Situational superko vs. Positional superko
ands...@gmail.com
superko over positional superko.
It has been stated that positional superko is superior to situational
superko on the grounds of simplicity. Indeed the usual definitions of
situational superko are hard to grasp immediately, even though once you
understand what they say you are unlikely to ever be confused.
Nonetheless, they all sound a bit awkward in comparison to positional
superko rule.
Here are a few examples of the situational superko rule:
1) A play may not recreate a previous board position from the game,
with the same player to move next.
2) A player is not allowed to make a play so that the resulting board
position repeats the whole board position as it was after any of that
player's previous moves.
3) A board play may not recreate a position previously left by the
player.
In contrast, positional superko rules are straightforward:
1) A play may not recreate a previous board position from the game.
2) A board play may not recreate a position.
3) A move should not repeat an earlier grid coloring.
All these definitions refer to the concept of "board position" or
simply "position" (or "grid coloring" in the language of Tromp-Taylor
ruleset). This concept however is deficient in that it completely
ignores one important aspect of the game, i.e. which player will move
next. In any position during the game it is always either black to move
or white to move. Therefore, we need to introduce the concept of
"state", which refers to the combination of board position and whose
turn to play it is. For example, at the start of the game the board is
empty (all grid is colored empty) and it is black to move. This defines
the game state. With this definition, situational superko rule simply
reads: "A play may not recreate a previous board state from the game."
This is just as easy as positional superko rule. What's more,
positional superko rule seems to imply that it does not matter whose
turn to play it is, which is ridiculous.
So, the punch-line is that simplicity of definition cannot be used to
decide on whether to adopt positional or situational superko as both
rules can be stated simply. The other guiding principle that can be
used to decide on the best rules is to require the rules to impose as
few restrictions on possible moves as possible. With this guiding
principle, situational superko rule is clearly superior to positional
superko. The point of all ko-rules is to prevent infinite cycles of
repetion and situational superko does exactly that. Positional superko,
on the other hand, forbids some moves that don't lead to infinite
cycles.
By the same guiding principle, suicide should be allowed as there are
no good reasons to forbid it. Forbidding suicide just limits the number
of possible plays available to a player. As for area scoring vs.
territory scoring, it's been said many times already that area scoring
is much better and I will not repeat those arguments here. So, to sum
up, a good ruleset should use area scoring, situational superko, and
allow suicide. New Zealand ruleset (supplemented by Tromp-Taylor
explanations) follows this and I'd like to encourage everyone to use
this ruleset when playing on KGS and other servers.
The only thing I don't like about this ruleset is that it does not
define which player will move first if the players disagree on the
status of some groups and need to resume play (after two consecutive
passes) to clarify the situation. However, this problem never arises on
KGS as it preserves the order of moves, so that if white was last to
pass then black will be the first to resume play.
none
> The point of all ko-rules is to prevent infinite cycles of
> repetion and situational superko does exactly that. Positional superko,
> on the other hand, forbids some moves that don't lead to infinite
> cycles.
Good point, but not entirely accurate. Situational superko can also
declare moves illegal that would not necessarily lead to infinite cycles.
Erik
PS I think I showed it here before, but in case you missed it here's a
fun superko problem from my 'stellingen'.
. O O # # # O O .
O . O # . # O O #
O O # . # O O # .
# # # # # O O # #
O O O O # # O O .
. . O # # # # O O
O O # # O O O # O
O . O # O . O # #
O . O # O O . # .
9x9 board, area scoring, 6 komi for White,
Super ko (positional or situational does not matter).
It also does not matter who plays first.
For a hint see below.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
Assuming both players know how to play optimally, the position can be
scored directly as it stands. They would then agree that White's upper
left group is dead (which can be confirmed by optimal playout under for
example Tromp-Taylor rules) and the score would be jigo. Of course White
could decide to spare his 'unconditionally alive' group, but then Black
would win.
Under rules that distinguish between balanced and unbalanced cycles
Black would of course loose...
Juha Nieminen
> superko. The point of all ko-rules is to prevent infinite cycles of
> repetion and situational superko does exactly that. Positional superko,
> on the other hand, forbids some moves that don't lead to infinite
> cycles.
Situational superko is not completely free of problems either.
There's for example an exceptional case where a dead group becomes
alive simply because the opponent would have to repeat a board position
in order to capture it. It really doesn't make sense for such group
to magically become alive just because the opponent has no other moves
than the one which would break situational superko. (If superko was
not in use then he could capture the group in question by repeating
the board position just once while the owner of the dead group cannot
do anything about it.)
Superko is tempting because it's so short to formulate and so easy
to understand. However, when there's a very complex situation with
lots of kos and other similar things, especially in a small board,
it may become quite difficult to impose superko. It's trivial for
a computer to do, but for a human it can become quite difficult.
And what should be done if the players disagree on who broke superko
first?
> By the same guiding principle, suicide should be allowed as there are
> no good reasons to forbid it. Forbidding suicide just limits the number
> of possible plays available to a player.
Suicide changes some tactics. In some rare situations it provides
ko threats, and in some quite exceptional cases it can even change
the winner of a semeai.
I'm not saying this is a bad thing. I'm just saying that one should
be aware that allowing suicide is not a no-op.
> As for area scoring vs.
> territory scoring, it's been said many times already that area scoring
> is much better and I will not repeat those arguments here.
For online playing, definitely. For teaching beginners about
dead-stone removal, definitely. For experienced players who do not
have problems with dead-stone removal... Well, only if they are fluent
at area counting. I personally am not although I have tried. I'm really
not very eager to start playing live games using area counting...
-
none <""erikw\"@(none)"> wrote:
>Erik
>
> PS I think I showed it here before, but in case you missed
> it here's a fun superko problem from my 'stellingen'.
Did not find a prior reference, actually.
> . O O # # # O O .
> O . O # . # O O #
> O O # . # O O # .
> # # # # # O O # #
> O O O O # # O O .
> . . O # # # # O O
> O O # # O O O # O
> O . O # O . O # #
> O . O # O O . # .
>
> 9x9 board, area scoring, 6 komi for White,
> Super ko (positional or situational does not matter).
> It also does not matter who plays first.
>
> -( spoiler alert snipped )-
>
> Assuming both players know how to play optimally, the
> position can be scored directly as it stands. They would
> then agree that White's upper left group is dead (which
> can be confirmed by optimal playout under for example
> Tromp-Taylor rules) and the score would be jigo. Of
> course White could decide to spare his 'unconditionally
> alive' group, but then Black would win.
>
> Under rules that distinguish between balanced and
> unbalanced cycles Black would of course loose...
Clear as mud. For benefit of other readers (and myself)
would Erik be so kind as to explain each of the steps?
Could the upshot of this justify Japanese counting rules?
- regards
- jb
----------------------------------------------------------------------
From: erikvanderwerf at gmail.com (Erik van der Werf)
Date: Fri Jul 29 01:52:44 2005
Subject: [computer-go] Protocol B
[ ... ]
http://computer-go.org/pipermail/computer-go/2005-July.txt
----------------------------------------------------------------------
Chris Dams
On Sun, 2005-10-09 at 18:23 +0000, - wrote:
> > . O O # # # O O .
> > O . O # . # O O #
> > O O # . # O O # .
> > # # # # # O O # #
> > O O O O # # O O .
> > . . O # # # # O O
> > O O # # O O O # O
> > O . O # O . O # #
> > O . O # O O . # .
> >
> Clear as mud. For benefit of other readers (and myself)
> would Erik be so kind as to explain each of the steps?
> Could the upshot of this justify Japanese counting rules?
Maybe I can be of some help too. I think Erik is talking about the
following possibility that Black has
. O O # # # O O 1
O . O # . # O O #
O O # . # O O # 2
# # # # # O O # #
O O O O # # O O .
. . O # # # # O O
O O # # O O O # O
O 3 O # O . O # #
O . O # O O . # .
(moves 1 to 3)
. O O # # # O O .
O . O # . # O O 7
O O # . # O O # O
# # # # # O O # #
O O O O # # O O .
6 5 O # # # # O O
O O # # O O O # O
O # O # O . O # #
O 4 O # O O . # .
(moves 4 to 7)
Now White might think that he better pass, but if we have 8: pass, we
get
. O O # # # O O 9
O . O # . # O O #
O O # . # O O # .
# # # # # O O # #
O O O O # # O O .
O . O # # # # O O
O O # # O O O # O
O . O # O . O # #
O O O # O O . # .
(move 9)
And now, unfortunately for White, he may not capture the two UR black
stones, because of superko. So, he had better not pass, but instead has
to give up the group in the UL-corner as this is the smallest loss.
The existence of such examples is by some considered to be an argument
against superko. I think seeing it as an argument in favour of Japanese
counting goes to far, as it is imaginable to use area scoring with other
ko-rules than superko.
Best wishes,
Chris
Erik van der Werf
> Maybe I can be of some help too. I think Erik is talking about the
> following possibility that Black has
Right, you found the solution. Below is an sgf to clarify things a bit more.
Best,
Erik
(;GM[1]FF[3]
RU[New Zealand]CP[Erik van der Werf]SZ[9]HA[0]KM[6]
PW[White]
PB[Black]
GN[Stelling 3 behorende bij het proefschrift "AI techniques for the game
of Go"]
DT[2005-01-27]
AW[ba][ca][ga][ha][ab][cb][gb][hb][ac][bc][fc][gc][fd][gd][ae][be][ce][de][ge][he][cf][hf][if][ag][bg][eg][fg][gg][ig][ah][ch][eh][gh][ai][ci][ei][fi]
AB[da][ea][fa][db][fb][ib][cc][ec][hc][ad][bd][cd][dd][ed][hd][id][ee][fe][df][ef][ff][gf][cg][dg][hg][dh][hh][ih][di][hi]
C[Super ko,
Area scoring,
6 komi for White
"Super-ko regels kunnen spelers er toe dwingen de ogen te vullen van
groepen die normaal gesproken onconditioneel leven"
]
(;B[ia]
C[Let's assume that Black plays first. He attacks the large white group
that would normally be alive in seki...]
;W[ic];B[bh];W[bi];B[bf];W[af]C[Remember this position!];B[ib]
C[Here passing seems like the natural choice, but...]
(;W[tt]C[White's pass would be a mistake (under superko rules)];B[ia]
C[because now White is not allowed to play at J7]
(;W[tt];B[ie]TW[aa][bb][bf][bh]
TB[ga][ha][eb][gb][hb][dc][fc][gc][ic][fd][gd][ge][he][hf][if][eg][fg][gg][ig][eh][fh][gh][ei][fi][gi][ii]
C[The game is over. Final score:
White = 4 territory + 19 living stones + 0 dame + 6 komi = 29
Black = 26 territory + 32 living stones + 0 dame = 58
Black wins by 29.])
(;W[ic]C[This is illegal by Super ko (look back 4 positions).])
)
(;W[bb]
C[To prevent loss by Super ko, White is now forced to sacrifice his
smallest group to change the board position/situation.]
;B[tt];W[tt]TW[bf][bh][fh]
TB[aa][ba][ca][ab][bb][cb][eb][ac][bc][dc][ic][ii]
C[With superko this is the optimal result.
The game is over. Final score:
White = 3 territory + 33 living stones + 1.5 dame + 6 komi = 43.5
Black = 12 territory + 30 living stones + 1.5 dame = 43.5
Jigo (Tie game).])
)
(;B[bh];W[bi];B[ia];W[ic];B[bf];W[af]
C[Variations such as this one transpose back to the previous solution.])
(;B[tt];W[tt]TW[aa][bb][af][bf][bh][fh][bi]TB[eb][dc][ic][ii]
C[Without superko this would of course be the normal result.
The game is over. Final score:
White = 7 territory + 37 living stones + 1.5 dame + 6 komi = 51.5
Black = 4 territory + 30 living stones + 1.5 dame = 35.5
White wins by 16.])
(;W[bi]
C[If White plays first he can only remove one of the two superko threaths.]
;B[ia];W[ic];B[bf];W[af]C[This transposes back to the first solution])
(;W[tt]TW[af][bf][bh][fh][bi]
TB[aa][ba][ca][ab][bb][cb][eb][ac][bc][dc][ic][ii]
C[So both players might just as well agree that White's upper left
'unconditionally alive' group is dead...
The game is over. Final score:
White = 5 territory + 31 living stones + 1.5 dame + 6 komi = 43.5
Black = 12 territory + 30 living stones + 1.5 dame = 43.5
Jigo (Tie game).])
)
-
> Chris Dams wrote:
>> Maybe I can be of some help too. I think Erik is talking about the
Erik van der Werf <E.vander...@CSdotUnimaas.nl> wrote:
> Right, you found the solution. Below is an sgf to clarify things a bit more.
Thanks to both of you and much appreciate the effort and work.
This has probably been discussed before, however isn't "score" a
component of the position? If the table game score has changed
even though the stone arrangements are identical, then hasn't the
position also been made different? For argument, these definitions:
( Oxford English Dictionary )
==================================== begin
position, n.
[ ... ]
4. The action of positing or placing, esp. in a particular order or
arrangement; disposition. Obs.
1735 BERTIN Chess iii, The Game of Chess consists of two parts,
the Offensive, and Defensive;..the Defensive [consists] in the due
position of your own [forces], by guarding against your enemy's attack.
5. a. The manner in which a body as a whole, or the several parts of it,
are disposed or arranged; disposition, posture, attitude.
1790 PHILIDOR Chess II. 90 In this position it is a drawn game.
DRAFT ADDITIONS DECEMBER 2001
position, n.
As a mass noun: the state of being favourably or advantageously
placed, esp. in a competitive situation; advantageous placing or
condition; see also to jockey for position s.v. JOCKEY v. 1d. In
Chess, Snooker, and other games of strategy: configuration of
play which confers an advantage over an opponent; favourable
location or disposition.
1950 L. H. DAWSON Hoyle's Games Modernized (ed. 20) III. 340
The most successful [pool] player is not necessarily he who can ‘pot’
with the deadliest accuracy, but he who combines potting with
effectively playing for position.
==================================== end
Ironically, among the senses for "position" is "disposition." :-)
Disposition connotes "relative position" that is relative to something
by which they are each relative. Game positions would be relative to
each other, and an ordering is conceivable owing to CGT formulation.
In particular, "the state of being favourably or advantageously
placed..." would seem to be altered by a situation where an apparent
table arrangement looked the same however prisoner counts for each
player were in different quantity so as to modify the total score. The
term "position" applies to a (game) state, not merely the arrangement.
Additionally a "position" description might also need to include extra
information about status of _ko_ point(s) because there is a game state
distinction between being capable or incapable of capturing a _ko_.
Counting rules that ignore prisoners would be justified if and only
if they can be harmonized with prisoner accounting methods, else we
have instead competing notions for what the Go Game ought to be.
If one argues for competing notions of the game itself, then there is
no effective procedure for determining which game version ought to
be applied for any given random tournament.
- regards
- jb
---------------------------------------------------------------------------
9/11 Videos: THE CONTROLLED COLLAPSE OF WTC 7
http://www.whatreallyhappened.com/wtc7.html
---------------------------------------------------------------------------
FBI Claims 84 Videos Show NO Flight 77 Impact
http://www.rense.com/general67/fbicl.htm
---------------------------------------------------------------------------
Simon Goss
>By the same guiding principle, suicide should be allowed as there are
>no good reasons to forbid it.
Rule sets that allow suicide usually don't allow single-stone suicide,
since single-stone suicide allows a player to obstruct the completion of
the game by giving up the turn without either playing or passing.
>New Zealand ruleset
<...>
>The only thing I don't like about this ruleset is that it does not
>define which player will move first if the players disagree on the
>status of some groups and need to resume play (after two consecutive
>passes) to clarify the situation.
That's a good rule. It means that if both players have passed when the
winner depends on who plays first, then neither side can benefit. If you
can win in such a situation regardless of who plays first, you just let
your opponent choose who plays first.
--
Simon
UK Go Challenge for schools
http://www.ukgochallenge.com
denis feldmann
> ands...@gmail.com writes
>
>
>>By the same guiding principle, suicide should be allowed as there are
>>no good reasons to forbid it.
>
>
> Rule sets that allow suicide usually don't allow single-stone suicide,
> since single-stone suicide allows a player to obstruct the completion of
> the game by giving up the turn without either playing or passing.
Superko rules already prevent that
-
> Simon Goss a écrit :
>> Rule sets that allow suicide usually don't allow single-stone suicide,
>> since single-stone suicide allows a player to obstruct the completion
>> of the game by giving up the turn without either playing or passing.
denis feldmann (ignoramus froggie killfiler) wrote:
> Superko rules already prevent that
Not if score is changing under a game state "position" definition.
- regards
- jb
----------------------------------------------------------------------
How the Human Brain works
http://my.fcc.net/~workgroup5/sup/howbrainworks.htm
----------------------------------------------------------------------
Robert Jasiek
>I'd like to share a few ideas about the advantages of situational
>superko over positional superko.
A fair comparison should regard not just a few arbitrary aspects.
Therefore here is a more profound discussion:
Aspects:
1) simplicity
2) familiarity
3) explicability
4) perception
5) strategic consequences
6) strategic variation
7) fulfilment of rules purposes
8) consistency
One may doubt whether this list is complete, but then so far no more
complete list has been studied. It depends on one's preferences which
aspects should be weighed how much. Now the aspects are considered for
- PSK = positional superko
- SSK = situational superko
- NSK = natural situal superko
(Of course, one could also analyse the dozens of other superko
variants.)
PSK:
(1) The rule text is the simplest of all superko variants.
(2) For most players, this is believed to be THE superko rule. The
Japanese 1949 Rules, the WAGC 1979 Rules, the Chinese 1988 Rules all
used this superko variant.
(3) Explanation of this rule is easier than explanation of other
superko rules.
(4) The rule refers to visible information only. This makes an
intuitive application easier than for other superko variants.
(5) There is exactly one known rare (occurred once) special shape
class with very strange behaviour (Molasses Ko) and the even rarer
(has never occurred) no-good-tenuki class with surprising behaviour. A
few classes of positions have other behaviour than SSK and NSK.
(6) Strategic variation is not maximized.
(7) The rule can always be applied in principle. It cannot be applied
in practice in case of some shape classes that are so rare that they
have never occurred so far.
(8) The rule is applied consistently.
SSK:
(1) The rule text is not the simplest of all superko variants.
(2) For most players, this is not believed to be the superko rule. The
New Zealand Rules and the AGA 1991 Rules use this superko variant.
(3) Explanation of this rule is more difficult than explanation of
PSK.
(4) The rule refers to visible and invisible information. This makes
an intuitive application more difficult than with PSK.
(5) There is the extremely rare (has never occurred) known
no-good-tenuki class with surprising behaviour. A few classes of
positions have other behaviour than PSK and NSK.
(6) Some attempt to increase strategic variation beyond PSK is made.
(7) The rule can always be applied in principle. It cannot be applied
in practice in case of some shape classes that are so rare that they
have never occurred so far.
(8) There is inconsistency: Causes of prohibiton can be due to board
plays or pass plays - application of prohibition only for board plays.
NSK:
(1), (3), (4), (6), (7) are as for SSK.
(2) For most players, this is not believed to be the superko rule. No
ruleset has used it so far.
(5) There is the extremely rare (has never occurred) known
no-good-tenuki class with surprising behaviour. After some pass play a
few classes of positions (like double ko seki) have unfamiliar move
sequences compared to PSK and SSK. A few classes of positions have
other behaviour than PSK and SSK.
(8) There is consistency: Causes of prohibiton and application of
prohibition are both only for board plays.
>All these definitions refer to the concept of "board position" or
>simply "position" (or "grid coloring" in the language of Tromp-Taylor
>ruleset). This concept however is deficient in that it completely
>ignores one important aspect of the game, i.e. which player will move
>next. [...]
It ignores further aspects. See above and besides recall a) thinking
time and b) prior history of prohibited plays during a game. There is
no a priori reason to consider only one particular extra aspect.
>the concept of
>"state", which refers to the combination of board position and whose
>turn to play it is.
This concept has had the term "situation".
>[...]This defines
>the game state. With this definition, situational superko rule simply
>reads:
One can replace the entire ruleset by an abbreviation if all one wants
is a shorter text. Thus your argument is very weak.
>What's more,
>positional superko rule seems to imply that it does not matter whose
>turn to play it is, which is ridiculous.
Call something "ridiculous" and you are expressing your opinion, not
facts.
>The other guiding principle that can be
>used to decide on the best rules
Nonsense, since there is more than one other aspect, see above. Thus
one may not speak of THE (one and only) other guiding principle.
>is to require the rules to impose as
>few restrictions on possible moves as possible.
One can like greater strategic variation or the opposite. There is no
a priori reason why one would have to prefer either, if only the
variation is great enough to challenge many generations of players.
>The point of all ko-rules is to prevent infinite cycles of
>repetion and situational superko does exactly that.
As do all superko variants.
>Positional superko,
>on the other hand, forbids some moves that don't lead to infinite
>cycles.
