is an open question which of K1992, Ing1986, Ing1991, J1949, J1979 turns
out to be the most horrible ruleset ever. K1992 has some chances here
because of its reference to ko-threats.

--
robert jasiek

Bernd Gramlich

ungelesen,
13.07.2004, 04:49:1613.07.04
an
Robert Jasiek wrote:

> The KBA 1992 Rules have the following precedental rule:
>
>>>
> # O a . # O . . .
> . # # # # O . . .
> # # O O O O . . .
> O O O . . . . . .
> . . . . . . . . .
>
> In [the figure], if Black clearly has [more ko-threats], Black a is
> not needed.
> <<
>
> There are further precedents, but before we can understand all of
> them, we might as well first try to understand the above one.

How about the following simple explanation? "Suppose that there are no
passes for ko and that two passes end hypothetical play. If white does
not have a hypothetical strategy which achieves a better score than the
score with the whole corner declared black territory, black need not
play a ."

***

Could you please point me to an English version of the complete KBA
rules? Thanks in advance!

--
Bernd Gramlich [bE6nd "gRamlIC]

Robert Jasiek

ungelesen,
13.07.2004, 05:00:0313.07.04
an
On Tue, 13 Jul 2004 10:49:16 +0200, Bernd Gramlich <be...@tenuki.de>
wrote:

>How about the following simple explanation? "Suppose that there are no
>passes for ko and that two passes end hypothetical play. If white does
>not have a hypothetical strategy which achieves a better score than the
>score with the whole corner declared black territory, black need not
>play a ."

I have had some such simplistic approach in mind indeed. (Nice that you
see it, too!) However, before I specify it, it must be clarified what
"the whole corner is" and whether the current position is considered or
a position after a local play out favouring Black, White, or either
player.

>Could you please point me to an English version of the complete KBA
>rules? Thanks in advance!

They are in the Hankuk Kiwon GuideBook and at

http://english.dashn.com/korea_baduk/koreabaduk_05.htm

--
robert jasiek

robert pauli

ungelesen,
14.07.2004, 07:23:5614.07.04
an
jas...@snafu.de (Robert Jasiek) wrote in message news:<40f3a464...@news.snafu.de>...

Dear Robert,

could you _please_ also develop a theory explaining

Figure 9 - All the Black groups are dead
because of Hujeolsu (snapback).

(http://english.dashn.com/korea_baduk/koreabaduk_05_06.htm)

;-)

Best
RP

Robert Jasiek

ungelesen,
14.07.2004, 08:00:0914.07.04
an
On 14 Jul 2004 04:23:56 -0700, robert...@epost.de (robert pauli)
wrote:

>could you _please_ also develop a theory explaining
>
> Figure 9 - All the Black groups are dead
> because of Hujeolsu (snapback).

Before I will explain such comparatively simple things, I try to solve
the more difficult parts of the rules. Thereby I know that I would not
study something essentially in vain.

(Snapbacks under K1992 will be explained by a pointer to J2003's
capturable-1, I guess. You know, I have developed J2003 and NAJ only to
explain Korean, Ing, and Chinese rulesets ;( )

--
robert jasiek

Nick Wedd

ungelesen,
14.07.2004, 08:56:3314.07.04
an
In message <40f51e81...@news.snafu.de>, Robert Jasiek
<jas...@snafu.de> writes

I'm not sure whether Robert is joking.

Here is the first group of which it is said "All the Black groups are

dead because of Hujeolsu (snapback)."

. # O . .
. # O . . upper left corner


# # O . .

. # O . .
# # O . .


O O O . .
. . . . .
. . . . .

Nick
--
Nick Wedd ni...@maproom.co.uk

Robert Jasiek

ungelesen,
14.07.2004, 09:24:2914.07.04
an
On Wed, 14 Jul 2004 13:56:33 +0100, Nick Wedd <ni...@maproom.co.uk>
wrote:

>Here is the first group of which it is said "All the Black groups are
>dead because of Hujeolsu (snapback)."
>
> . # O . .
> . # O . . upper left corner
> # # O . .
> . # O . .
> # # O . .
> O O O . .
> . . . . .
> . . . . .

Both the GuideBook and the dashn versions contain heavy editing
mistakes. Luckily, those differ, so it is possible to identify them;)

The handbook says: "Figure 7 / In A and B Black's groups are alive. In C
and D the internal groups are alive."

--
robert jasiek

Allen antworten
Dem Autor antworten
Weiterleiten
0 neue Nachrichten