So does even the basic ko rule: It prohibits some moves and thus
avoids such infinite cycles otherwise created by them.
>suicide should be allowed as there are
>no good reasons to forbid it.
"Suicide should be prohibited as there are no good reasons to allow
it."
And now what? The fact remains that rules need suicide either allowed
or prohibited.
>Forbidding suicide just limits the number
>of possible plays available to a player.
The necessiety for some suicide rule is something else than the
variation possible under it.
>As for area scoring vs. territory scoring, [...]
Now it becomes OT.
--
robert jasiek
Robert Jasiek
wrote:
> Situational superko is not completely free of problems either.
> There's for example an exceptional case where a dead group becomes
>alive simply because the opponent would have to repeat a board position
>in order to capture it. It really doesn't make sense for such group
>to magically become alive just because the opponent has no other moves
>than the one which would break situational superko.
You value maintenance of the status of particular shapes higher than
simplicity of the rules. I value simplicity of the rules higher than
considering any shapes together with rules.
BTW, go is about changing status of stones throughout the alternation.
>when there's a very complex situation with
>lots of kos and other similar things,
Indeed. When. It has NEVER occurred so far. The most complex thing was
a quadruple ko. That makes 2^4 = 16 states within the ko mouths. If
you cannot compare 16 positions with each other, then maybe you should
not play go...
>especially in a small board,
Which 19x19 is not.
>for a human it can become quite difficult.
Yawn. Life and death is much more difficult than remembering 16
positions that differ on only 8 intersections.
> And what should be done if the players disagree on who broke superko
>first?
Never heard of "referee"?
--
robert jasiek
ands...@gmail.com
of various ko-rules on your web site. In particular, your statement
that SSK rule is not as simple as PSK rule.
> SSK:
>
> (1) The rule text is not the simplest of all superko variants.
I am grateful to you and the others for exposing my fallacies. However,
I would like to reiterate the main point of my post, which I believe is
still valid even though you disagree with me. I think every go player
is familiar with the concept of state (or situation) intuitively even
if they don't have a name for it. In my post I simply gave a name for
something that everyone knows already (I would have used the term
situation but it didn't occur to me). I think it is a simple, natural,
and above all useful concept.
So now, my point was that SSK rule could be formulated as simply as PSK
rule if the term position is replaced with the term state. Obviously I
needed to define this term first, so one could argue that I just hid
the complexity of the SSK rule behind some extra definitions. This
could be a valid objection but one has to look at whether the new term
actually makes things simpler and easier to understand and formulate.
My definition of the game state helped me clarify and simplify my
intuitive undestanding of the SSK rule. I thought it might help the
others as well, so I made my post.
Chris Dams
On Mon, 2005-10-10 at 03:10 +0000, - wrote:
> Counting rules that ignore prisoners would be justified if and only
> if they can be harmonized with prisoner accounting methods, else we
> have instead competing notions for what the Go Game ought to be.
Really? If you ask me for my opinions on the what the go game ought to
be, I really wouldn't care about prisoner accounting methods. To me the
Go Game ought to
(1) be about controlling the biggest part of the board;
(2) include alternating play and the capture rule;
(3) be well-defined in the sense that it should be clear from the
wording of the rules that they can be, in a practical way, applied to
every game that is played obeying those rules.
> If one argues for competing notions of the game itself, then there is
> no effective procedure for determining which game version ought to
> be applied for any given random tournament.
There certainly are competing notions on what the go game ought to be.
For instance, in this newsgroup there are people who do not consider
point (3) of the above to belong to the what the go game ought to be.
These are the ones that don't care about the various attempts of
formalization of rules.
Best,
Chris
-
> "-" wrote:
>> Counting rules that ignore prisoners would be justified if and only
>> if they can be harmonized with prisoner accounting methods, else we
>> have instead competing notions for what the Go Game ought to be.
Chris Dams <C.D...@science.ru.nl> wrote:
> Really? If you ask me for my opinions on the what the go game ought
> to be, I really wouldn't care about prisoner accounting methods.
> To me the Go Game ought to
> (1) be about controlling the biggest part of the board;
Some groups are quite small, controlling only a small part of a board.
> (2) include alternating play and the capture rule;
More than simply "include" -- these are indeed "essential" to Go.
> (3) be well-defined in the sense that it should be clear from the
> wording of the rules that they can be, in a practical way, applied
> to every game that is played obeying those rules.
I seek to do this by having "score" be a component of "position."
>> If one argues for competing notions of the game itself, then there is
>> no effective procedure for determining which game version ought to
>> be applied for any given random tournament.
> There certainly are competing notions on what the go game ought
> to be. For instance, in this newsgroup there are people who do not
> consider point (3) of the above to belong to the what the go game
> ought to be. These are the ones that don't care about the various
> attempts of formalization of rules.
If they're not contributing to this newsgroup nor reading what we
have to say, then we can't do very much about that here, can we?
Yet what is unclear about the proposition that, during hypothetical
play ("confirmation and agreement"), all ko threats may first be filled
before approaching the problem situation under jurisdiction? What
is unclear about the proposition that "score" be made a property of
"postion" so that a table arrangement of stones is -not- the same if
the net score has changed? If, on successive iterations, the same
color player keeps pulling ahead, so that it is no longer necessary
to delay further attempts to score the game?
What is unclear about the propostion that the panoply of various
counting methods be brought into agreement and harmonization so
that players arriving from different countries and traditions do not
face difficulties at language comprehension over the intricacies of
competing rulesets? Barring special exemptions for Ing Rules, etc.
What is unclear about proposing that a special referee, equipped
with rules encyclopedia, be detailed to Mr. Jasiek at tournaments?
- regards
- jb
----------------------------------------------------------------------
What Really Happened on September 11th?
http://911research.wtc7.net/sept11/analysis/
----------------------------------------------------------------------
Chris Dams
On Mon, 2005-10-10 at 19:19 +0000, - wrote:
> Yet what is unclear about the proposition that, during hypothetical
> play ("confirmation and agreement"), all ko threats may first be filled
> before approaching the problem situation under jurisdiction?
For somebody who regularly points out ambiguities and errors in posts,
way too unclear. I'm still doubting whether it is unhelpful, wrong or
not applicable in any case whatsoever.
> What
> is unclear about the proposition that "score" be made a property of
> "postion" so that a table arrangement of stones is -not- the same if
> the net score has changed? If, on successive iterations, the same
> color player keeps pulling ahead, so that it is no longer necessary
> to delay further attempts to score the game?
Even though it is clear, I do not believe we need such a modification.
There are several reasons for this.
(1) Anomaly is in the eye of the beholder, to quote Bill Spight.
(2) On a 9x9 board, this anomaly is already very rare. On a 19x19 board
it is even less probable.
(3) Such a modification should not be discussed in isolation, but in the
context of a complete proposal for a rule text as that is the only way
to be able to discuss other, maybe much worse, anomalies that might
arise from the fix.
(4) If every perceived anomaly is going to result in an addition to the
rules, we might as well give up the quest for applicable and clear
rules.
> What is unclear about the propostion that the panoply of various
> counting methods be brought into agreement and harmonization so
> that players arriving from different countries and traditions do not
> face difficulties at language comprehension over the intricacies of
> competing rulesets?
This is very unclear. Sounds great, though. Maybe you should go work for
the UN. You'll be great at writing speeches that sound beautiful, yet
accomplish nothing.
> What is unclear about proposing that a special referee, equipped
> with rules encyclopedia, be detailed to Mr. Jasiek at tournaments?
If this is necessary, clearly the tournament is using rules that do not
satisfy point (3).
Best,
Chris
-
> "-" wrote:
>> Yet what is unclear about the proposition that, during hypothetical
>> play ("confirmation and agreement"), all ko threats may first be filled
>> before approaching the problem situation under jurisdiction?
Chris Dams <C.D...@science.ru.nl> wrote:
> For somebody who regularly points out ambiguities and errors in posts,
> way too unclear. I'm still doubting whether it is unhelpful, wrong or
> not applicable in any case whatsoever.
I trust that you won't hold me to my any previous standards. :-)
Earlier we had the discussion on "`Chinese' vs `Japanese' scoring"
where Mr. Jasiek had asserted that there is no "remove all the ko
threats rule" in order to assess the status of ambiguous positions:
================== begin
Newsgroups: rec.games.go
From: jazze...@hotmail.com (-)
Date: Fri, 30 Sep 2005 16:16:13 GMT
Subject: Re: "Chinese" vs "Japanese" scoring
>> Robert Jasiek <jas...@snafu.de> wrote:
>>> He was right in that since there is no "remove all
>>> the ko threats" rule...
> jazze...@hotmail.com (-) wrote:
>> Indeed there is: that's covered under hypothetical
>> play, or "confirmation and agreement" (J1989).
Robert Jasiek <jas...@snafu.de> wrote:
> That is not a concept of removing ko threats but a concept
> that lets ko-passes be the only available useful ko threats.
Surely it cannot be correct to disallow a player the
move which removes a ko threat during "confirmation
and agreement."
================== end
That discussion was certainly applicable to a scoring question
concerning methodologies utilized in pursuing dispute resolution.
>> What
>> is unclear about the proposition that "score" be made a property of
>> "postion" so that a table arrangement of stones is -not- the same if
>> the net score has changed? If, on successive iterations, the same
>> color player keeps pulling ahead, so that it is no longer necessary
>> to delay further attempts to score the game?
> Even though it is clear, I do not believe we need such a modification.
> There are several reasons for this.
> (1) Anomaly is in the eye of the beholder, to quote Bill Spight.
Why is Bill Spight not in the eye of the beholder?
> (2) On a 9x9 board, this anomaly is already very rare. On a 19x19 board
> it is even less probable.
Would seem to me that any 9x9 problem could be recreated onto
a 19x19 board, employing proper construction of walls and so forth.
> (3) Such a modification should not be discussed in isolation, but in
> the context of a complete proposal for a rule text as that is the only
> way to be able to discuss other, maybe much worse, anomalies that
> might arise from the fix.
We have one question of arriving at a "rules fix" but quite another
question of arriving at general public understanding of a "rules fix."
I'm certain that you have beforehand found yourself in the situation
of understanding the answer to a question but being pestered by
incorrigible newbies intent on a refusal to understand what you do.
> (4) If every perceived anomaly is going to result in an addition to the
> rules, we might as well give up the quest for applicable and clear
> rules.
Is a large glass window "less clear" than a small glass window?
Japanese Rules cases may be handled by first filling ko threats.
>> What is unclear about the propostion that the panoply of various
>> counting methods be brought into agreement and harmonization so
>> that players arriving from different countries and traditions do not
>> face difficulties at language comprehension over the intricacies of
>> competing rulesets?
> This is very unclear. Sounds great, though. Maybe you should go
> work for the UN. You'll be great at writing speeches that sound
> beautiful, yet accomplish nothing.
Let's not criticize the UN for accomplishing nothing. At least
the UN is not marching around into all countries with a UN military.
>> What is unclear about proposing that a special referee, equipped
>> with rules encyclopedia, be detailed to Mr. Jasiek at tournaments?
> If this is necessary, clearly the tournament is using rules that do not
> satisfy point (3).
There is a distinction between whether rules satisfy point (3) and
whether such satisfication is satisfactory to Mr. Jasiek's satisfaction.
Here the Subject: line is "Situational superko vs. Positional superko."
Obviously there was a difference, and amounting to a tremendous
difference when computing the space of possible game graphs.
I am uncertain whether tying "score" to "position" can resolve
matters of result quibbling: create another color-reversed position
and include that into the overall game position score being tallied.
The "SST vs PST" questions might devolve to whether score has
increased or decreased when returning to a previous position. If the
score becomes the same then a position may not be repeated. Yet
if a score change has occurred then clearly the player who benefits
by change seeks position recycling while the other player does not.
- regards
- jb
Juha Nieminen
> You value maintenance of the status of particular shapes higher than
> simplicity of the rules. I value simplicity of the rules higher than
> considering any shapes together with rules.
Simplicity of rules cannot be the only principle.
Rules would be much simpler if there was no ko rule at all. In fact,
other things can be removed too. Passing is too complicated, so for the
sake of simplicity we could remove that as well. Counting is WAY too
complicated, so let's remove counting too. The player who cannot make
a legal move loses the game. Do we even need a rule for capturing? It
makes the ruleset needlessly complicated.
So, I propose this ruleset:
1. Players alternate at placing stones of their own color on
intersections on the board.
2. The first player who cannot make a move loses.
There. Very simple.
Oh, wait, too simple? Well, you valued simplicity of rules, didn't you?
By the way, just for a second consider this thought: Superko
(compared to not having superko) makes the game needlessly complicated.
Don't dismiss that idea right away. Think about it.
>>when there's a very complex situation with
>>lots of kos and other similar things,
>
> Indeed. When. It has NEVER occurred so far.
You seriously claim that in the about 3000 years of the history of Go
it has NEVER occurred that two beginners have got confused about who
broke the superko rule first?
Or is a "to me" missing from that sentence? Well, I have never had
any of my games voided using Japanese rules, yet game voiding sucks?
> The most complex thing was
> a quadruple ko. That makes 2^4 = 16 states within the ko mouths. If
> you cannot compare 16 positions with each other, then maybe you should
> not play go...
Yes, I'll make sure to tell exactly that to a beginner who gets
confused. Thanks for the tip.
>> And what should be done if the players disagree on who broke superko
>>first?
>
> Never heard of "referee"?
And the referee does what? Decides at random who broke the superko
rule? Decides at random who loses the game? Voids the game? What?
Compare that to *not* having superko: After a while the players will
notice that the thing is repeating and going nowhere, at which point
they either play elsewhere or they end the game. Is this more unfair
than a referee deciding at random who wins or voiding the game?
Robert Jasiek
wrote:
> Simplicity of rules cannot be the only principle.
Yes, and that I did say.
>Superko (compared to not having superko) makes the game needlessly complicated.
Simplicity of the rules text and strategic complexity are different
things.
> You seriously claim that in the about 3000 years of the history of Go
>it has NEVER occurred that two beginners have got confused about who
>broke the superko rule first?
I have claimed that complex cycles have never occurred. I have not
claimed that there would not have been players confused by easy
cycles.
>> The most complex thing was
>> a quadruple ko. That makes 2^4 = 16 states within the ko mouths. If
>> you cannot compare 16 positions with each other, then maybe you should
>> not play go...
>
> Yes, I'll make sure to tell exactly that to a beginner who gets
>confused. Thanks for the tip.
You are going to tell a beginner right away of triple ko, quadruple
ko, etc.? That must become a strong player quickly;)
> And the referee does what? Decides at random who broke the superko
>rule?
Decides factually who did. If he cannot, he declares a default jigo.
> Compare that to *not* having superko: After a while the players will
>notice that the thing is repeating and going nowhere, at which point
>they either play elsewhere or they end the game.
or they continue cycling forever.
>Is this more unfair
>than a referee deciding at random who wins or voiding the game?
A referee ought never to decide anything at random. So let us assume
he declares a default jigo.
More unfair? We have to distinguish two basic cases: The referee was
present or not. If he was, he should not make a wrong decision. Only
if he was not present, then he might have to consider default jigo.
This is very similar to a possibly illegally recaptured basic ko under
the bsaic ko rule. The difference in nature is so small that I would
not speak of "more" or "less unfair".
--
robert jasiek
Simon Goss
>> Rule sets that allow suicide usually don't allow single-stone suicide,
>> since single-stone suicide allows a player to obstruct the completion of
>> the game by giving up the turn without either playing or passing.
>
>
>Superko rules already prevent that
PSK only.
Simon Goss
>> And what should be done if the players disagree on who broke superko
>>first?
Sounds like a good rule question.
Robert Jasiek replied:
>Never heard of "referee"?
Uh, isn't he the guy who enforces the rules? I'm surprised at you,
Robert.
ian
Simon Goss wrote:
> Juha Nieminen wrote:
> >> And what should be done if the players disagree on who broke superko
> >>first?
>
> Sounds like a good rule question.
>
> Robert Jasiek replied:
> >Never heard of "referee"?
>
> Uh, isn't he the guy who enforces the rules? I'm surprised at you,
> Robert.
>
are not allowed to bring this matter to the referee's attention
ourselves. So just be patient.
Can we have Incidental Super Ko as well?
ian
ian wrote:
> Simon Goss wrote:
> > Juha Nieminen wrote:
> > >> And what should be done if the players disagree on who broke superko
> > >>first?
> >
> > Sounds like a good rule question.
> >
> > Robert Jasiek replied:
> > >Never heard of "referee"?
> >
> > Uh, isn't he the guy who enforces the rules? I'm surprised at you,
> > Robert.
> >
> Note that the players involved have to ask the referee what to do, we
> are not allowed to bring this matter to the referee's attention
> ourselves.
Bah that's wrong isn't it :-<
ands...@gmail.com
move that does not lead to an infinite cycle. Obviously a player may
decide to deviate from the infinite cycle at any time (by passing for
instance). But provided that previous moves were the best possible (at
least in the eyes of the players), deviating from the cycle will lead
to an inferior result for the player who breaks the cycle, so that both
players will not want to break it. So theoretically infinite cycles
only appear if the players are willing to play this way, as they are
not obliged by the rules to play like that. So, the moves that
situational superko (or any other ko rule) declares illegal may not
lead to an infinite cycle if one of the players decides to deviate (at
the expense of losing the game as the case may be). Having said all
this, I am sure you did not mean this when you made your statement, as
I am sure you have an example that illustrates your assertion. I'd be
grateful if you could provide such an example here.
Juha Nieminen
> Could you please give an example when situational superko prohibits a
> move that does not lead to an infinite cycle.
http://warp.povusers.org/go/RulesElegance/diagram2.png
Black captures by playing on the squared intersection, white is
prohibited from recapturing, thus black lives. There's no infinite
cycle involved.
Robert Jasiek
wrote:
>http://warp.povusers.org/go/RulesElegance/diagram2.png
Also under a basic-2-move-ko-rule with a 2-pass-rule Black remains on
the board until the end of the alternation.
--
robert jasiek
Eduardo Lopez Herrero
Robert, sorry for the nitpicking, but I think "jigo" is not what you mean.
This word refers exclusively to the situation in which both players have the
same amount of points at the end of the game.
Eduardo
Robert Jasiek
<eduard...@gol.com> wrote:
>> Decides factually who did. If he cannot, he declares a default jigo.
>This word refers exclusively to the situation in which both players have the
>same amount of points at the end of the game.
"default jigo" is the technical term for the game tournament result
"0.5:0.5(!)" as declared by the referee's arbitrating ruling.
--
robert jasiek
Bob Hearn
ands...@gmail.com wrote:
> I think every go player
> is familiar with the concept of state (or situation) intuitively even
> if they don't have a name for it. In my post I simply gave a name for
> something that everyone knows already (I would have used the term
> situation but it didn't occur to me). I think it is a simple, natural,
> and above all useful concept.
Well, the "board state" would seem to be simply the configuration of
stones on the board - the position. The "game state" is much more than
that, as Robert pointed out. You can't have a rule that prohibits ever
repeating a game state, because unlike board states, there are an
infinite number of game states, so you can get nonterminating games.
E.g., with territory rules, number of prisoners is part of the game
state. For tournament purposes, time left is part of the game state.
More generally, the entire history of the game is part of the game
state, and of course the history changes every time you make a move.
So, the choice of including player to move and excluding other factors
is not really fundamental, but is motivated by more complex concerns
that are not at all obvious.
> So now, my point was that SSK rule could be formulated as simply as PSK
> rule if the term position is replaced with the term state. Obviously I
> needed to define this term first, so one could argue that I just hid
> the complexity of the SSK rule behind some extra definitions. This
> could be a valid objection but one has to look at whether the new term
> actually makes things simpler and easier to understand and formulate.
If you have to add a definition, how have you made it easier to
formulate?
Bob
Eduardo Lopez Herrero
> "default jigo" is the technical term for the game tournament result
> "0.5:0.5(!)" as declared by the referee's arbitrating ruling.
Who is responsible for the coinage of that term?
Is it too late to change it to something more logical, linguistically
speaking?
I don't see the need to misuse a Japanese word, when there is at least one
perfectly appropriate English word to describe that sort of result.
Eduardo
Robert Jasiek
<eduard...@gol.com> wrote:
>Who is responsible for the coinage of that term?
I do not recall, except that it was not I.
>there is at least one
>perfectly appropriate English word to describe that sort of result.
Which? "default tie"?
--
robert jasiek
Eduardo Lopez Herrero
(R. Jasiek)
Either that, or "default draw", would do better than "default jigo", IMO.
I'd like to hear the opinion of a native speaker of English, though.
Eduardo
-
>> Which? "default tie"?
"Eduardo Lopez Herrero" <eduard...@gol.com> wrote:
> Either that, or "default draw", would do better than "default jigo", IMO.
> I'd like to hear the opinion of a native speaker of English, though.
After suffering disbarment today from "native speaker" status,
I'm not certain I should rise to the occasion. Two questions occur:
(a) why it should be suitable in cases of severe contest to simply
walk away from the game and declare a draw, especially when it
may be apparent that one player had obtained a favorable position
over the other, all things considering, (b) if the draw declaration is
unavoidable then it should not matter too much how one dresses
it up elegantly with favorite stock-in-trade technical phrases.
Most would be happier arriving at a "default" consideration
which could resort to fair-minded resolution of the rules dispute,
declaring why one color was winner and another color was loser.
This having been said, in Chess it may seem unsatisfying to obtain
a draw, unless of course one was matched against a much stronger
player and managed to avoid the sequences leading to checkmate.
So the party desirous of draw is actually the weaker player, and to
declare something as "default draw" is to assume that such things in
gameplay should be tactically dominated by weaker player mentality.
- regards
- jb
----------------------------------------------------------------------
A repunit prime is a repunit that is also a prime number.
http://en.wikipedia.org/wiki/Repunit
----------------------------------------------------------------------
An absolute prime cannot contain all four of the digits 1, 3, 7 and 9.
http://www.mathsyear2000.org/xplusyfiles/home/issue_5/numbers_with_names_absolute_prime_numbers/
----------------------------------------------------------------------
Simon Goss
I think "default draw" is best. "Tie" really means equal score.
Chris Dams
On Mon, 2005-10-10 at 22:51 +0000, - wrote:
> I trust that you won't hold me to my any previous standards. :-)
You can be sure of that.
> Earlier we had the discussion on "`Chinese' vs `Japanese' scoring"
> where Mr. Jasiek had asserted that there is no "remove all the ko
> threats rule" in order to assess the status of ambiguous positions:
Yes, I had read that discussion and suspect that Robert Jasiek left the
discussion because you seemed too misguided on this issue and providing
education would probably take a lot of effort. Nevertheless, I will
attempt to provide some clarification.
(1) There can be unremovable ko-threats.
(2) The main text of J89 does not mention "ko-threats". Such can be
verified by a simple text-search. It is unclear to me why you seem to be
asserting otherwise. Are you confusing "rule" as it occurs in a rule-
text with the offering of "rules" that really are strategic advice?
(3) As in the confirmation of life and death after the game stops
recapturing in the same ko is prohibited unless a ko-pass occurs, what
business does the concept of a ko-threat have during confirmation
anyway?
> > (2) On a 9x9 board, this anomaly is already very rare. On a 19x19 board
> > it is even less probable.
>
>
> Would seem to me that any 9x9 problem could be recreated onto
> a 19x19 board, employing proper construction of walls and so forth.
Certainly. However, I was actually referring to the frequency with which
such an anomaly would occur in games. Is it really worthwhile adding
complexity to the rule-text when the phenomenon that one tries to
prevent is never going to occur in a game? This can only be decided on a
case-by-case basis, of course. The danger is that if all perceived
anomalies result in additions to rule-texts, we might end up with
something as complicated as J2003.
> > (4) If every perceived anomaly is going to result in an addition to the
> > rules, we might as well give up the quest for applicable and clear
> > rules.
>
>
> Is a large glass window "less clear" than a small glass window?
> Japanese Rules cases may be handled by first filling ko threats.
All analogies are false, so let me offer another one ;-). Is a thick
window less clear than a thin one? If they are made of the same kind of
glass, certainly.
About filling ko-threats, I already remarked that unremovable ko-threats
do occur.
> > This is very unclear. Sounds great, though. Maybe you should go
> > work for the UN. You'll be great at writing speeches that sound
> > beautiful, yet accomplish nothing.
>
> Let's not criticize the UN for accomplishing nothing. At least
> the UN is not marching around into all countries with a UN military.
I grant you this one.
Best,
Chris
Chris Dams
On Mon, 2005-10-10 at 22:51 +0000, - wrote:
> I trust that you won't hold me to my any previous standards. :-)
You can be sure of that.
> Earlier we had the discussion on "`Chinese' vs `Japanese' scoring"
> where Mr. Jasiek had asserted that there is no "remove all the ko
> threats rule" in order to assess the status of ambiguous positions:
Yes, I had read that discussion and suspect that Robert Jasiek left the
discussion because you seemed too misguided on this issue and providing
education would probably take a lot of effort. Nevertheless, I will
attempt to provide some clarification.
(1) There can be unremovable ko-threats.
(2) The main text of J89 does not mention "ko-threats". Such can be
verified by a simple text-search. It is unclear to me why you seem to be
asserting otherwise. Are you confusing "rule" as it occurs in a rule-
text with the offering of "rules" that really are strategic advice?
(3) As in the confirmation of life and death after the game stops
recapturing in the same ko is prohibited unless a ko-pass occurs, what
business does the concept of a ko-threat have during confirmation
anyway?
> > (2) On a 9x9 board, this anomaly is already very rare. On a 19x19 board
> > it is even less probable.
>
>
> Would seem to me that any 9x9 problem could be recreated onto
> a 19x19 board, employing proper construction of walls and so forth.
Certainly. However, I was actually referring to the frequency with which
such an anomaly would occur in games. Is it really worthwhile adding
complexity to the rule-text when the phenomenon that one tries to
prevent is never going to occur in a game? This can only be decided on a
case-by-case basis, of course. The danger is that if all perceived
anomalies result in additions to rule-texts, we might end up with
something as complicated as J2003.
> > (4) If every perceived anomaly is going to result in an addition to the
> > rules, we might as well give up the quest for applicable and clear
> > rules.
>
>
> Is a large glass window "less clear" than a small glass window?
> Japanese Rules cases may be handled by first filling ko threats.
All analogies are false, so let me offer another one ;-). Is a thick
window less clear than a thin one? If they are made of the same kind of
glass, certainly.
About filling ko-threats, I already remarked that unremovable ko-threats
do occur.
> > This is very unclear. Sounds great, though. Maybe you should go
> > work for the UN. You'll be great at writing speeches that sound
> > beautiful, yet accomplish nothing.
>
> Let's not criticize the UN for accomplishing nothing. At least
> the UN is not marching around into all countries with a UN military.
I grant you this one.
Best,
Chris
-
> "-" wrote:
>> Earlier we had the discussion on "`Chinese' vs `Japanese' scoring"
>> where Mr. Jasiek had asserted that there is no "remove all the ko
>> threats rule" in order to assess the status of ambiguous positions:
Chris Dams <C.D...@science.ru.nl> wrote:
> Yes, I had read that discussion and suspect that Robert Jasiek left
> the discussion because you seemed too misguided on this issue
> and providing education would probably take a lot of effort.
Which then appears to be a conclusion in "bad faith" because
Mr. Jasiek DOES expend considerable effort with wrong directions.
Would that be a European Phenomenon, basically, to let things
slide without engaging in corrective dialogue? Perhaps this may
be due to the fact that Europeans do not grasp "We the people..."
> Nevertheless, I will attempt to provide some clarification.
> (1) There can be unremovable ko-threats.
Example? Why not play after "ko pass"?
> (2) The main text of J89 does not mention "ko-threats". Such can
> be verified by a simple text-search. It is unclear to me why you
> seem to be asserting otherwise. Are you confusing "rule" as it
> occurs in a rule-text with the offering of "rules" that really are
> strategic advice?
The rule is implied, too simple for the designation of strategic
advice. A general consensus exists that J1989 is for pros, so it
would not be asking too much to observe how pros score games.
I also queried some stronger Go Players who seemd to indicate
that Mr. Jasiek was incorrect when he had asserted that the phase
for "confirmation and agreement" did not include a "concept of
removing ko threats." Could be a language translation issue...
> (3) As in the confirmation of life and death after the game stops
> recapturing in the same ko is prohibited unless a ko-pass occurs,
> what business does the concept of a ko-threat have during
> confirmation anyway?
Let us agree that kos may be filled. You have a counter-example?
For scoring corner bent-4 under hypothetical play first fill all ko
threats and then capture the bent-4. This demonstrates that bent-4 is
dead. Now refer to the board prior to hypothetical play. The bent-4
position has already been shown as dead under hypothetical play.
In the present discussion it does not suffice merely to refer to the
discrepancy between "SSK vs. PSK." We need also the concept
of "game state" to consider situations and/or positions, since the
notion of "position" can also reference "a mass noun" coordinated
with advantage/disadvantage, result and game outcome. We may
measure a quantity called "score change" by observing the tallies
of respective prisoners among putatively identical "board states."
If the difference in prisoner count is the same -and- an identical
"board state" has been reached then we may apply the "SSK or
PSK" determination. If an illegal position was reached, under
either interpretation, then the move must be retracted. Under
some rulesets the move may be replayed, or it may be forfeit,
or a stone representing that move is handed to the opponent.
The notion that we may embrace rulesets which purport to pay
no attention to prisoner accounting, is miscast and utterly false.
The fact of having rulesets without prisoner accounting is due
to tournament convenience should prisoner accounting be lost.
The last consideration would concern somebody who kept
pitching losing stones into a dissolving ko, racking up a huge
prisoner debt, while attempting to run out opponent clock.
Such behavior would be addressed under other etiquette rules.
> ... Is it really worthwhile adding complexity to the rule-text when the
> phenomenon that one tries to prevent is never going to occur in a game?
Bear in mind that a rules-text applies to circumstantial context.
The type of "rules-text implementation" for a computer would be
radically different from readable "rules-texts" for human beings.
The recent paper presented by Mr. Tromp and Mr. Farneback
( at http://homepages.cwi.nl/~tromp/go/gostate.ps ) describes the
Game of Go succinctly, in one paragraph, using 16 lines of text.
If you don't have a postscript reader capability then you can take
the URL and punch it into the "browse/open" window of an online
converter such as http://www.ps2pdf.com/convert/convert.cgi .
The rules-text available to a tournament director might need
to be more extensive than the rules-text sought by casual players
(other than Mr. Jasiek) when they show up for tournament days.
> This can only be decided on a case-by-case basis, of course.
No, I don't believe it needs to devolve to that. Computers
can score positions if they are provided with enough rules-text.
> The danger is that if all perceived anomalies result in additions to
> rule-texts, we might end up with something as complicated as J2003.
Was J2003 adopted? If it's necessary for players to think that
way, then why not indulge in a stretch of the imagination? Most
players, preferring to be casual about rules, can get by just fine.
> About filling ko-threats, I already remarked that unremovable
> ko-threats do occur.
Were they unremovable with that "no score change" consideration
I had mentioned, _prerequisite_ for considering "SSK vs. PSK" ?
- regards
- jb
-------------------------------------------------------------------
Overview of "Mathematician's Secret Room"
-- Challenges to the Unsolved Problems in Number Theory --
http://www.asahi-net.or.jp/~KC2H-MSM/mathland/overview.htm
-------------------------------------------------------------------
Clive Hunt
"Simon Goss" <si...@gosoft.demon.co.uk> wrote in message
news:tHk0gVAg...@gosoft.demon.co.uk...
...as in cricket ( a native English game ), where "tie" refers to the (rare)
situation whereby the game has met all conditions for satisfactory
completion, but the scores come out identical. "draw" is the deplorably
frequent situation whereby a satisfactory completion has not been possible.
There is no recourse to God, except indirectly through His agency, The
Weather, or to a statistical model of Himself by Duckworth & Lewis.
Robert will point out that attaching the label "unsatisfactory" [negation of
satisfactory] to a draw brought about by several days of rain, is "matter of
opinion rather than logic".
Clive N.E.S.
Perth WA
Robert Jasiek
"making the pairing".
--
robert jasiek
ian
Robert Jasiek wrote:
> I do not like "default draw" because "draw" is also associated with
> "making the pairing".
>
I don't see what's so wrong about default jigo, surely it's indicating
both have not scored any points, since the scoring phase was never
reached. In that sense they finished on the same score.
You'll have to make up a new word for it at this rate.
Eduardo Lopez Herrero
> "making the pairing".
"Jigo" is also associated with "ending the game with equal points for both
sides". I'd say that confusion is more likely to occur with this term,
because a jigo is sometimes a draw. The distinction between the two meanings
of "draw", on the other hand, should be obvious from context.
Eduardo
Chris Dams
On Wed, 2005-10-12 at 22:06 +0000, - wrote:
> Chris Dams <C.D...@science.ru.nl> wrote:
> > Yes, I had read that discussion and suspect that Robert Jasiek left
> > the discussion because you seemed too misguided on this issue
> > and providing education would probably take a lot of effort.
>
>
> Which then appears to be a conclusion in "bad faith" because
> Mr. Jasiek DOES expend considerable effort with wrong directions.
> Would that be a European Phenomenon, basically, to let things
> slide without engaging in corrective dialogue? Perhaps this may
> be due to the fact that Europeans do not grasp "We the people..."
What would this be? A generalization from one or fewer examples?
> > Nevertheless, I will attempt to provide some clarification.
> > (1) There can be unremovable ko-threats.
>
>
> Example? Why not play after "ko pass"?
Two examples can be found at http://senseis.xmp.net/?
UnremovableKoThreat . What do you mean by "Why not play after 'ko
pass'?"?
> > (2) The main text of J89 does not mention "ko-threats". Such can
> > be verified by a simple text-search. It is unclear to me why you
> > seem to be asserting otherwise. Are you confusing "rule" as it
> > occurs in a rule-text with the offering of "rules" that really are
> > strategic advice?
>
>
> The rule is implied, too simple for the designation of strategic
> advice. A general consensus exists that J1989 is for pros, so it
> would not be asking too much to observe how pros score games.
> I also queried some stronger Go Players who seemd to indicate
> that Mr. Jasiek was incorrect when he had asserted that the phase
> for "confirmation and agreement" did not include a "concept of
> removing ko threats." Could be a language translation issue...
If it is implied in a way "too simple for the designation of strategic
advice" then surely it will not be too much to ask how this implication
is read in J1989. The commentary seems to read in a different way as it
says "Even if a player has unlimited ko threats (in a double-ko seki for
example), he cannot use them to recapture a ko.". If these "unlimited ko
threats" cannot be used, why should they be filled? Perhaps your
"stronger Go Players" could also provide some arguments?
> > (3) As in the confirmation of life and death after the game stops
> > recapturing in the same ko is prohibited unless a ko-pass occurs,
> > what business does the concept of a ko-threat have during
> > confirmation anyway?
>
> Let us agree that kos may be filled. You have a counter-example?
> For scoring corner bent-4 under hypothetical play first fill all ko
> threats and then capture the bent-4. This demonstrates that bent-4 is
> dead. Now refer to the board prior to hypothetical play. The bent-4
> position has already been shown as dead under hypothetical play.
Except that this filling of threats is useless. A ko-threat is used to
be able to capture again in the ko after the threat is answered.
However, in hypothetical play one may not recapture unless a ko-pass has
occurred. So why would anyone want to play a ko-threat in hypothetical
play if it is not going to help winning the ko. If ko-threats are not
helpful in winning a ko in hypothetical analysis, what purpose would
filling them have?
> The rules-text available to a tournament director might need
> to be more extensive than the rules-text sought by casual players
> (other than Mr. Jasiek) when they show up for tournament days.
IMO rules of that kind should disappear and something simpler should be
adopted.
> > The danger is that if all perceived anomalies result in additions to
> > rule-texts, we might end up with something as complicated as J2003.
>
> Was J2003 adopted? If it's necessary for players to think that
> way, then why not indulge in a stretch of the imagination? Most
> players, preferring to be casual about rules, can get by just fine.
I seem to remember having seen a computer go project that choose to
adopt them. Besides, if you are asked to score a position according to
J89, the only way to do so may well be by actually using J2003, as J89
is not even able reproduce its own examples while J2003 gets all of them
correct.
Best,
Chris
-
>> Chris Dams <C.D...@science.ru.nl> wrote:
>> > Yes, I had read that discussion and suspect that Robert Jasiek left
>> > the discussion because you seemed too misguided on this issue
>> > and providing education would probably take a lot of effort.
> "-" wrote:
>> Which then appears to be a conclusion in "bad faith" because
>> Mr. Jasiek DOES expend considerable effort with wrong directions.
>> Would that be a European Phenomenon, basically, to let things
>> slide without engaging in corrective dialogue? Perhaps this may
>> be due to the fact that Europeans do not grasp "We the people..."
Chris Dams <C.D...@science.ru.nl> wrote:
> What would this be? A generalization from one or fewer examples?
It's the private assessment of a good number of American lawyers
who may also work with issues in International Law, and simply not
given much airplay owing to the problem of beating a dead horse.
The mechanisms for the European Union, for example, are being
imposed from above, or externally, not really within the sovereign
jurisdiction of Europeans. A few earlier newsgroup examples:
Results 1 - 6 of 6 for "we the people" group:rec.games.go (0.26 seconds)
http://groups.google.com/groups?as_q=&num=100&scoring=d&hl=en&as_epq=we+the+people&as_oq=&as_eq=&as_ugroup=rec.games.go&as_usubject=&as_uauthors=&lr=&as_drrb=q&as_qdr=&as_mind=1&as_minm=1&as_miny=1981&as_maxd=14&as_maxm=10&as_maxy=2005&safe=off
>>> Nevertheless, I will attempt to provide some clarification.
>>> (1) There can be unremovable ko-threats.
>> Example? Why not play after "ko pass"?
> Two examples can be found at
> http://senseis.xmp.net/?UnremovableKoThreat . What do you mean
> by "Why not play after 'kopass'?"?
These have been termed "unremovable" because they describe
_seki_. If both players have filled their _ko_ threats, leaving only
bent-4 and a _seki_ on the board, how would you utilize play into
this _seki_ in an effort to alter the status of the bent-4 shape?
I question the characterization of "unremovable" here because
play into any _seki_ is a legal (and game possible) move. Just
goes to show what sorts of "trouble" can be encountered when
a community webpage encyclopedia project goes awry. Who
is `unkx80' -- a graduate computer student from Singapore?
The term "unremovable" has been qualified by "without also
executing the threat." It is not _unconditinally_ "unremovable"
so what are we talking about, when referencing the conditions?
The definitive analysis proceeds instead from examination of all
-legal- positions in the game.
The "ko pass" allows a player to defer moving without entering
one of the passes in the "consecutive pass" rule which signifies
game stop. After "ko pass" and "opponent pass" play may again
revert to the ko without a risk of repeating "board position" (owing
to the "ko pass"). The ko is recaptured, the opponent says "ko
pass" and then other ko material is captured, resolving the ko.
>>> (2) The main text of J89 does not mention "ko-threats". Such can
>>> be verified by a simple text-search. It is unclear to me why you
>>> seem to be asserting otherwise. Are you confusing "rule" as it
>>> occurs in a rule-text with the offering of "rules" that really are
>>> strategic advice?
>> The rule is implied, too simple for the designation of strategic
>> advice. A general consensus exists that J1989 is for pros, so it
>> would not be asking too much to observe how pros score games.
>> I also queried some stronger Go Players who seemd to indicate
>> that Mr. Jasiek was incorrect when he had asserted that the phase
>> for "confirmation and agreement" did not include a "concept of
>> removing ko threats." Could be a language translation issue...
> If it is implied in a way "too simple for the designation of strategic
> advice" then surely it will not be too much to ask how this implication
> is read in J1989. The commentary seems to read in a different way as
> it says "Even if a player has unlimited ko threats (in a double-ko seki
> for example), he cannot use them to recapture a ko."
That statement applies to the opposite player. In fact it's another
way of saying the same thing: even "unlimited ko threats" could be
first filled so that they don't exist anymore. They are thereby made
unavailable as material in contest for the ko recapture.
> If these "unlimited ko threats" cannot be used, why should they be filled?
> Perhaps your "stronger Go Players" could also provide some arguments?
It doesn't matter whether or not they become filled, since players
have declared game stop. They might need to be filled in order to
settle and determine game score. There are two sides to the proof
for establishing non-seki status of groups: (a) it is established that
one of the players cannot kill stones, even with ulimited ko threats;
(b) it is established that the other player can survive by showing
how remaining ko threats may first be filled, and ko prevented.
Robert Jasiek does not prefer to have "ensuring" accomplished
by both sides in the (a) & (b) form. His J2003 describes non-seki
groups in terms of characteristics entirely within the analysis by
one side or the other, not using some dual proof from both sides.
J2003 offers advantages for machine-implementation of scoring.
>>> (3) As in the confirmation of life and death after the game stops
>>> recapturing in the same ko is prohibited unless a ko-pass occurs,
>>> what business does the concept of a ko-threat have during
>>> confirmation anyway?
>> Let us agree that kos may be filled. You have a counter-example?
>> For scoring corner bent-4 under hypothetical play first fill all ko
>> threats and then capture the bent-4. This demonstrates that bent-4 is
>> dead. Now refer to the board prior to hypothetical play. The bent-4
>> position has already been shown as dead under hypothetical play.
> Except that this filling of threats is useless. A ko-threat is used to
> be able to capture again in the ko after the threat is answered.
> However, in hypothetical play one may not recapture unless a ko-pass
> has occurred. So why would anyone want to play a ko-threat in
> hypothetical play if it is not going to help winning the ko.
The other guy might want to play a ko threat for _defending_ the ko.
> If ko-threats are not helpful in winning a ko in hypothetical analysis,
> what purpose would filling them have?
They're first filled to prevent the other guy from _defending_ the ko.
Let's also remember that the play of "useless moves" is not being
penalized once game stop has been declared. All that interests
them, at that point, is status of groups, then reverting to a position
at game stop, and then scoring the game with that position.
>> The rules-text available to a tournament director might need
>> to be more extensive than the rules-text sought by casual players
>> (other than Mr. Jasiek) when they show up for tournament days.
> IMO rules of that kind should disappear and something simpler
> should be adopted.
Tournament directors and referees have opinions, and players
have opinions. They are not mutually exclusive. Versions of the
same ruleset may be expressed for player readability and also for
the reference readability of tournament directors and referees.
>>> The danger is that if all perceived anomalies result in additions to
>>> rule-texts, we might end up with something as complicated as J2003.
>> Was J2003 adopted? If it's necessary for players to think that
>> way, then why not indulge in a stretch of the imagination? Most
>> players, preferring to be casual about rules, can get by just fine.
> I seem to remember having seen a computer go project that choose
> to adopt them. Besides, if you are asked to score a position according
> to J89, the only way to do so may well be by actually using J2003, as
> J89 is not even able reproduce its own examples while J2003 gets
> all of them correct.
Well, here's the dilemma: Robert Jasiek has published and
made copyright something called `J2003' which has the label of
"Japanese" Rules, though I do not recall whether any Japanese
organization has formally adopted them, even when provided with
the opportunity while Mr. Jasiek supplied a presentation, in person.
Until they do, it seems to me that the "J" in J2003 refers to Jasiek
and not to Japanese.
I do not agree that the "only way to do so" must use J2003.
J1989 may work quite well but not be comprehensible to those
who do not grasp the concept of first filling ko threats. Mr. Jasiek
demonstrated recently that he does not grasp this simple concept.
- regards
- jb
------------------------------------------------------------------
Dead White Women: Which Ones Matter Most?
http://www.strike-the-root.com/52/herman/herman17.html
------------------------------------------------------------------
Robert Jasiek
>> as J89 is not even able reproduce its own examples while J2003 gets
> all of them correct.
Did anybody claim that J2003 would be the only way? I have claimed
that J2003 has been the first way to do so ever. E.g., here is a
second way: replace the hypothetical-ko rule of J2003
>>
[...]
{hypothetical-ko:} between a player's ko-capture on one intersection
of a ko and the next ko-pass by him or his opponent does not include
an opponent's ko-capture on the other intersection.
<<
by the following:
>>
[...]
{hypothetical-ko:} between a player's ko-capture on one intersection
of a ko and the next ko-pass by him or his opponent does not include
an opponent's ko-capture on an intersection of the ko.
<<
Since Japanese professionals have never studied examples where these
two versions differ in practice, either might be used in a model of
Japanese rules.
> J1989 may work quite well but not be comprehensible to those
> who do not grasp the concept of first filling ko threats.
> Jasiek demonstrated recently that he does not grasp this simple concept.
You are going to teach us what "first filling ko threats" is.
--
robert jasiek
-
>> Chris Dams <C.D...@science.ru.nl> wrote:
>>> ... if you are asked to score a position according to J89,
> jazze...@hotmail.com (-) wrote:
>> I do not agree that the "only way to do so" must use J2003.
Robert Jasiek <jas...@snafu.de> wrote:
> Did anybody claim that J2003 would be the only way?
Chris Dams said "...may well be..."
> ... Since Japanese professionals have never studied examples
> where these two versions differ in practice, either might be
> used in a model of Japanese rules.
Seems to be a non-sequitur. To qualify as a proper model,
blessed by the Japanese, wouldn't Japanese professionals need
to study the examples?
>> J1989 may work quite well but not be comprehensible to those
>> who do not grasp the concept of first filling ko threats. Mr. Jasiek
>> demonstrated recently that he does not grasp this simple concept.
> You are going to teach us what "first filling ko threats" is.
The complementary proposition being "You are going to learn
what `first filling ko threats' is" ? Remember that game stop can be
re-instantiated, as a reference, if a preliminary attempt in hypothetical
play does not produce the sought-after hypothetical strategy. Some ko
threat, previously unanticipated, could be filled for next iteration of
the demonstrary hypothetical sequence. Not all ko threats need be
filled, but only those which a defender of ko capture might apply by
way of retaliation. If one was left off, then restart and fill anyway.
- regards
- jb
--------------------------------------------------------------------
Amazing UFO Taped By Security Cam In Mexico
http://www.rense.com/general67/amaziufo.htm
--------------------------------------------------------------------
Chris Dams
On Fri, 2005-10-14 at 20:01 +0000, - wrote:
> These have been termed "unremovable" because they describe
> _seki_. If both players have filled their _ko_ threats, leaving only
> bent-4 and a _seki_ on the board, how would you utilize play into
> this _seki_ in an effort to alter the status of the bent-4 shape?
Do you even realize what you are asking? As you may or may not have
noticed, I have been arguing in this very thread that making ko-threats
of any kind (be they removable or not) is not useful during hypothetical
analysis. Would it in this context not be a bit strange---to say the
very least---if I would now continue by showing how a ko-threat (of the
unremovable kind in this case) would be used to alter the status of a
bent-4 in the corner? Why don't you take upon yourself the task of
showing how the outcome of a bent-4 in the corner can be altered in the
case where a player fails to fill ko-threats? Such would prove your
point.
> The "ko pass" allows a player to defer moving without entering
> one of the passes in the "consecutive pass" rule which signifies
> game stop. After "ko pass" and "opponent pass" play may again
> revert to the ko without a risk of repeating "board position" (owing
> to the "ko pass"). The ko is recaptured, the opponent says "ko
> pass" and then other ko material is captured, resolving the ko.
Yes, this sounds like something that could happen in a hypothetical
sequence, however, it does not go anywhere near to proving your point.
In particular I note that no ko-threats were played or filled during the
sequence you outlined. Maybe you should now attempt to show a
hypothetical sequence in which either the filling or the playing of a
ko-threat serves a purpose.
> >>> (2) The main text of J89 does not mention "ko-threats". Such can
> >>> be verified by a simple text-search. It is unclear to me why you
> >>> seem to be asserting otherwise. Are you confusing "rule" as it
> >>> occurs in a rule-text with the offering of "rules" that really are
> >>> strategic advice?
>
> >> The rule is implied, too simple for the designation of strategic
> >> advice. A general consensus exists that J1989 is for pros, so it
> >> would not be asking too much to observe how pros score games.
> >> I also queried some stronger Go Players who seemd to indicate
> >> that Mr. Jasiek was incorrect when he had asserted that the phase
> >> for "confirmation and agreement" did not include a "concept of
> >> removing ko threats." Could be a language translation issue...
>
> > If it is implied in a way "too simple for the designation of strategic
> > advice" then surely it will not be too much to ask how this implication
> > is read in J1989. The commentary seems to read in a different way as
> > it says "Even if a player has unlimited ko threats (in a double-ko seki
> > for example), he cannot use them to recapture a ko."
>
>
> That statement applies to the opposite player. In fact it's another
> way of saying the same thing: even "unlimited ko threats" could be
> first filled so that they don't exist anymore. They are thereby made
> unavailable as material in contest for the ko recapture.
The example of a double ko-seki was offered in the commentary. How would
one precisely fill the "unlimited ko threats" in a double ko-seki?
Furthermore, I cannot help to notice that you fail to answer the
question how the implication that filling ko-threats is useful in
hypothetical sequences is read in the text of J1989.
> The other guy might want to play a ko threat for _defending_ the ko.
What do you mean by "defending the ko"? A simple example would suffice.
> I do not agree that the "only way to do so" must use J2003.
> J1989 may work quite well but not be comprehensible to those
> who do not grasp the concept of first filling ko threats.
Well, be my guest then. The examples I was referring to are "Life-and-
Death Example 16" and "Life-and-Death Example 17". Show how those who do
grasp the concept of first filling ko threats are able to use J1989 to
obtain the described result. In particular, please choose one of these
two examples and we will attempt to play a hypothetical sequence as
described in J1989 in which capture of the rightmost white string that
occurs in either example is all I ask of you.
> Mr. Jasiek demonstrated recently that he does not grasp this simple concept.
Let us not jump to conclusions before you have proved your point.
Prediction: you will not be able to.
Best,
Chris
-
> "-" wrote:
>> These have been termed "unremovable" because they describe
>> _seki_. If both players have filled their _ko_ threats, leaving only
>> bent-4 and a _seki_ on the board, how would you utilize play into
>> this _seki_ in an effort to alter the status of the bent-4 shape?
Chris Dams <C.D...@science.ru.nl> wrote:
> Do you even realize what you are asking? As you may or may not have
> noticed, I have been arguing in this very thread that making ko-threats
> of any kind (be they removable or not) is not useful during hypothetical
> analysis.
Yes, which is why (during hypothetical play) a player might elect
to fill ko-threats with what you could term "useless moves" for good
measure, so as to avoid any argument concerning whether such
ko-threats could be brought into play and disrupt contest over status.
> Would it in this context not be a bit strange---to say the very least---if
> I would now continue by showing how a ko-threat (of the unremovable
> kind in this case) would be used to alter the status of a bent-4 in the
> corner? Why don't you take upon yourself the task of showing how the
> outcome of a bent-4 in the corner can be altered in the case where a
> player fails to fill ko-threats? Such would prove your point.
The convention, during hypothetical play, does not consider a
corner bent-4 alive owing to analysis obtained by hypothetical play.
However, prior to game stop the corner bent-4 presents potential ko.
If the players wish to press corner bent-4 into ko material status then
under J1989 they would do so prior to declaring game stop. What
irks some rules mavens is the apparent "change in status" between
the corner bent-4 prior to game stop and after game stop. Undecided
prior to game stop but declared unconditionally dead after game stop.
If the players do not score correctly during hypothetical play, it
is conceivable that contesting ko threats would not first be filled prior
to making the final approach at demonstrating death of corner bent-4.
It is conceivable that players would forget the stipulations of J1989 on
the corner bent-4 status and score the game "incorrectly" in context
to the J1989 stipulations. It is conceivable that a "retry" back to the
start position of game stop would be disallowed, even though the
phase is termed "confirmation -and- agreement." However we are
instead reaching this conclusion that the corner bent-4 status was
being "altered" perhaps under tournament play conditions where a
determination of game score and result is handled incorrectly by the
players. As I've pointed out earlier, that is a very different scenario
than the situation involving the tournament director and referees,
with all of the third-party bystanders. There is a proper method for
showing why corner bent-4 is dead, and that proper method would
first fill the relevant ko-threats that an opponent might spring as a
form of contesting ko material.
>> The "ko pass" allows a player to defer moving without entering
>> one of the passes in the "consecutive pass" rule which signifies
>> game stop. After "ko pass" and "opponent pass" play may again
>> revert to the ko without a risk of repeating "board position" (owing
>> to the "ko pass"). The ko is recaptured, the opponent says "ko
>> pass" and then other ko material is captured, resolving the ko.
> Yes, this sounds like something that could happen in a hypothetical
> sequence, however, it does not go anywhere near to proving your point.
"Ko pass" may occur during actual play AND hypothetical sequences.
Perhaps if you state what you think my point is then I can decide
whether you are correctly processing the descriptions of my point.
> In particular I note that no ko-threats were played or filled during the
> sequence you outlined. Maybe you should now attempt to show a
> hypothetical sequence in which either the filling or the playing of a
> ko-threat serves a purpose.
Let's say that stones in the corner bent-4 are Black, so that
White must demonstrate their capture during hypothetical play.
Unfortunately, the players had reached game stop while leaving
additional ko threat material on the board, and Black has many
unplayed ko threats. White first plays in a manner which resolves
the threats Black could make, before capturing the two-step ko in
a corner. If White was not already dead in the process then White
must have an extra point in the shape which allows White's action.
>>> If it is implied in a way "too simple for the designation of strategic
>>> advice" then surely it will not be too much to ask how this implication
>>> is read in J1989. The commentary seems to read in a different way as
>>> it says "Even if a player has unlimited ko threats (in a double-ko seki
>>> for example), he cannot use them to recapture a ko."
>> That statement applies to the opposite player. In fact it's another
>> way of saying the same thing: even "unlimited ko threats" could be
>> first filled so that they don't exist anymore. They are thereby made
>> unavailable as material in contest for the ko recapture.
> The example of a double ko-seki was offered in the commentary. How
> would one precisely fill the "unlimited ko threats" in a double ko-seki?
Well, I don't agree that a double ko-seki offers "unlimited ko threats."
We were discussing this in context to the SSK vs PSK Subject: line.
Once reaching a non-repeatable position having identical score to
a prior position the game has gone too far and must back up once.
But J1989 might offer "double-ko seki" as an example of "unlimited
ko threats" (where there is no SSK or PSK rule) because in practice
we do not have an infinite Go Board. The upshot is that once game
stop occurs then there is no fixed limit upon the number of ko threats
which may be filled prior to demonstrating status of a certain group.
If you play out sequences with a corner bent-4 and a double ko seki
on the board, then you reach the SSK or PSK nexus where the ko
defender cannot recreate the prior position and must offer ko pass.
> Furthermore, I cannot help to notice that you fail to answer the
> question how the implication that filling ko-threats is useful in
> hypothetical sequences is read in the text of J1989.
Filling ko threats is -optionally- useful as a means for clarification
between the players during "confirmation and agreement." The
technical meaning of the English term "confirmation" connotes a
high degree of lattitude and good faith effort to get at the truth.
>> The other guy might want to play a ko threat for _defending_ the ko.
> What do you mean by "defending the ko"? A simple example would suffice.
The ko defender wants to keep the ko operational, while the ko
capturer wants to settle the ko by eventually capturing it. In this case
the corner bent-4 turns into koloser for Black and komaster for White.
>> I do not agree that the "only way to do so" must use J2003.
>> J1989 may work quite well but not be comprehensible to those
>> who do not grasp the concept of first filling ko threats.
> Well, be my guest then. The examples I was referring to are "Life-and-
> Death Example 16" and "Life-and-Death Example 17". Show how those
> who do grasp the concept of first filling ko threats are able to use J1989
> to obtain the described result. In particular, please choose one of these
> two examples and we will attempt to play a hypothetical sequence as
> described in J1989 in which capture of the rightmost white string that
> occurs in either example is all I ask of you.
I infer that we are referencing the webpage, or its equivalent,
at http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html . One's
early impression is the double ko seki, which suffices for "temporary
seki." We are given that groups with outside walls are independently
alive. There are no residual ko threats to speak of, and the double
ko seki is a wash altogether which eventually runs up against that
SSK vs PSK rule problem. The task for Example 16 is straightforward.
Black captures and White has no ko threat to play (pass or ko pass).
Black captures the next two stones and White may try recapture of
one. Black fills for recapturing White, and White has no ko threat
(other than the cycling within the double ko seki which then gets
back to another SSK vs PSK rule problem). Eventually Black gets
another seven stones and changes the status of the Black group
from "temporary seki" to "alive."
The task for Example 17 proceeds similarly but has a subtle
monkeywrench. The three stones "alive for black" do not form a
corner bent-4 shape during the sacrifice because White has no
outside liberty or extra eye. White must fill a point for capture and
then Black starts a corner ko inside the White shape which White
cannot win.
These Examples were not illustrative of pre-filling of multiple ko
threats, nor should it be likely that examples need to be provided.
>> Mr. Jasiek demonstrated recently that he does not grasp this simple concept.
> Let us not jump to conclusions before you have proved your point.
> Prediction: you will not be able to.
Hmmm. Something was unclear. I'll try posting it once more.
There, we were discussing resolution of the corner bent-4 status.
- regards
- jb
-----------------------------------------------------------------------
Newsgroups: rec.games.go
From: jazze...@hotmail.com (-)
Date: Fri, 30 Sep 2005 16:16:13 GMT
Subject: Re: "Chinese" vs "Japanese" scoring
>> Robert Jasiek <jas...@snafu.de> wrote:
>>> He was right in that since there is no "remove all
>>> the ko threats" rule...
> jazze...@hotmail.com (-) wrote:
>> Indeed there is: that's covered under hypothetical
>> play, or "confirmation and agreement" (J1989).
Robert Jasiek <jas...@snafu.de> wrote:
> That is not a concept of removing ko threats but a concept
> that lets ko-passes be the only available useful ko threats.
Surely it cannot be correct to disallow a player the
move which removes a ko threat during "confirmation
and agreement."
-----------------------------------------------------------------------
Chris Dams
On Sat, 2005-10-15 at 09:23 +0000, - wrote:
> Yes, which is why (during hypothetical play) a player might elect
> to fill ko-threats with what you could term "useless moves" for good
> measure, so as to avoid any argument concerning whether such
> ko-threats could be brought into play and disrupt contest over status.
Yes. However, in that case the opponent is apparently not playing using
J1989, so this issue actually has no bearing on the question whether it
is useful to fill ko-threats during hypothetical play .
> If the players do not score correctly during hypothetical play, it
> is conceivable that contesting ko threats would not first be filled prior
> to making the final approach at demonstrating death of corner bent-4.
> It is conceivable that players would forget the stipulations of J1989 on
> the corner bent-4 status and score the game "incorrectly" in context
> to the J1989 stipulations.
While all kinds of incorrectness are imaginable, I do not see how
filling ko-threats is going to help. After all, the one attempting to
use them will find them answered and after that the status of the ko's
with respect to being allowed or disallowed is simply unchanged. If
players are so weak they forget about the status of the involved ko's if
two moves are exchanged, they probably should not be playing using
J1989. If they do not know that recapturing is only allowed after a ko-
pass, they are not playing using J1989 and their behaviour is unrelated
to the question on the usefulness of filling ko-threats under J1989.
> It is conceivable that a "retry" back to the
> start position of game stop would be disallowed, even though the
> phase is termed "confirmation -and- agreement." However we are
> instead reaching this conclusion that the corner bent-4 status was
> being "altered" perhaps under tournament play conditions where a
> determination of game score and result is handled incorrectly by the
> players. As I've pointed out earlier, that is a very different scenario
> than the situation involving the tournament director and referees,
> with all of the third-party bystanders. There is a proper method for
> showing why corner bent-4 is dead, and that proper method would
> first fill the relevant ko-threats that an opponent might spring as a
> form of contesting ko material.
No it is not. If the bent-4 remains on the board together with one ko-
threat, let us say that we have the position
5 # #
4 . #
3 O # + ko threat for B.
2 O # #
1 O . #
a b c
with outer B-stones surrounded by W ones. A hypothetical sequence would
be Wb1, Ba4, Wa2, Ba1, Wb1 (ko-capture), B plays ko-threat, W answers, B
pass for ko (is not allowed to recapture), Wa1, and clearly the B-stones
are going to be captured. Every such a sequence is trivially transformed
into an equivalent one where no ko threat + corresponding answer is
played. Hence such a sequence and its transformed counterpart sever the
same purpose in hypothetical analysis. So why play the ko-threat if it
servers no purpose.
> We were discussing this in context to the SSK vs PSK Subject: line.
Yes, up to the point where you started asserting that "during
hypothetical play ('confirmation and agreement'), all ko threats may
first be filled before approaching the problem situation under
jurisdiction?", after which we also started discussing that this
statement is unhelpful/wrong/not applicable to any case whatsoever.
> > Furthermore, I cannot help to notice that you fail to answer the
> > question how the implication that filling ko-threats is useful in
> > hypothetical sequences is read in the text of J1989.
>
>
> Filling ko threats is -optionally- useful as a means for clarification
> between the players during "confirmation and agreement." The
> technical meaning of the English term "confirmation" connotes a
> high degree of lattitude and good faith effort to get at the truth.
So, now you are going to say that this "rule" is a matter of politeness?
I must say I am surprised and disappointed. Couldn't you have stated
this in a clearer way earlier in this thread? In that case I could have
just noted that jb asserts that filling ko-threats when performing
hypothetical analysis is polite and moved on as I am not interested in a
discussion about the politeness of filling ko-threats during
hypothetical analysis.
> >> The other guy might want to play a ko threat for _defending_ the ko.
>
> > What do you mean by "defending the ko"? A simple example would suffice.
>
>
> The ko defender wants to keep the ko operational, while the ko
> capturer wants to settle the ko by eventually capturing it. In this case
> the corner bent-4 turns into koloser for Black and komaster for White.
Ko-threats are not useful for that purpose either. The ko-capturer will
simply capture and fill the ko as ko-threats cannot prevent him form
doing this during hypothetical analysis.
Best,
Chris
-
> "-" wrote:
>> ...(during hypothetical play) a player might elect
>> to fill ko-threats with what you could term "useless moves" for
>> good measure, so as to avoid any argument concerning whether such
>> ko-threats could be brought into play and disrupt contest over status.
Chris Dams <C.D...@science.ru.nl> wrote:
> Yes. However, in that case the opponent is apparently not playing using
> J1989, so this issue actually has no bearing on the question whether it
> is useful to fill ko-threats during hypothetical play .
How did you reach -THAT- conclusion (not playing using J1989)? Are
you disallowing hypothetical play under "confirmation and agreement"?
>> If the players do not score correctly during hypothetical play, it is
>> conceivable that contesting ko threats would not first be filled prior
>> to making the final approach at demonstrating death of corner bent-4.
>> It is conceivable that players would forget the stipulations of J1989
>> on the corner bent-4 status and score the game "incorrectly" in context
>> to the J1989 stipulations.
> While all kinds of incorrectness are imaginable, I do not see how
> filling ko-threats is going to help. After all, the one attempting
> to use them will find them answered ...
Not necessarily. In my (corrected and augmented) diagrams further
below, note how the "corner bent-4 shape" occurs when the surrounding
stones of Black ("#") have zero, one, or two external liberties, here
designated at c3 & c4. They are dame points. After game stop either
player may object to the filling of dame for purposes of territory
counting. Suppose Black objects to White's proposal to fill c3,c4
and White is persuaded by Black's objection that these dame may not
be filled. In (A) Black retains four liberties. White sacrifices
four stones into a "corner bent-4 shape" and Black recaptures. Black
retains four liberties. The sequence follows as Wa2, Ba1, Wb1
(ko-capture) and Black still has three liberties. Black may surround
at a3 and White cannot make bent-3 for dead shape, owing to the "no
suicide" rule. In this case, even if White could suicide it would
still be Black's move. Not all ko threats are dame points, however.
If (B) one of the c3 or c4 liberties was filled prior to game stop,
Black is in a pickle. Once White ("O") takes the ko Black tries to
find a ko threat elsewhere. There is an on board decision: answer
the Black ko threat or risk not capturing the corner Black group.
In other words, the condition where players had entered game stop
without resolving residual ko on the board. Say White answers the
ko and Black retakes at Ba2. Now White needs to find a ko threat.
The J1989 rules which, for (B) or (C), award default capture of Black
upon corner bent-4, do so in the context of no residual ko threats on
the remainder of the board. Because dame fill can be protested by
either player, and many of the residual ko threats would be dame fill
anyway, they are declared as unavailable. This is not always the case,
however, as not all ko threats are dame fill, so either player could
continue finding ko threats that are not dame fill if any remaining
ko materials were left unresolved. If there is going to be any sort
of decision rendered on the status of "corner bent-4" for (B) or (C)
then the unresolved ko materials should first be filled and settled.
There is the question concerning whether player White ("O")
argues for moving the diagram from (A) to (B), or (A) to (B) to (C).
Then "prohibition on dame fill" (for purposes of territory counting)
is suspended because White argues for the capture sequence that first
fills all ko materials from the outside and then starts the interior
capture. If White can begin this way, then why should White be
prohibited from using dame fill during the sequences in (A)? Well,
to address this question there is a distinction between dame fill
when demonstrating the status of groups and dame fill for purposes
of territory counting. It is legitimate to permit dame fill for
determining the status of groups because this bears upon answering
the question of life and death of those groups. If the dame fill
is merely for the purposes of counting territory and not changing
the status of groups then that sort of dame fill may be prohibited
by either player, once players have already entered into game stop.
This rule type seems to penalize spurious "pass-pass" declarations.
I believe that Mr.Jasiek has previously argued that there are some
cases where it is not obvious whether one is counting territory or
determining the status of groups. I think Mr.Jasiek asserts that
the attempt at distinguishing prohibition of dame fill for territory
counting, versus acceptance of dame fill for ascertaining the status
of groups, must fail. We would need to review such cases once more
before passing judgment on whether such distinction is impossible.
Remember that the question after game stop was to ascertain the
status of groups. This means _answering_the_question_ concerning the
status of all groups in each of these three cases (A), (B) and (C).
Here, assume that player White ("O") accepted Black's argument that
`external' dame at c3,c4 were not fillable under the territory rule:
8 | . . . . . . 8 | . . . . . . 8 | . . . . . .
7 | . . . . . . 7 | . . . . . . 7 | . . . . . .
6 | O O O . . . 6 | O O O . . . 6 | O O O . . .
5 | # # O O . . 5 | # # O O . . 5 | # # O O . .
4 | . # . O . . 4 | . # O O . . 4 | . # O O . .
3 | O # . O . . 3 | O # . O . . 3 | O # O O . .
2 | O # # O . . 2 | O # # O . . 2 | O # # O . .
1 | O . # O . . 1 | O . # O . . 1 | O . # O . .
+------------- +------------- +-------------
a b c d e f a b c d e f a b c d e f
(A) (B) (C)
> ... If players are so weak they forget
> about the status of the involved ko's if two moves are exchanged,
> they probably should not be playing using J1989. If they do not know
> that recapturing is only allowed after a ko-pass, they are not playing
> using J1989 and their behaviour is unrelated to the question on the
> usefulness of filling ko-threats under J1989.
We don't need to speculate on what the players don't know. If they
score incorrectly then that's not a reflection on the terms of J1989.
>> We were discussing this in context to the SSK vs PSK Subject: line.
> Yes, up to the point where you started asserting that "during
> hypothetical play ('confirmation and agreement'), all ko threats
> may first be filled before approaching the problem situation under
> jurisdiction?", after which we also started discussing that this
> statement is unhelpful/wrong/not applicable to any case whatsoever.
Your point is unclear. Care to rephrase for another attempt?
The English "may ... be filled" offers merely permission to do so.
>>> Furthermore, I cannot help to notice that you fail to answer the
>>> question how the implication that filling ko-threats is useful in
>>> hypothetical sequences is read in the text of J1989.
>> Filling ko threats is -optionally- useful as a means for clarification
>> between the players during "confirmation and agreement." The
>> technical meaning of the English term "confirmation" connotes a
>> high degree of lattitude and good faith effort to get at the truth.
> So, now you are going to say that this "rule" is a matter of politeness?
> I must say I am surprised and disappointed. Couldn't you have stated
> this in a clearer way earlier in this thread? In that case I could have
> just noted that jb asserts that filling ko-threats when performing
> hypothetical analysis is polite and moved on as I am not interested
> in a discussion about the politeness of filling ko-threats during
> hypothetical analysis.
It's certainly not illegal to make a move intended to be useless
or polite. The purpose, for demonstration, was to show why corner
bent-4 is dead. In order to do so, "elsewhere ko threats" are filled.
What, specifically, do you not understand about a simple description?
>>>> The other guy might want to play a ko threat for _defending_ the ko.
>>> What do you mean by "defending the ko"? A simple example would suffice.
>> The ko defender wants to keep the ko operational, while the ko
>> capturer wants to settle the ko by eventually capturing it. In this case
>> the corner bent-4 turns into koloser for Black and komaster for White.
> Ko-threats are not useful for that purpose either. The ko-capturer will
> simply capture and fill the ko as ko-threats cannot prevent him form
> doing this during hypothetical analysis.
I gather that you reference diagrams (A) or (B). In the complex
occurrance of multiple interspersing groups it might not be feasible
to distinguish ignorable ko-threats versus non-ignorable ko-threats
when both group status and territory results are being determined.
Players might persuade each other to adopt less than correct means.
To avoid misunderstandings, and to keep the discourse of scoring
civil, players would not object even to useless and polite methods.
Why should White agree to letting Black capture three stones, when
proceeding from diagram (A), where White could simply take off the
stones when starting from diagram (C) ?
- regards
- jb
-------------------------------------------------------------------------
In Sacred Groves - Paganism in Russia
http://www.nationalvanguard.org/story.php?id=6534
-------------------------------------------------------------------------
Chris Dams
On Sun, 2005-10-16 at 05:50 +0000, - wrote:
> > "-" wrote:
> >> ...(during hypothetical play) a player might elect
> >> to fill ko-threats with what you could term "useless moves" for
> >> good measure, so as to avoid any argument concerning whether such
> >> ko-threats could be brought into play and disrupt contest over status.
>
> Chris Dams <C.D...@science.ru.nl> wrote:
> > Yes. However, in that case the opponent is apparently not playing using
> > J1989, so this issue actually has no bearing on the question whether it
> > is useful to fill ko-threats during hypothetical play .
>
> How did you reach -THAT- conclusion (not playing using J1989)?
One can have (at least) two hypotheses about the play of the opponent.
The first is that he is playing using J1989. The second is that he is
playing using a set of rules that in most respects resembles J1989 but
is different in the sense that not only ko-passes but also other moves
are sufficient to remove the ko-ban during hypothetical play. The
observed behaviour of the opponent of starting an argument that he can
use ko-threats is readily explained if the second hypothesis holds,
however is strange if the first holds. Therefore, such behaviour would
make me conclude that the player exhibiting this behaviour is not
playing using J1989.
> Are you disallowing hypothetical play under "confirmation and agreement"?
No, of course, I am not.
> >> We were discussing this in context to the SSK vs PSK Subject: line.
>
> > Yes, up to the point where you started asserting that "during
> > hypothetical play ('confirmation and agreement'), all ko threats
> > may first be filled before approaching the problem situation under
> > jurisdiction?", after which we also started discussing that this
> > statement is unhelpful/wrong/not applicable to any case whatsoever.
>
>
> Your point is unclear. Care to rephrase for another attempt?
> The English "may ... be filled" offers merely permission to do so.
Very well then, I will explain that under varying assumptions your
statement can be wrong/unapplicable/unhelpful,
(1) Wrong
I think (a thought that my dictionary confirms) that, besides indicating
allowedness, "may" can also mean that something is possible. Because of
this, I saw it as a possibility that your statement is wrong because it
can be untrue that it is possible to fill ko-threats because there can
be unremovable ones.
(2) Unapplicable
Another way of reading your statement. Now we read "may" as "is allowed
to", While this as such renders your statement true, now we face the
question why this is stated. The context talks about hypothetical play
(and hence Japanese rules) and says "before approaching the problem
situation under jurisdiction" so, presumably, this being allowed of
filling all ko-threats would help solving a problem situation in
hypothetical play. However, as I argued in this thread, actually it is
not useful (besides matters of politeness), to fill all ko-threats in
hypothetical play. This possibility of reading your statement makes it
not applicable to any case whatsoever.
(3) Unhelpful
Reading even more context, still considering the possibility that "may"
means "is allowed to", I note that your remark came up in a discussion
about the peculiarities of superko. If we take this into account for the
original position that Erik van der Werf posted, we see that while Black
is allowed to remove ko-threats (in this case by making them and forcing
White to respond) before doing something in the upper-right corner, he
does not want to, thereby rendering the statement that he is allowed to
unhelpful for the purpose of avoiding the anomaly.
> >>> Furthermore, I cannot help to notice that you fail to answer the
> >>> question how the implication that filling ko-threats is useful in
> >>> hypothetical sequences is read in the text of J1989.
>
> >> Filling ko threats is -optionally- useful as a means for clarification
> >> between the players during "confirmation and agreement." The
> >> technical meaning of the English term "confirmation" connotes a
> >> high degree of lattitude and good faith effort to get at the truth.
>
> > So, now you are going to say that this "rule" is a matter of politeness?
> > I must say I am surprised and disappointed. Couldn't you have stated
> > this in a clearer way earlier in this thread? In that case I could have
> > just noted that jb asserts that filling ko-threats when performing
> > hypothetical analysis is polite and moved on as I am not interested
> > in a discussion about the politeness of filling ko-threats during
> > hypothetical analysis.
>
>
> It's certainly not illegal to make a move intended to be useless
> or polite. The purpose, for demonstration, was to show why corner
> bent-4 is dead. In order to do so, "elsewhere ko threats" are filled.
> What, specifically, do you not understand about a simple description?
I gather that you are asking about the description "ko-threats are
filled". The "in order to do so"-part is strange as it indicates a
necessity where there is no necessity.
You *still* seem to be thinking that it is necessary to fill ko-threats
to come to a judgement of the life/death status of groups. I conclude
that from the fact that your previous post also contains the sentences
> Once White ("O") takes the ko Black tries to find a ko threat
> elsewhere. There is an on board decision: answer the Black ko threat
> or risk not capturing the corner Black group.
The ko-taking of White you are talking about occurs in this position
8 | . . . . . .
7 | . . . . . .
6 | O O O . . .
5 | # # O O . .
4 | # # O O . .
3 | . # . O . .
2 | O # # O . .
1 | # . # O . .
+-------------
a b c d e f
Taking the ko we find the position
8 | . . . . . .
7 | . . . . . .
6 | O O O . . .
5 | # # O O . .
4 | # # O O . .
3 | . # . O . .
2 | O # # O . .
1 | , O # O . .
+-------------
a b c d e f
where the comma at a1 indicates that this intersection is currently
forbidden. As you say, now Black plays a ko-threat and White answers the
ko-threat (not shown in the diagram because it is somewhere else) we
have the corner situation
8 | . . . . . .
7 | . . . . . .
6 | O O O . . .
5 | # # O O . .
4 | # # O O . .
3 | . # . O . .
2 | O # # O . .
1 | , O # O . .
+-------------
a b c d e f
Note that nothing changed. The comma is still there, indicating an
intersection on which Black is not allowed not capture. This is because
the only way to remove such a comma is by a ko-pass. However, a ko-pass
has not occured. Just a ko-threat + answer. If he plays sensibly, White
is simply going to play at a1, perhaps after many other useless ko-
threats by B + their respective answers by W. Do you *finally* get that?
Best,
Chris
-
>>> "-" wrote:
>>>> ...(during hypothetical play) a player might elect to fill ko-threats
>>>> with what you could term "useless moves" for good measure, so
>>>> as to avoid any argument concerning whether such ko-threats
>>>> could be brought into play and disrupt contest over status.
>> Chris Dams <C.D...@science.ru.nl> wrote:
>>> Yes. However, in that case the opponent is apparently not playing
>>> using J1989, so this issue actually has no bearing on the question
>>> whether it is useful to fill ko-threats during hypothetical play .
> "-" wrote:
>> How did you reach -THAT- conclusion (not playing using J1989)?
Chris Dams <C.D...@science.ru.nl> wrote:
> One can have (at least) two hypotheses about the play of the opponent.
> The first is that he is playing using J1989. The second is that he is
> playing using a set of rules that in most respects resembles J1989 but
> is different in the sense that not only ko-passes but also other moves
> are sufficient to remove the ko-ban during hypothetical play. The
> observed behaviour of the opponent of starting an argument that he
> can use ko-threats is readily explained if the second hypothesis holds,
> however is strange if the first holds. Therefore, such behaviour would
> make me conclude that the player exhibiting this behaviour is not
> playing using J1989.
In summary, you would conclude that a player is not applying
J1989 if (s)he plays an elsewhere ko-threat rather than waiting for
the ko-pass? Yet J1989 could be in effect for both players and
there was simply a misunderstanding about exercising options.
Returning to the main point, however: not all demonstrations of
group status will supply a ko-pass in the sequence. An elsewhere
ko threat may be produced, which complicates status decidability.
>>>> We were discussing this in context to the SSK vs PSK Subject: line.
>>> Yes, up to the point where you started asserting that "during
>>> hypothetical play ('confirmation and agreement'), all ko threats
>>> may first be filled before approaching the problem situation under
>>> jurisdiction?", after which we also started discussing that this
>>> statement is unhelpful/wrong/not applicable to any case whatsoever.
>> Your point is unclear. Care to rephrase for another attempt?
>> The English "may ... be filled" offers merely permission to do so.
> Very well then, I will explain that under varying assumptions your
> statement can be wrong/unapplicable/unhelpful,
> (1) Wrong
> I think (a thought that my dictionary confirms) that, besides indicating
> allowedness, "may" can also mean that something is possible. Because
> of this, I saw it as a possibility that your statement is wrong because it
> can be untrue that it is possible to fill ko-threats because there can
> be unremovable ones.
Well, we wouldn't be talking about the unremovable ones.
> (2) Unapplicable
> Another way of reading your statement. Now we read "may" as "is
> allowed to", While this as such renders your statement true, now
> we face the question why this is stated. The context talks about
> hypothetical play (and hence Japanese rules) and says "before
> approaching the problem situation under jurisdiction" so,
> presumably, this being allowed of filling all ko-threats ...
Ok, the filling of all -allowable- ko-threats... But why would
we need to specify that only the allowable plays are allowed?
> ... would help solving a problem situation in hypothetical play.
> However, as I argued in this thread, actually it is not useful (besides
> matters of politeness), to fill all ko-threats in hypothetical play. This
> possibility of reading your statement makes it not applicable to any
> case whatsoever.
It doesn't take long to formally state: "since you had passed
earlier in the `pass-pass' convention for game stop, you would
not object to these preliminary plays here ... and here ... and here
... which fill certain ko-threats which you might use during our
settlement of the group status under contention." Say, for example,
that your method of direct playing out is tried first. The opponent
uses a ko opportunity during settlement to initiate something else
on the board and then tries to claim compensation. Of course
that was an avoidable sequence altogether, and so one reverts
to the game stop position to reach the hypothetical strategy by
another, more appropriate, hypothetical sequence. At that point
one demonstrates to the opponent that ko-threats could first be
filled so that there is no compensation to be wrought from the
settlement of some particular group status under contention.
Even for the corner bent-4 case we are discussing, there
is a ko involved and so residual elsewhere ko on the board
might complicate the determination of corner bent-4 status.
> (3) Unhelpful
> Reading even more context, still considering the possibility that
> "may" means "is allowed to", I note that your remark came up in
> a discussion about the peculiarities of superko. If we take this into
> account for the original position that Erik van der Werf posted, we
> see that while Black is allowed to remove ko-threats (in this case by
> making them and forcing White to respond) before doing something
> in the upper-right corner, he does not want to, thereby rendering the
> statement that he is allowed to unhelpful for the purpose of avoiding
> the anomaly.
Suppose it is White's first move, however. Then White would
play to remove ko threats so that Black could not apply these. With
regards to Erik van der Werf's "stellingen" position, I responded by
observing that the "position" might not be repeated if reaching the
same "board state" when score had been changed. Any change
in score could imply a possible change in game result status, and
game result status bears upon the advantage/disadvantage aspect
of position, taken to be "a mass noun" under OED's 2001 addenda.
>> It's certainly not illegal to make a move intended to be useless
>> or polite. The purpose, for demonstration, was to show why corner
>> bent-4 is dead. In order to do so, "elsewhere ko threats" are filled.
>> What, specifically, do you not understand about a simple description?
> I gather that you are asking about the description "ko-threats are
> filled". The "in order to do so"-part is strange as it indicates a
> necessity where there is no necessity.
A "necessity" could arise for the case where direct play does
not suffice. As mentioned previously, an opponent might not be
persuaded on the corner bent-4 status when elsewhere residual
ko exists. Given these -two- conditions, a restart from game stop
might -become- necessary, to show how residual ko materials
may first be nullified. Remember that both players had already
previously passed to establish game stop, so it does not matter
who plays first for the hypothetical sequence.
> You *still* seem to be thinking that it is necessary to fill ko-threats
> to come to a judgement of the life/death status of groups. I conclude
> that from the fact that your previous post also contains the sentences
No, it is not -always- necessary. Having been presented with
the logic in these verbal arguments, one could just as well persuade
an opponent by talking about fill without actually playing the fills.
Because the general case allows for fill of residual ko threat prior
to settling the postion, the general case may be invoked when
addressing specific instances.
>> Once White ("O") takes the ko Black tries to find a ko threat
>> elsewhere. There is an on board decision: answer the Black
>> ko threat or risk not capturing the corner Black group.
> ... As you say, now Black plays a ko-threat and White answers the
> ko-threat (not shown in the diagram because it is somewhere else)
> we have the corner situation
>
> 8 | . . . . . .
> 7 | . . . . . .
> 6 | O O O . . .
> 5 | # # O O . .
> 4 | # # O O . .
> 3 | . # . O . .
> 2 | O # # O . .
> 1 | , O # O . .
> +-------------
> a b c d e f
>
> Note that nothing changed. The comma is still there, indicating
> an intersection on which Black is not allowed not capture.
No, Black -can- capture into the comma because "board state"
has changed by playing and responding to an elsewhere ko threat.
> This is because the only way to remove such a comma is by a ko-pass.
I'm not aware of this rule in J1989. :-)
> However, a ko-pass has not occured. Just a ko-threat + answer. If he
> plays sensibly, White is simply going to play at a1, perhaps after many
> other useless ko- threats by B + their respective answers by W. Do
> you *finally* get that?
You won't find anybody who agrees with you. The "board state"
changed from play and response to elsewhere ko threats. Sorry.
- regards
- jb
-----------------------------------------------------------
Immigration Costs More Than Thought
http://www.amren.com/mtnews/archives/2005/10/immigration_cos.php
-----------------------------------------------------------
Chris Dams
Actually, the commentary of J1989 appears to read article 7.2 in the
same way as I just presented. Let us have a look at one of the examples
that are presented there.
# O . . # O . . # O . O # .
. # # # # O . . # O O # # #
# # O O O O . . # # O O # .
O O O . . . . . . # O . O #
. . . . . . . . . # O O O O
. . . . . . . . . # # # # #
. . . . . . . . . . . . . .
(upper part of the board, made smaller in order to have to type fewer
dots)
Moves 1--6 in the hypothetical sequence (W starts):
4 O 3 . # O . . # O 6 O # .
1 # # # # O . . # O O # # #
# # O O O O . . # # O O # 5
O O O . . . . . . # O . O #
. . . . . . . . . # O O O O
. . . . . . . . . # # # # #
. . . . . . . . . . . . . .
2: pass-for-ko
And now the commentary says: "White 7 passes for the left ko, because
recapturing without first passing is prohibited." Note that after Black
4 when this ko was taken the last time, the "board state" did change.
Nevertheless, capturing in the ko is forbidden.
Best,
Chris
Tomek
> > This is because the only way to remove such a comma is by a ko-pass.
>
> I'm not aware of this rule in J1989. :-)
Article 7.2:
In the confirmation of life and death after the game stops in Article
9, recapturing in the same ko is prohibited. A player whose stone has
been captured in a ko may, however, capture in that ko again after
passing once for that particular ko capture.
> You won't find anybody who agrees with you. The "board state"
> changed from play and response to elsewhere ko threats. Sorry.
In the confirmation of life and death, recapturing in the same ko
without a ko-pass for that ko is prohibited by the Article 7.2 cited
above. Even if the "board state" changes (in fact, there is no mention
of "board state" in the ko rules).
I do not claim to understand the Japanese 1989 rules, but this one
point is quite clear.
Tomek
-
> "-" wrote:
>>> This is because the only way to remove such a comma is by a ko-pass.
>> I'm not aware of this rule in J1989. :-)
"Tomek" <tcz...@gmail.com> wrote:
> Article 7.2:
> In the confirmation of life and death after the game stops in Article 9,
> recapturing in the same ko is prohibited. A player whose stone has
> been captured in a ko may, however, capture in that ko again after
> passing once for that particular ko capture.
Article 9.3:
If a player requests resumption of a stopped game, his
opponent must oblige and has the right to play first.
So one player does not agree with the prohibition against removing
the comma by ordinary ko under Article 7.2. That player then requests
resumption of play started from the most previous game stop. Since the
request for resumption occured -during- the phase of "confirmation and
agreement" it is instead Article 9.3 which overrides Article 7.2. If
only Article 7.2 were the last word then in effect it has tacitly assumed
"all other ko threats have been filled" to make them unavailable as
alternate ko materials disruptive to settling questionable group status.
Is it the claim that "confirmation and agreement" is no longer being
pursued merely because players have resumed play under provisions
of Article 9.3 in order to achieve their "confirmation and agreement"?
This hinges upon whether "retries" are allowed under "confirmation
and agreement." Since Mr. Jasiek has spoken of many hypothetical
sequences in the study of hypothetical strategy it seems that "retries"
were permitted in the analysis he supplies, as long as resumption of
play has not yet been requested and applied. Once trying several
hypothetical sequences, a player could request resumption of play
from the game stop. At that point "retries" would be disallowed, of
course, until reaching another game stop with consecutive passes.
Let's use a practical example. Imagine that you are playing in a
very large tournament and you reach game stop with your opponent.
You proceed to "confirmation and agreement" under Article 7 to score
a "corner bent-4" (our discussion example) and then your opponent
wants to play elsewhere ko threats to which you reply, then responds
by efforts to retake corner ko in violation of Article 7.2. Moreover,
your opponent disputes your claim to the wording of Article 7.2 You
may waste 15-20 minutes in consult with the tournament directors and
referees, rulebooks, maybe to obtain the wrong answer, or you may
just say "Alright, let's begin again from game stop" and simply fill the
elsewhere alternative ko first. If dispute continues for that you can
also request resumption of play under Article 9.3 from game stop,
where it does not matter who moves first because both players had
previously entered consecutive passes.
>> You won't find anybody who agrees with you. The "board state"
>> changed from play and response to elsewhere ko threats. Sorry.
> In the confirmation of life and death, recapturing in the same ko
> without a ko-pass for that ko is prohibited by the Article 7.2 cited
> above. Even if the "board state" changes (in fact, there is no
> mention of "board state" in the ko rules).
Article 12: No result
When the same whole-board position is repeated during
a game, if the players agree, the game ends without result.
I leave it up to the semantics mavens to quibble over whether
to use "whole-board position" or "board state" for rendering the
Japanese original. I would also interpret "whole-board position"
to include the difference scoring among prisoner stones. One
problem with the wording, which has perhaps been previously
noted by Mr.Jasiek and others, is whether Article 12 is invoked
during "confirmation and agreement" after a game stop. The
"game" in effect after "game stop" seems to be that of testing
the player skills during "confirmation and agreement" (of those
procedures of "confirmation and agreement") so I would argue
that Article 12 still applies during Article 7.2 even without an
invocation of Article 9.3 within Article 7.2.
> I do not claim to understand the Japanese 1989 rules, but this one
> point is quite clear.
It is clear that the point was considerably less clear than you thought.
"Chris Dams" <C.D...@science.ru.nl> wrote:
> ( Examples supplied in prior post -- thank you very much ... )
> And now the commentary says: "White 7 passes for the left ko,
> because recapturing without first passing is prohibited." Note that
> after Black 4 when this ko was taken the last time, the "board state"
> did change. Nevertheless, capturing in the ko is forbidden.
The commentary assumes you will abide in Article 7.2 without
resumption of play, because the commentary addresses only the
hypothetical mode. Of course the advantage for both players in
hypothetical mode is the opportunity for allowing "retries" from a
game stop position. Even with Article 7.2 tattooed on your forehead
your opponent might not agree with its wording or interpretation.
You can then offer to fill kos or ask for resumption of play under
Article 9.3 where you also proceed to fill kos and settle the dispute,
or you can engage in time-wasting prevarication with tournament
directors, referees, rulebooks, and extensive typing on newsgroups.
- regards
- jb
-----------------------------------------------------------------
Putin Stymies Neo-con Iran War Drive
http://www.nationalvanguard.org/story.php?id=6539
-----------------------------------------------------------------
denis feldmann
> - wrote:
>
>>>This is because the only way to remove such a comma is by a ko-pass.
>>
>> I'm not aware of this rule in J1989. :-)
>
>
I believe the smiley here to be the (not fist) tongue-in-cheek
admittance that jb doesnt know what he is talking about, but always love
a good(?) discussion. Well, at least he didn't draw too many racist
considerations in this debate...
Robert Jasiek
<denis-f...@club-internet.fr> wrote:
>> - wrote:
>>>>This is because the only way to remove such a comma is by a ko-pass.
>>> I'm not aware of this rule in J1989. :-)
>I believe the smiley here to be the (not fist) tongue-in-cheek
>admittance that jb doesnt know what he is talking about
I guess that the smiley refers to not knowing about any "comma" rule
in J1989.
--
robert jasiek
Chris Dams
On Mon, 2005-10-17 at 03:22 +0000, - wrote:
> > "-" wrote:
> >>> This is because the only way to remove such a comma is by a ko-pass.
>
> >> I'm not aware of this rule in J1989. :-)
>
> "Tomek" <tcz...@gmail.com> wrote:
> > Article 7.2:
> > In the confirmation of life and death after the game stops in Article 9,
> > recapturing in the same ko is prohibited. A player whose stone has
> > been captured in a ko may, however, capture in that ko again after
> > passing once for that particular ko capture.
>
>
> Article 9.3:
> If a player requests resumption of a stopped game, his
> opponent must oblige and has the right to play first
Resumption is not relevant to this discussion as after a resumption we
are back in standard rules for play, as opposed to the special rules
that govern hypothetical play.
> Let's use a practical example. Imagine that you are playing in a
> very large tournament and you reach game stop with your opponent.
> You proceed to "confirmation and agreement" under Article 7 to score
> a "corner bent-4" (our discussion example) and then your opponent
> wants to play elsewhere ko threats to which you reply, then responds
> by efforts to retake corner ko in violation of Article 7.2. Moreover,
> your opponent disputes your claim to the wording of Article 7.2 You
> may waste 15-20 minutes in consult with the tournament directors and
> referees, rulebooks,
In a practical tournament I might in this case indeed fill the ko-
threats in order to avoid spending these 15-20 minutes, even though,
this way of handling the situation lacks the beauty of avoiding moves
that, strictly speaking from a rules-perspective, are not useful.
> maybe to obtain the wrong answer,
As the matter of whether my answer is wrong is still under discussion,
let us not jump to conclusions.
> >> You won't find anybody who agrees with you. The "board state"
> >> changed from play and response to elsewhere ko threats. Sorry.
>
> > In the confirmation of life and death, recapturing in the same ko
> > without a ko-pass for that ko is prohibited by the Article 7.2 cited
> > above. Even if the "board state" changes (in fact, there is no
> > mention of "board state" in the ko rules).
>
>
> Article 12: No result
> When the same whole-board position is repeated during
> a game, if the players agree, the game ends without result.
Also the possibility of "no result" is not relevant to this issue, as
this applies to regular play, as opposed to hypothetical play.
> > I do not claim to understand the Japanese 1989 rules, but this one
> > point is quite clear.
>
>
> It is clear that the point was considerably less clear than you thought.
I cannot help to notice that the prediction that I would not be able to
find anybody who agrees with me is not just untrue but that actually I
did not look for someone to agree with me but am nevertheless found by
such persons.
> "Chris Dams" <C.D...@science.ru.nl> wrote:
> > ( Examples supplied in prior post -- thank you very much ... )
> > And now the commentary says: "White 7 passes for the left ko,
> > because recapturing without first passing is prohibited." Note that
> > after Black 4 when this ko was taken the last time, the "board state"
> > did change. Nevertheless, capturing in the ko is forbidden.
>
>
> The commentary assumes you will abide in Article 7.2 without
> resumption of play, because the commentary addresses only the
> hypothetical mode. Of course the advantage for both players in
> hypothetical mode is the opportunity for allowing "retries" from a
> game stop position. Even with Article 7.2 tattooed on your forehead
> your opponent might not agree with its wording or interpretation.
> You can then offer to fill kos or ask for resumption of play under
> Article 9.3 where you also proceed to fill kos and settle the dispute,
> or you can engage in time-wasting prevarication with tournament
> directors, referees, rulebooks, and extensive typing on newsgroups.
As I would prefer to avoid having Article 7.2 tattooed on my forehead, I
might indeed fill the ko-threats in the case of a tournament-dispute.
However, as stated earlier, this way of handling the situation lacks the
beauty of avoiding moves that, strictly speaking from a rules-
perspective, are not useful.
Best,
Chris
-
>>>> Chris Dams <C.D...@science.ru.nl> wrote:
>>>>> This is because the only way to remove such a comma is by a ko-pass.
>>> "-" wrote:
>>>> I'm not aware of this rule in J1989. :-)
Obviously the same score result is available under Article 9.3 which
resumes play and establishes that a questionable ko may be recaptured.
So it was FALSE to state that "the ONLY way to remove such a comma is
by a ko-pass."
"denis feldmann" (killfiler) wrote:
> I believe the smiley here to be ...
Does Mr. Feldman believe ko-pass is the ONLY way to remove comma?
"Robert Jasiek" <jas...@snafu.de> wrote:
> I guess that the smiley refers to ...
Does Mr. Jasiek believe ko-pass is the ONLY way to remove comma?
>> "Tomek" <tcz...@gmail.com> wrote:
>>> Article 7.2:
>>> In the confirmation of life and death after the game stops in
>>> Article 9, recapturing in the same ko is prohibited. A player whose
>>> stone has been captured in a ko may, however, capture in that ko
>>> again after passing once for that particular ko capture.
>>> "-" wrote:
>> Article 9.3:
>> If a player requests resumption of a stopped game, his
>> opponent must oblige and has the right to play first
Chris Dams <C.D...@science.ru.nl> wrote:
> Resumption is not relevant to this discussion as after a resumption
> we are back in standard rules for play, as opposed to the special
> rules that govern hypothetical play.
The key to resolving this matter occurs from a remark once made
by James Kerwin, which perhaps I paraphrase: after game stop the
score cannot change. It is simply up to the players in "confirmation
and agreement" to find that score. If they have to resume play using
Article 9.3 when they cannot agree on the reading of Article 7.2, then
they shall exercise that option available to them in the J1989 features.
What's "relevant to this discussion" concerns how players arrive at
"confirmation and agreement" and whether any provision at all occurs
which allows for filling of ko to establish the logic behind Article 7.2
which leads to the same score result as continuation under Article 9.3.
Neither Article 7.2 nor Article 9.3 disallow pre-filling of ko threats.
In that fact is harmonization which redeems it as a beautiful notion.
>> Let's use a practical example. Imagine that you are playing in a
>> very large tournament and you reach game stop with your opponent.
>> You proceed to "confirmation and agreement" under Article 7 to score
>> a "corner bent-4" (our discussion example) and then your opponent
>> wants to play elsewhere ko threats to which you reply, then responds
>> by efforts to retake corner ko in violation of Article 7.2. Moreover,
>> your opponent disputes your claim to the wording of Article 7.2 You
>> may waste 15-20 minutes in consult with the tournament directors and
>> referees, rulebooks,
> In a practical tournament I might in this case indeed fill the ko-
> threats in order to avoid spending these 15-20 minutes, even though,
> this way of handling the situation lacks the beauty of avoiding moves
> that, strictly speaking from a rules-perspective, are not useful.
Once you've scored the game and posted the game result, then
you can haul out the rulebook for indulging in opponent education.
That, too, is "optional behavior" by the way.
>> ... maybe to obtain the wrong answer,
> As the matter of whether my answer is wrong is still under discussion,
> let us not jump to conclusions.
No, your answer could be correct but in the context of a tournament
you would need to abide by whatever answer the -referees- supply.
If you're willing to take a chance on that, with more time-wasting
opportunities, then you "may" as well exercise that option, too.
>> Article 12: No result
>> When the same whole-board position is repeated during
>> a game, if the players agree, the game ends without result.
> Also the possibility of "no result" is not relevant to this issue, as
> this applies to regular play, as opposed to hypothetical play.
If you do not have a resolution for problems of repetition, via
"SSK or PSK" during hypothetical play, then how would you ever
call a halt to repetition during hypothetical play? Whatever the
disposition -- "no result" or "take back one move" -- you cannot
have infinite sequences for hypothetical play in tournament games.
Article 12 would come into effect again within the Article 9.3 that
got invoked during the hypothetical play out of Article 7.2. At
least that's one possibility for what -could- occur, hypothetically.
>>> I do not claim to understand the Japanese 1989 rules, but this one
>>> point is quite clear.
>> It is clear that the point was considerably less clear than you thought.
> I cannot help to notice that the prediction that I would not be able to
> find anybody who agrees with me is not just untrue but that actually
> I did not look for someone to agree with me but am nevertheless
> found by such persons.
Did they agree with you, specifically, that "ONLY way to remove
such a comma is by a ko-pass"? Remember Mr. Jasiek's formulation
concerning hypothetical strategies: that each is arrived at by means of
ANY combination of hypothetical sequences leading to that identical
hypothetical strategy. Including the "hypothetical sequence" which
RESUMES PLAY in order to establish the same hypothetical strategy.
Can resumption of play occur -WITHIN- a hypothetical sequence, or
would hypothetical sequences need to be REPLAYED from game stop?
Does "Tomek" believe ko-pass is the ONLY way to remove comma?
>> The commentary assumes you will abide in Article 7.2 without
>> resumption of play, because the commentary addresses only the
>> hypothetical mode. Of course the advantage for both players in
>> hypothetical mode is the opportunity for allowing "retries" from a
>> game stop position. Even with Article 7.2 tattooed on your forehead
>> your opponent might not agree with its wording or interpretation.
>> You can then offer to fill kos or ask for resumption of play under
>> Article 9.3 where you also proceed to fill kos and settle the dispute,
>> or you can engage in time-wasting prevarication with tournament
>> directors, referees, rulebooks, and extensive typing on newsgroups.
> As I would prefer to avoid having Article 7.2 tattooed on my forehead,
> I might indeed fill the ko-threats in the case of a tournament-dispute.
> However, as stated earlier, this way of handling the situation lacks the
> beauty of avoiding moves that, strictly speaking from a rules-
> perspective, are not useful.
In the eye of the beholder, apparently. I find it a rather beautiful
idea that the prohibition of ko recapture under Article 7.2 is not
different from the pre-filling of ko threats under Article 9.3 invoked
within the context of Article 7.2, or that pre-filling of ko threats can
occur anyway within the context of Article 7.2 even irrespective of
whether Article 9.3 was formally referenced, just in case a player
might be considering re-initiating Article 9.3 for play resumption.
If you prefer, the player filling ko threats can simply declare to the
opponent: "All (potentially competing) ko threats are pre-filled..."
If the idea was to enable players to score properly on their own,
so that arbitrarion by mysterious officials was made unnecessary,
then pre-filling of ko threats retains player sovereignty over results.
And -THAT- is yet another meaning for "we the people ..."
- regards
- jb
-------------------------------------------------------------------------
Russia Asks Japan to Take Part in Kliper Spacecraft Development
http://www.nationalvanguard.org/story.php?id=6515
-------------------------------------------------------------------------
ian
denis feldmann wrote:
> Tomek a écrit :
> > - wrote:
> >
> >>>This is because the only way to remove such a comma is by a ko-pass.
> >>
> >> I'm not aware of this rule in J1989. :-)
> >
> >
>
> admittance that jb doesnt know what he is talking about, but always love
> a good(?) discussion. Well, at least he didn't draw too many racist
> considerations in this debate...
>
mais denis,
Is it not the case that every post he makes is in itself an admission
of this?
-
"ian" (yet another killfiler ignoramus) <siva...@gmail.com> wrote:
> Is it not the case that every post he makes is in itself an admission
> of this?
I admit that five posters already have failed to account for Article 9.3.
- regards
- jb
----------------------------------------------------------------
Disaster Response: Spot the Difference
http://www.nationalvanguard.org/story.php?id=6509
----------------------------------------------------------------
denis feldmann
-
denis feldmann (fucken frogie killfiler) wrote:
> implicitly, yes. But he doesn't often put the smiley..
You can see what standards Feldmann has for research.
He simply does not review the evidence opposite to his bigotry.
- regards
- jb
----------------------------------------------------------------
Lies, Swinging Chickens, and a License to Steal
http://www.nationalvanguard.org/story.php?id=6504
----------------------------------------------------------------
Chris Dams
On Mon, 2005-10-17 at 09:48 +0000, - wrote:
> >>>> Chris Dams <C.D...@science.ru.nl> wrote:
> >>>>> This is because the only way to remove such a comma is by a ko-pass.
>
> >>> "-" wrote:
> >>>> I'm not aware of this rule in J1989. :-)
>
>
> Obviously the same score result is available under Article 9.3 which
> resumes play and establishes that a questionable ko may be recaptured.
> So it was FALSE to state that "the ONLY way to remove such a comma is
> by a ko-pass."
No, it was not false to state this. Actually, your assertion that the
same score result is available under article 9.3 is not generally true.
Let us combine your earlier example of a bent-4 in the corner with the
second example offered in http://senseis.xmp.net/?UnremovableKoThreat .
(10 x 7 board)
7 . . . . . . . . . .
6 O O O . . . . . . .
5 # # O . . O O O O O
4 . # O O . O # # # #
3 O # . O . O # O . #
2 O # # O . O # . O .
1 O . # O . O # # . O
a b c d e f g h i j
Note that in the LR-corner Black has a ko-threat but that White cannot
play there or he is going to loose the entire corner. In particular,
White cannot remove Blacks ko-threat. Now, White would like to
demonstrate that the black string in the LL-corner is dead. He can do
this in hypothetical play because the ko-pass rule is rendering Blacks
ko-threat useless. However, what useful thing can White do if play is
resumed? He should not start the ko in the bent-4, obviously, as he is
going to lose it (or else the LR-corner) under normal ko-rules as Black
has a ko-threat. Also filling at c3 is not an option as it costs a
point. A move in the LR-corner then? No, that is going to cost that
corner. Black too has nothing useful to do under resumption. After
resumption the right way to play would be to have two passes. This
brings us back to where we were: a position where hypothetical play and
the ko-pass rule are absolutely necessary to find the right score.
You can dream on about this "harmonization which redeems it as a
beautiful notion" if you wish, however this "harmonization" does not
happen in the example I showed above.
> >> Let's use a practical example. Imagine that you are playing in a
> >> very large tournament and you reach game stop with your opponent.
> >> You proceed to "confirmation and agreement" under Article 7 to score
> >> a "corner bent-4" (our discussion example) and then your opponent
> >> wants to play elsewhere ko threats to which you reply, then responds
> >> by efforts to retake corner ko in violation of Article 7.2. Moreover,
> >> your opponent disputes your claim to the wording of Article 7.2 You
> >> may waste 15-20 minutes in consult with the tournament directors and
> >> referees, rulebooks,
>
> > In a practical tournament I might in this case indeed fill the ko-
> > threats in order to avoid spending these 15-20 minutes, even though,
> > this way of handling the situation lacks the beauty of avoiding moves
> > that, strictly speaking from a rules-perspective, are not useful.
>
>
> Once you've scored the game and posted the game result, then
> you can haul out the rulebook for indulging in opponent education.
> That, too, is "optional behavior" by the way.
Yes. It is true that a tournament is not the most suitable place to
"indulge in opponent education". Therefore, I would be reluctant to do
this under these circumstances. Ideally, this would happen at a go-club
meeting (or part of a meeting) specifically allocated for this purpose.
You do not seem to have much trust in my abilities at appropriate
behaviour :-(.
> >> Article 12: No result
> >> When the same whole-board position is repeated during
> >> a game, if the players agree, the game ends without result.
>
> > Also the possibility of "no result" is not relevant to this issue, as
> > this applies to regular play, as opposed to hypothetical play.
>
>
> If you do not have a resolution for problems of repetition, via
> "SSK or PSK" during hypothetical play, then how would you ever
> call a halt to repetition during hypothetical play? Whatever the
> disposition -- "no result" or "take back one move" -- you cannot
> have infinite sequences for hypothetical play in tournament games.
This is a good question, however, you should remember that they are
*hypothetical* sequences. If we want to be fully accurate, we must
realize that they are actually only imagined. In the case of an imagined
sequence, there is not so much wrong with an infinite one (except that
one might start to wonder if the process of proving life and death can
actually be decided in a finite amount of time, but such is life (no pun
intended) if it is lived under J1989). In fact, a possible hypothetical
sequence in "Life-and-Death Example 25: Double-Ko Seki" is the infinite
B-capture, W-capture, B pass for the one ko, W pass for the other ko, B-
capture, W-capture, and so on. This sequence can be imagined and if both
players think that this is a strategically optimal one, it follows that
the two single stones are dead.
> Article 12 would come into effect again within the Article 9.3 that
> got invoked during the hypothetical play out of Article 7.2. At
> least that's one possibility for what -could- occur, hypothetically.
No, sorry, this is not the way it is done. A hypothetical sequence has a
result, even if it is infinitely long. E.g. a hypothetical sequence that
is infinitely long can have the property that the string of which the
status is currently under dispute is never captured during it. If this
infinitely long hypothetical sequence is strategically optimal, this
would prove that this disputed string is alive.
> > I cannot help to notice that the prediction that I would not be able to
> > find anybody who agrees with me is not just untrue but that actually
> > I did not look for someone to agree with me but am nevertheless
> > found by such persons.
>
>
> Did they agree with you, specifically, that "ONLY way to remove
> such a comma is by a ko-pass"? Remember Mr. Jasiek's formulation
> concerning hypothetical strategies: that each is arrived at by means of
> ANY combination of hypothetical sequences leading to that identical
> hypothetical strategy. Including the "hypothetical sequence" which
> RESUMES PLAY in order to establish the same hypothetical strategy.
> Can resumption of play occur -WITHIN- a hypothetical sequence, or
> would hypothetical sequences need to be REPLAYED from game stop?
> Does "Tomek" believe ko-pass is the ONLY way to remove comma?
I read his statement as such, however I think it is impolite to start
discussing the details of somebodies wording when not talking to that
person. At least, I would not feel comfortable about it. It should be
Tomeks choice if he wants to further disambiguify his statement. Of
course, if you wish, you could post a request to Tomek in this group
asking him to clarify his statement.
> In the eye of the beholder, apparently. I find it a rather beautiful
> idea that the prohibition of ko recapture under Article 7.2 is not
> different from the pre-filling of ko threats under Article 9.3 invoked
> within the context of Article 7.2, or that pre-filling of ko threats can
> occur anyway within the context of Article 7.2 even irrespective of
> whether Article 9.3 was formally referenced, just in case a player
> might be considering re-initiating Article 9.3 for play resumption.
You can find it a "rather beautiful idea", however, as argued above, it
is wrong.
> If the idea was to enable players to score properly on their own,
> so that arbitrarion by mysterious officials was made unnecessary,
> then pre-filling of ko threats retains player sovereignty over results.
> And -THAT- is yet another meaning for "we the people ..."
I do appreciate this sentiment. The best way to make it happen would be
to stop the use of Japanese rules. In J1989 there are some highly arcane
features, such as ko-passes and infinite hypothetical sequences.
Best,
Chris
-
>>>>>> Chris Dams <C.D...@science.ru.nl> wrote:
>>>>>>> ... the only way to remove such a comma is by a ko-pass.
>>>>> "-" wrote:
>>>>>> I'm not aware of this rule in J1989. :-)
> "-" wrote:
>> Obviously the same score result is available under Article 9.3
>> which resumes play and establishes that a questionable ko may
>> be recaptured. So it was FALSE to state that "the ONLY way to
>> remove such a comma is by a ko-pass."
Yes, this is specifically what you are talking about and brings
into sharp relief the problems with Article 9.3. You are correct
about the problem of Unremovable Ko Threats in applying general
pre-fill of residual ko, and why analysis reverts back to Article 7.2.
Please accept my apology for misunderstanding your argument.
I suggest that your send this to the various bestiaries at (or via):
http://www.cs.cmu.edu/~wjh/go/rules/bestiary.html
http://www.cs.cmu.edu/~wjh/go/rules/siivola.html
http://www.cs.cmu.edu/~wjh/go/rules/fearnley.html
http://www.cs.cmu.edu/~wjh/go/rules/jasiek.html
I see also that I should read more carefully at webpage:
http://senseis.xmp.net/?BentFourInTheCornerIsDead
An expanded size for the Unremovable Ko Threat shape does
not help White, because White would reply to Black's Ko Threats,
allowing Black to resume a defense of the LL corner. Because a
higher temperature in the Unremovable Ko Threat does not help
White it is correct to label this shape as an anomaly. So "corner
bent-4" is not always dead, as previously assumed. We also have
Black taking the LR corner as consequence of losing LL, so the
score change has occured when a guiding precept was "score
cannot change after game stop." At that "game stop" (prior to
forcing continuation of hypothetical play under Article 7.2) both
positions were effective seki. Neither player wished to fill dame.
Moreover, there are problems in the wording of Article 7.2:
Article 7.2:
" [ ... ] A player whose stone has been captured in a ko may,
however, capture in that ko again after passing once for that
particular ko capture."
Black could argue that, if compelled to sustain hypothetical
play after having reached a second "pass pass" with game stop
under Article 9.3, and being disallowed more delaying tactics
at repeated efforts to request resumption of play, that benefits of
the 2nd sentence of Article 7.2 have been rendered unavailable.
Where was the opportunity for Black to capture "in THAT ko again"
after having made a "ko-pass" if White is going to make an inside
three shape? Black's capture of three stones, instead of just one,
has formed _another_ ko shape type, which White then destroys.
>> Once you've scored the game and posted the game result, then
>> you can haul out the rulebook for indulging in opponent education.
>> That, too, is "optional behavior" by the way.
> You do not seem to have much trust in my abilities at
> appropriate behaviour :-(.
Eh? I said "optional behavior."
> ... a possible hypothetical sequence in "Life-and-Death Example 25:
> Double-Ko Seki" is the infinite B-capture, W-capture, B pass for the
> one ko, W pass for the other ko, B-capture, W-capture, and so on.
> This sequence can be imagined and if both players think that this is a
> strategically optimal one, it follows that the two single stones are dead.
Well, back to Article 7.2:
" [ ... ] A player whose stone has been captured in a ko may,
however, capture in that ko again after passing once for that
particular ko capture."
Capture in the ko is permitted after passing -ONCE- but there
is -NO- mention of permission after passing _more_ than ONCE.
Here distinguish "capturing that stone" from "capturing in that ko."
>> Article 12 would come into effect again within the Article 9.3 that
>> got invoked during the hypothetical play out of Article 7.2. At
>> least that's one possibility for what -could- occur, hypothetically.
> No, sorry, this is not the way it is done. A hypothetical sequence has
> a result, even if it is infinitely long. E.g. a hypothetical sequence that
> is infinitely long can have the property that the string of which the
> status is currently under dispute is never captured during it. If this
> infinitely long hypothetical sequence is strategically optimal, this
> would prove that this disputed string is alive.
Do you refer to a basis in precedent for an "infinite hypothetical
sequence"? I don't see how this could be permitted by Article 7.2.
>> In the eye of the beholder, apparently. I find it a rather beautiful
>> idea that the prohibition of ko recapture under Article 7.2 is not
>> different from the pre-filling of ko threats under Article 9.3 invoked
>> within the context of Article 7.2, or that pre-filling of ko threats can
>> occur anyway within the context of Article 7.2 even irrespective of
>> whether Article 9.3 was formally referenced, just in case a player
>> might be considering re-initiating Article 9.3 for play resumption.
> You can find it a "rather beautiful idea", however, as argued above,
> it is wrong.
Owing to the fact that Article 7.2 is compulsory, it should not
matter whether the players arrive at their score via Article 7.2 or
by means of courteous play under Article 9.3. That's still a rather
beautiful idea, and not wrong at all. The "score cannot change
after game stop."
>> If the idea was to enable players to score properly on their own,
>> so that arbitrarion by mysterious officials was made unnecessary,
>> then pre-filling of ko threats retains player sovereignty over results.
>> And -THAT- is yet another meaning for "we the people ..."
> I do appreciate this sentiment. The best way to make it happen would
> be to stop the use of Japanese rules. In J1989 there are some highly
> arcane features, such as ko-passes and infinite hypothetical sequences.
There's no search result for "infinite hypothetical sequence" on
http://senseis.xmp.net . The only other Go reference I find on google
stems from Mr. Jasiek's commentary, itself without another reference.
Where, and when, does that -ever- occur under J1989 ?
- regards
- jb
-----------------------------------------------------------------
Tell us who fabricated the Iraq evidence
http://www.nationalvanguard.org/story.php?id=6495
-----------------------------------------------------------------
Chris Dams
On Tue, 2005-10-18 at 09:25 +0000, - wrote:
> Yes, this is specifically what you are talking about and brings
> into sharp relief the problems with Article 9.3. You are correct
> about the problem of Unremovable Ko Threats in applying general
> pre-fill of residual ko, and why analysis reverts back to Article 7.2.
I am glad that this has become clear.
> >> Once you've scored the game and posted the game result, then
> >> you can haul out the rulebook for indulging in opponent education.
> >> That, too, is "optional behavior" by the way.
>
> > You do not seem to have much trust in my abilities at
> > appropriate behaviour :-(.
>
>
> Eh? I said "optional behavior."
Yes. Sorry. Language problem. I thought that "to indulge in" had a
negative connotation. Checking with my dictionary suggests that it is
actually supposed to be pleasant to be indulged into something.
> > ... a possible hypothetical sequence in "Life-and-Death Example 25:
> > Double-Ko Seki" is the infinite B-capture, W-capture, B pass for the
> > one ko, W pass for the other ko, B-capture, W-capture, and so on.
> > This sequence can be imagined and if both players think that this is a
> > strategically optimal one, it follows that the two single stones are dead.
>
>
> Well, back to Article 7.2:
>
> " [ ... ] A player whose stone has been captured in a ko may,
> however, capture in that ko again after passing once for that
> particular ko capture."
>
>
> Capture in the ko is permitted after passing -ONCE- but there
> is -NO- mention of permission after passing _more_ than ONCE.
> Here distinguish "capturing that stone" from "capturing in that ko."
Note *for that particular ko-capture*. In J1989 a ko-pass is always for
a particular ko-capture (this is by the way a difference between J2003
and J1989: in J2003 a ko-pass removes *all* ko-bans). Because there are
two ko-captures, it is possible that first a ko-pass for one of them
occurs and then a ko-pass for the other.
> >> Article 12 would come into effect again within the Article 9.3 that
> >> got invoked during the hypothetical play out of Article 7.2. At
> >> least that's one possibility for what -could- occur, hypothetically.
>
> > No, sorry, this is not the way it is done. A hypothetical sequence has
> > a result, even if it is infinitely long. E.g. a hypothetical sequence that
> > is infinitely long can have the property that the string of which the
> > status is currently under dispute is never captured during it. If this
> > infinitely long hypothetical sequence is strategically optimal, this
> > would prove that this disputed string is alive.
>
>
> Do you refer to a basis in precedent for an "infinite hypothetical
> sequence"? I don't see how this could be permitted by Article 7.2.
I think it is permitted in the just-discussed example of a double ko-
seki. However, if we are going to enter the issue of infinite
hypothetical sequences I must admit that now I am on much less firmer
ground than in the discussion about ko-passes. I know how Robert Jasiek
interprets this and am not aware of any other interpretations (apart
from a possible simplification, that I will explain below).
> >> If the idea was to enable players to score properly on their own,
> >> so that arbitrarion by mysterious officials was made unnecessary,
> >> then pre-filling of ko threats retains player sovereignty over results.
> >> And -THAT- is yet another meaning for "we the people ..."
>
> > I do appreciate this sentiment. The best way to make it happen would
> > be to stop the use of Japanese rules. In J1989 there are some highly
> > arcane features, such as ko-passes and infinite hypothetical sequences.
>
>
> There's no search result for "infinite hypothetical sequence" on
> http://senseis.xmp.net . The only other Go reference I find on google
> stems from Mr. Jasiek's commentary, itself without another reference.
> Where, and when, does that -ever- occur under J1989 ?
Well, you yourself brought up the issue that a hypothetical sequence
could be infinitely long. As this indeed is problematic, something ought
to be said/done about it. Robert Jasieks proposal is a way, albeit an
impractical one, to look at it. I myself suggested in this group a more
practical way to handle the problem. It is in
http://groups.google.nl/group/rec.games.go/browse_frm/thread/9219beff7379f2b0/5ada69c946b6e361?lnk=st&rnum=1&hl=nl#5ada69c946b6e361
The proposal is in the sentence "If the position repeats during
hypothetical play, one could stop the hypothetical sequence and treat it
as if an infinite sequence was played."
In the post
http://groups.google.nl/group/rec.games.go/browse_frm/thread/3acbd0b789a36f8a/277b30b39b4b33e1?lnk=st&rnum=1&hl=nl#277b30b39b4b33e1
I am more concrete on how this could work in practice.
Best,
Chris
Chris Dams
> Well, you yourself brought up the issue that a hypothetical sequence
> could be infinitely long.
Actually, I meant to say "Well, you yourself brought up the issue that
the position might repeat in a hypothetical sequence."
-
> "-" wrote:
>> Yes, this is specifically what you are talking about and brings
>> into sharp relief the problems with Article 9.3. You are correct
>> about the problem of Unremovable Ko Threats in applying general
>> pre-fill of residual ko, and why analysis reverts back to Article 7.2.
Chris Dams <C.D...@science.ru.nl> wrote:
> I am glad that this has become clear.
Please realize that it is not entirely clear. Increasing the size
of that Unremovable Ko Threat sufficiently merely compels White
to reply and then Black survives.
>>>> Once you've scored the game and posted the game result, then
>>>> you can haul out the rulebook for indulging in opponent education.
>>>> That, too, is "optional behavior" by the way.
>>> You do not seem to have much trust in my abilities at
>>> appropriate behaviour :-(.
>> Eh? I said "optional behavior."
> Yes. Sorry. Language problem. I thought that "to indulge in" had a
> negative connotation. Checking with my dictionary suggests that it is
> actually supposed to be pleasant to be indulged into something.
Indulge can have a negative connotation but not in terms of
judgmentalism on "appropriate versus inappropriate" behavior.
>>> ... a possible hypothetical sequence in "Life-and-Death Example 25:
>>> Double-Ko Seki" is the infinite B-capture, W-capture, B pass for the
>>> one ko, W pass for the other ko, B-capture, W-capture, and so on.
>>> This sequence can be imagined and if both players think that this is a
>>> strategically optimal one, it follows that the two single stones are dead.
>> Well, back to Article 7.2:
>>
>> " [ ... ] A player whose stone has been captured in a ko may,
>> however, capture in that ko again after passing once for that
>> particular ko capture."
>>
>> Capture in the ko is permitted after passing -ONCE- but there
>> is -NO- mention of permission after passing _more_ than ONCE.
>> Here distinguish "capturing that stone" from "capturing in that ko."
> Note *for that particular ko-capture*. In J1989 a ko-pass is always for
> a particular ko-capture (this is by the way a difference between J2003
> and J1989: in J2003 a ko-pass removes *all* ko-bans). Because there
> are two ko-captures, it is possible that first a ko-pass for one of them
> occurs and then a ko-pass for the other.
I'm uncertain what was originally intended by the Japanese, yet
in English there is ambiguity in the sense. Associating the ko-pass
with that ko is not at issue. If both "ko" and "capture" are nouns, as
a compound noun, then "ko capture" is the compound term being
addressed. Not the different stone being captured which changes
the term but the space event itself among various possible other
"ko captures" on display in a problem task. Nothing is said about
whether a player can play THAT "ko capture" after passing TWICE,
even if the second ko pass occurs after the ko has cycled. When
permission is not granted perhaps permission is not being granted.
Another reason for this interpretation may stem from a proof that no
condition occurs which requires more than one ko-pass on a ko.
>> There's no search result for "infinite hypothetical sequence" on
>> http://senseis.xmp.net . The only other Go reference I find on google
>> stems from Mr. Jasiek's commentary, itself without another reference.
>> Where, and when, does that -ever- occur under J1989 ?
> Well, you yourself brought up the issue that a hypothetical sequence
> could be infinitely long. As this indeed is problematic, something
> ought to be said/done about it. Robert Jasieks proposal is a way,
> albeit an impractical one, to look at it. I myself suggested in this
> group a more practical way to handle the problem. It is in
> http://groups.google.nl/group/rec.games.go/browse_frm/thread/9219beff7379f2b0/5ada69c946b6e361?lnk=st&rnum=1&hl=nl#5ada69c946b6e361
> The proposal is in the sentence "If the position repeats during
> hypothetical play, one could stop the hypothetical sequence and
> treat it as if an infinite sequence was played."
In any case, tests for SSK or PSK would need to be in effect.
Whether such tests are components of the scoring procedure
is a completely different question. Article 7.2 of J1989 wants to
bypass the "no repetition rule" by means of imposing a variant
style of Go where ko cannot be recaptured without a ko-pass.
The assumption here is that "infinite hypothetical sequences"
do not need to occur under the stipulations of Article 7.2 and
perhaps we can devise a proof to illustrate that assumption.
I think there was a guiding aesthetic at work when drafting
the terms of Article 7.2, and to abandon that guiding aesthetic
misses its intention. Abandonment would need a compelling
justifying reason sufficient to usurp the aesthetics of Article 7.2.
Mr. Jasiek has not directed his energies toward the question of
comparitive aethetics, concerning "no ko retake without ko pass"
versus the complexities of SSK or PSK, or neither of them.
> In the post
> http://groups.google.nl/group/rec.games.go/browse_frm/thread/3acbd0b789a36f8a/277b30b39b4b33e1?lnk=st&rnum=1&hl=nl#277b30b39b4b33e1
> I am more concrete on how this could work in practice.
Each of these interesting links. I am fairly confident that
the "no result" effect on violations of SSK or PSK will not last.
Article 7.2 was one method for obtaining a result which does
not portend to run up against SSK or PSK by -forcing- the
"ko pass" requirement amidst the prospect of ko recapture. At
the end of the day Mr. Jasiek may return to appreciate J1989.
- regards
- jb
-----------------------------------------------------------------
UK: Imprisoned Pagans Win Right to Worship
http://www.nationalvanguard.org/story.php?id=6569
-----------------------------------------------------------------
Simon Goss
>> Note *for that particular ko-capture*. In J1989 a ko-pass is always for
>> a particular ko-capture (this is by the way a difference between J2003
>> and J1989: in J2003 a ko-pass removes *all* ko-bans). Because there
>> are two ko-captures, it is possible that first a ko-pass for one of them
>> occurs and then a ko-pass for the other.
>
>
> I'm uncertain what was originally intended by the Japanese,
>yet
> in English there is ambiguity in the sense. Associating the ko-
>pass
> with that ko is not at issue. If both "ko" and "capture" are
>nouns, as
> a compound noun, then "ko capture" is the compound term being
> addressed. Not the different stone being captured which changes
> the term but the space event itself among various possible other
> "ko captures" on display in a problem task. Nothing is said about
> whether a player can play THAT "ko capture" after passing TWICE,
> even if the second ko pass occurs after the ko has cycled. When
> permission is not granted perhaps permission is not being granted.
> Another reason for this interpretation may stem from a proof that
>no
> condition occurs which requires more than one ko-pass on a ko.
I think there is no real ambiguity. It is as if any stone that captures
in a ko is thereby marked with a dot. A stone with a dot cannot be
captured. A ko-pass removes exactly one dot from exactly one stone, and
is the only way to remove a dot from a stone. A player who makes a ko
pass needs to indicate which stone he is de-dotting, if there are more
than one.
--
Simon
UK Go Challenge for schools
http://www.ukgochallenge.com
-
>>> Note *for that particular ko-capture*. In J1989 a ko-pass is always for
>>> a particular ko-capture (this is by the way a difference between J2003
>>> and J1989: in J2003 a ko-pass removes *all* ko-bans). Because there
>>> are two ko-captures, it is possible that first a ko-pass for one of them
>>> occurs and then a ko-pass for the other.
> "-" <jazze...@hotmail.com> writes
>> I'm uncertain what was originally intended by the Japanese, yet
>> in English there is ambiguity in the sense. Associating the ko-pass
>> with that ko is not at issue. If both "ko" and "capture" are nouns, as
>> a compound noun, then "ko capture" is the compound term being
>> addressed. Not the different stone being captured which changes
>> the term but the space event itself among various possible other
>> "ko captures" on display in a problem task. Nothing is said about
>> whether a player can play THAT "ko capture" after passing TWICE,
>> even if the second ko pass occurs after the ko has cycled. When
>> permission is not granted perhaps permission is not being granted.
>> Another reason for this interpretation may stem from a proof that no
>> condition occurs which requires more than one ko-pass on a ko.
Simon Goss <si...@gosoft.demon.co.uk> wrote:
> I think there is no real ambiguity. It is as if any stone that captures
> in a ko is thereby marked with a dot. A stone with a dot cannot be
> captured. A ko-pass removes exactly one dot from exactly one stone,
> and is the only way to remove a dot from a stone. A player who makes
> a kopass needs to indicate which stone he is de-dotting, if there are
> more than one.
The question, Simon, concerns whether a position for scoring
under Article 7.2 would ever need more than one ko-pass by that
player for that particular ko.
- regards
- jb
-------------------------------------------------------------
New book finds evidence of historical
"face that launched a thousand ships"
http://www.nationalvanguard.org/story.php?id=6570
-------------------------------------------------------------
Chris Dams
On Wed, 2005-10-19 at 06:32 +0000, - wrote:
> > "-" wrote:
> >> Yes, this is specifically what you are talking about and brings
> >> into sharp relief the problems with Article 9.3. You are correct
> >> about the problem of Unremovable Ko Threats in applying general
> >> pre-fill of residual ko, and why analysis reverts back to Article 7.2.
>
> Chris Dams <C.D...@science.ru.nl> wrote:
> > I am glad that this has become clear.
>
>
> Please realize that it is not entirely clear. Increasing the size
> of that Unremovable Ko Threat sufficiently merely compels White
> to reply and then Black survives.
I see that I should have read your previous post better. Could you show
hypothetical sequences for this survival by Black? I really have
difficulty believing that the size of the unremovable ko-threat changes
life/death status.
I do not find this prose enlightening. I would *really* appreciate it if
you could write sentences that I can understand without having to read
them ten or more times first, at the same time acquiring a headache. Let
me just add that the commentary *does* have ko-captures after two ko-
passes in a row. "White 9 and Black 10 pass for kos on the right. White
11 takes ko; this is legal because White has passed at play 9. Black 12
takes ko; this is legal because Black has passed at play 10." This
occurs in the commentary of Article 7.2 and it *really* is quite
consistent with my interpretation.
Best,
Chris
Chris Dams
Ah, you want to add to (or read into) article 7.2 that after a
particular ko has (during a hypothetical sequence) seen a capture, a ko-
pass and a recapture (possibly with moves elsewhere in between them), no
further recapturing and/or ko-passing is allowed. I am afraid that it is
not going to work. Life-and-Death Example 16 gives the wrong result if
we try this.
Best,
Chris
T Mark Hall
<C.D...@science.ru.nl> writes
>Dear jb,
>
<Snip>
>Ah, you want to add to (or read into) article 7.2 that after a
>particular ko has (during a hypothetical sequence) seen a capture, a ko-
>pass and a recapture (possibly with moves elsewhere in between them), no
>further recapturing and/or ko-passing is allowed. I am afraid that it is
>not going to work. Life-and-Death Example 16 gives the wrong result if
>we try this.
>
>Best,
>Chris
>
depressing thought that there were people in this world who had nothing
better to do with their spare time than post this stuff onto a news
group. Watching paint dry would be more interesting, but I don't have
any paint. I think I'll play Go on one of the servers I haven't been
banned from. Bet some pedant starts criticising my sentence structure.
Have fun.
Best wishes.
--
T Mark Hall
www.gogod.demon.co.uk
www.gogod.demon.co.uk/NewInGo/NewInGo.htm
-
> "-" wrote:
>> Please realize that it is not entirely clear. Increasing the size
>> of that Unremovable Ko Threat sufficiently merely compels White
>> to reply and then Black survives.
Chris Dams <C.D...@science.ru.nl> wrote:
> I see that I should have read your previous post better. Could you
> show hypothetical sequences for this survival by Black? I really have
> difficulty believing that the size of the unremovable ko-threat changes
> life/death status.
(10 x 7 board)
+---------------------+
7 | . . . O # # # # # # |
6 | O O O O # O O O O # |
5 | # # O O # O O O O # |
4 | . # O O # O O O O # |
3 | O # . O # # # O . # |
2 | O # # O O O # . O . |
1 | O . # O . O # # . O |
+---------------------+
a b c d e f g h i j
White will probably reply to Black's ko threat on LR. Black lives LL.
>> I'm uncertain what was originally intended by the Japanese, yet
>> in English there is ambiguity in the sense. Associating the ko-pass
>> with that ko is not at issue. If both "ko" and "capture" are nouns, as
>> a compound noun, then "ko capture" is the compound term being
>> addressed. Not the different stone being captured which changes
>> the term but the space event itself among various possible other
>> "ko captures" on display in a problem task. Nothing is said about
>> whether a player can play THAT "ko capture" after passing TWICE,
>> even if the second ko pass occurs after the ko has cycled. When
>> permission is not granted perhaps permission is not being granted.
>> Another reason for this interpretation may stem from a proof that no
>> condition occurs which requires more than one ko-pass on a ko.
> I do not find this prose enlightening. I would *really* appreciate it
> if you could write sentences that I can understand without having
> to read them ten or more times first, at the same time acquiring a
> headache. Let me just add that the commentary *does* have
> ko-captures after two ko-passes in a row. "White 9 and Black 10
> pass for kos on the right. White 11 takes ko; this is legal because
> White has passed at play 9. Black 12 takes ko; this is legal because
> Black has passed at play 10." This occurs in the commentary of
> Article 7.2 and it *really* is quite consistent with my interpretation.
Each player, White and Black, has the privilege of passing
ONCE for each ko. If you add up the number of kos (k) in that
example and multiply by the number of players in the game (2),
then the total number of passes cannot exceed 2*k. I assume
the URL is http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html .
>> Simon Goss <si...@gosoft.demon.co.uk> wrote:
>>> I think there is no real ambiguity. It is as if any stone that
>>> captures in a ko is thereby marked with a dot. A stone with
>>> a dot cannot be captured. A ko-pass removes exactly one dot
>>> from exactly one stone, and is the only way to remove a dot
>>> from a stone. A player who makes a kopass needs to indicate
>>> which stone he is de-dotting, if there are more than one.
> "-" wrote:
>> The question, Simon, concerns whether a position for scoring
>> under Article 7.2 would ever need more than one ko-pass by that
>> player for that particular ko.
"Chris Dams" <C.D...@science.ru.nl> wrote:
> Ah, you want to add to (or read into) article 7.2 that after a particular
> ko has (during a hypothetical sequence) seen a capture, a ko-pass
> and a recapture (possibly with moves elsewhere in between them),
> no further recapturing and/or ko-passing is allowed. I am afraid that
> it is not going to work. Life-and-Death Example 16 gives the wrong
> result if we try this.
To begin with, let's define "wrong result" as something other
than what Example 16 wishes to say. In the double-ko seki one
could have both players simply playing for capture indefinitely.
After one player opts for a ko-pass the other player had better opt
for a ko-pass too, else there could be no defending ko recapture.
Because Article 12 does not apply, the double-ko seki could be
recycling without a scoring resolution: capture, capture, ko-pass,
ko-pass, capture, capture, ko-pass, ko-pass, capture, capture ...
(stop). If each player is allowed ko-pass only -ONCE- for each ko
then this scoring sequence must stop after four ko-passes into that
double-ko seki. Otherwise Article 7.2 fails to score the position,
which was its original goal, and failing to score the position would
be a "wrong result" in the context to what we've defined it to be.
After exhausting all possibilities in the double-ko seki perhaps
it is then easier to understand how Example 16 must be scored.
- regards
- jb
----------------------------------------------------------------
Gun Rights Group Calls for Immediate Action
http://www.nationalvanguard.org/story.php?id=6559
----------------------------------------------------------------
-
T Mark Hall (son-of-a-bitch british killfiler) wrote:
> You know, as I went to get myself a stronger drink, I had the
Perhaps the cause of all your problems, (and making them ours).
Notice that "T Mark Hall" contributed -NOTHING- to the discussion.
- regards
- jb
-----------------------------------------------------------------
A Study in Zionist Censorship
http://www.nationalvanguard.org/story.php?id=6563
-----------------------------------------------------------------
Chris Dams
On Wed, 2005-10-19 at 19:57 +0000, - wrote:
> Chris Dams <C.D...@science.ru.nl> wrote:
> > I see that I should have read your previous post better. Could you
> > show hypothetical sequences for this survival by Black? I really have
> > difficulty believing that the size of the unremovable ko-threat changes
> > life/death status.
>
>
> (10 x 7 board)
> +---------------------+
> 7 | . . . O # # # # # # |
> 6 | O O O O # O O O O # |
> 5 | # # O O # O O O O # |
> 4 | . # O O # O O O O # |
> 3 | O # . O # # # O . # |
> 2 | O # # O O O # . O . |
> 1 | O . # O . O # # . O |
> +---------------------+
> a b c d e f g h i j
>
>
> White will probably reply to Black's ko threat on LR. Black lives LL.
To me Black looks as dead as always. Hypothetical sequence for showing
that the black string in the LL-corner is dead: Wb1, Ba4, Wa2, Ba1, Wb1,
Bi1, Wj2. The position is now
+---------------------+
7 | . . . O # # # # # # |
6 | O O O O # O O O O # |
5 | # # O O # O O O O # |
4 | # # O O # O O O O # |
3 | . # . O # # # O . # |
2 | O # # O O O # . O O |
1 | , O # O . O # # # O |
+---------------------+
a b c d e f g h i j
There is a comma on a1 indicating a ko-banned intersection and Black has
nothing. Do you have anything better for Black?
> Each player, White and Black, has the privilege of passing
> ONCE for each ko. If you add up the number of kos (k) in that
> example and multiply by the number of players in the game (2),
> then the total number of passes cannot exceed 2*k. I assume
> the URL is http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html .
Thanks. Clear as glass. Unfortunately, it still does not work. See
below.
> "Chris Dams" <C.D...@science.ru.nl> wrote:
> > Ah, you want to add to (or read into) article 7.2 that after a particular
> > ko has (during a hypothetical sequence) seen a capture, a ko-pass
> > and a recapture (possibly with moves elsewhere in between them),
> > no further recapturing and/or ko-passing is allowed. I am afraid that
> > it is not going to work. Life-and-Death Example 16 gives the wrong
> > result if we try this.
>
>
> To begin with, let's define "wrong result" as something other
> than what Example 16 wishes to say.
Yes, of course.
> In the double-ko seki one
> could have both players simply playing for capture indefinitely.
> After one player opts for a ko-pass the other player had better opt
> for a ko-pass too, else there could be no defending ko recapture.
> Because Article 12 does not apply, the double-ko seki could be
> recycling without a scoring resolution: capture, capture, ko-pass,
> ko-pass, capture, capture, ko-pass, ko-pass, capture, capture ...
> (stop). If each player is allowed ko-pass only -ONCE- for each ko
> then this scoring sequence must stop after four ko-passes into that
> double-ko seki. Otherwise Article 7.2 fails to score the position,
> which was its original goal, and failing to score the position would
> be a "wrong result" in the context to what we've defined it to be.
>
> After exhausting all possibilities in the double-ko seki perhaps
> it is then easier to understand how Example 16 must be scored.
No. Not working. Hypothetical sequence to show the capture of the
rightmost white string:
B: capture in UL-corner,
W: capture in DKS,
B: capture in DKS,
W: ko-pass for DKS,
B: ko-pass for DKS,
W: capture in DKS,
B: capture in DKS,
W: ko-pass for DKS,
B: ko-pass for DKS
(from this point on, no further ko-passes for DKS allowed),
W: capture in DKS,
B: capture in DKS,
(from this point on, no further captures in DKS allowed),
However, no further captures in the DKS means that B is never ever going
to capture the rightmost white string as it has two permanent liberties.
Best,
Chris
ian
Chris Dams wrote:
*snip*
Have you ever considered it might be fatuous argueing a position with
somebody who was not interested in it.
-
------------------------------------------------------------
Serious problem, BGA. Suggest you foward an email copy to:
"ian" ("best-of-the-british" killfilers) <siva...@gmail.com> wrote:
> Have you ever considered it might be fatuous argueing a
> position with somebody who was not interested in it.
Was "ian" certain that he is posting to the correct newsgroup?
- regards
- jb
--------------------------------------------------------
http://www.whatreallyhappened.com/
--------------------------------------------------------
Simon Goss
>Simon Goss <si...@gosoft.demon.co.uk> wrote:
>> I think there is no real ambiguity. It is as if any stone that captures
>> in a ko is thereby marked with a dot. A stone with a dot cannot be
>> captured. A ko-pass removes exactly one dot from exactly one stone,
>> and is the only way to remove a dot from a stone. A player who makes
>> a kopass needs to indicate which stone he is de-dotting, if there are
>> more than one.
>
>
> The question, Simon, concerns whether a position for scoring
> under Article 7.2 would ever need more than one ko-pass by that
> player for that particular ko.
The answer, Jeff, is yes. More than one pass for a particular ko, but
_not_ for a particular ko _capture_, which is what article 7.2 says.
-
>> Simon Goss <si...@gosoft.demon.co.uk> wrote:
>>> I think there is no real ambiguity. It is as if any stone that
>>> captures in a ko is thereby marked with a dot. A stone with a
>>> dot cannot be captured. A ko-pass removes exactly one dot
>>> from exactly one stone, and is the only way to remove a dot
>>> from a stone. A player who makes a kopass needs to indicate
>>> which stone he is de-dotting, if there are more than one.
> "-" <jazze...@hotmail.com> writes:
>> The question, Simon, concerns whether a position for scoring
>> under Article 7.2 would ever need more than one ko-pass by
>> that player for that particular ko.
Simon Goss <si...@gosoft.demon.co.uk> wrote:
> The answer, Jeff, is yes. More than one pass for a particular ko, but
> _not_ for a particular ko _capture_, which is what article 7.2 says.
No, Simon. That is not an answer to my question. I am not
asking what Article 7.2 says, though that is certainly a very critical
question to ask as well, and a discussion in its own right. I might
also add that I do not agree with your interpretation of Article 7.2.
The question I am asking refers to whether there exists a position
that players prefer to score under Article 7.2 (rather than by playing
it out) which must require a repeated ko-pass by one of the players
for a particular ko -- or for a particular "ko capture" -- (not to mean
a ko rendered slightly different from any previous ko in that area).
If there is no position then your reading of Article 7.2 is superfluous.
Let me thank you, however, for trying to address the matter.
- regards
- jb
-------------------------------------------------------------------
England Expects: With Nelson to Victory
http://www.nationalvanguard.org/story.php?id=6576
-------------------------------------------------------------------
Simon Goss
> No, Simon. That is not an answer to my question. I am not
> asking what Article 7.2 says, though that is certainly a very
>critical
> question to ask as well, and a discussion in its own right.
> I
>might
> also add that I do not agree with your interpretation of Article
>7.2.
> The question I am asking refers to whether there exists a position
> that players prefer to score under Article 7.2 (rather than by
>playing
> it out) which must require a repeated ko-pass by one of the players
> for a particular ko -- or for a particular "ko capture" -- (not to
>mean
> a ko rendered slightly different from any previous ko in that
>area).
> If there is no position then your reading of Article 7.2 is
>superfluous.
>
> Let me thank you, however, for trying to address the matter.
I see. Well, when you earlier wrote ...
> I'm uncertain what was originally intended by the Japanese, yet
> in English there is ambiguity in the sense.
... I was led to imagine that you were talking about the rule, since the
Japanese addresses the rules, not the existence of positions with
particular attributes. Apparently I misunderstood. Oh well.
-
> "-" <jazze...@hotmail.com> writes:
>> The question I am asking refers to whether there exists a position
>> that players prefer to score under Article 7.2 (rather than by playing
>> it out) which must require a repeated ko-pass by one of the players
>> for a particular ko -- or for a particular "ko capture" -- (not to mean
>> a ko rendered slightly different from any previous ko in that area).
>> If there is no position then your reading of Article 7.2 is superfluous.
>>
>> Let me thank you, however, for trying to address the matter.
Simon Goss <si...@gosoft.demon.co.uk> wrote:
> I see. Well, when you earlier wrote ...
>>>> I'm uncertain what was originally intended by the Japanese,
>>>> yet in English there is ambiguity in the sense.
> ... I was led to imagine that you were talking about the rule, since
> the Japanese addresses the rules, not the existence of positions
> with particular attributes. Apparently I misunderstood. Oh well.
Bear in mind that the NiHon-KiIn was dragging its heels on a
formulation of rule corrections, more than 30 years, which led to
Mr. Ing's complicated ruleset, even after they had promised to Go
Seigen that they would work on rules revisions. The Japanese
were not particularly concerned whether they had the rules exact.
It's true that one of my inquiries concerns semantic interpretation
however in the general sense I am not satisfied with a definitional
fallacy. For positions being scored under Article 7.2 will we find
any situations where a player needs to register more than one
"ko-pass" for a _particular_ ko (that ko shape, not one in the same
area which had formerly a different shape)? If the answer is "no"
then there is no need to consider the problems of an "infinite
hypothetical sequence" which would preclude finite analysis.
At that point I would be justified in claiming the interpretation of
Article 7.2 to read that a "ko-pass" for a particular ko capture may
be limited to only ONCE, and still be scoring all potential positions.
- regards
- jb
-------------------------------------------------------------------------
English City Apologises for Destroying Welsh-speaking Community
http://www.nationalvanguard.org/story.php?id=6589
-------------------------------------------------------------------------
Simon Goss
> Bear in mind that the NiHon-KiIn was dragging its heels on a
> formulation of rule corrections, more than 30 years, which led to
> Mr. Ing's complicated ruleset, even after they had promised to Go
> Seigen that they would work on rules revisions. The Japanese
> were not particularly concerned whether they had the rules exact.
> It's true that one of my inquiries concerns semantic interpretation
> however in the general sense I am not satisfied with a definitional
> fallacy. For positions being scored under Article 7.2 will we find
> any situations where a player needs to register more than one
> "ko-pass" for a _particular_ ko (that ko shape, not one in the same
> area which had formerly a different shape)? If the answer is "no"
> then there is no need to consider the problems of an "infinite
> hypothetical sequence" which would preclude finite analysis.
> At that point I would be justified in claiming the interpretation
>of
> Article 7.2 to read that a "ko-pass" for a particular ko capture
>may
> be limited to only ONCE, and still be scoring all potential
>positions.
Ah, light begins to dawn.
However, I don't quite understand what you mean by ...
>a _particular_ ko (that ko shape, not one in the same
> area which had formerly a different shape)
What do you consider the scope of "the same area", here? The definition
of ko is "a shape in which the players can alternately capture and
recapture one opposing stone. There are three such shapes: this one in
the middle
# O
# a b O where there is either a white
# O stone at a or else a black at b
and one each in the corners and on the sides away from the corners. I
would suggest that stones and intersections other than those shown are
not part of the ko shape. If you want it larger, why not the whole
board?
But I think you may be mistaken in your claim that
>If the answer is "no"
> then there is no need to consider the problems of an "infinite
> hypothetical sequence" which would preclude finite analysis.
Infinite hypothetical sequences exist with no ko at all. The place to
look is L&D example 6 (chosei).
-
> "-" <jazze...@hotmail.com> writes:
>> ... For positions being scored under Article 7.2 will we find
>> "ko-pass" for a _particular_ ko (that ko shape, not one in the same
>> area which had formerly a different shape)? If the answer is "no"
>> then there is no need to consider the problems of an "infinite
>> hypothetical sequence" which would preclude finite analysis.
>> At that point I would be justified in claiming the interpretation of
>> Article 7.2 to read that a "ko-pass" for a particular ko capture may
>> be limited to only ONCE, and still be scoring all potential positions.
Simon Goss <si...@gosoft.demon.co.uk> wrote:
> Ah, light begins to dawn.
> However, I don't quite understand what you mean by ...
>> ... a _particular_ ko (that ko shape, not one in the same
>> area which had formerly a different shape)
> What do you consider the scope of "the same area", here?
Key phrase is "...IN the same area" or "...in the same region."
The recent ko occurs on points shared with the former ko shape.
> The definition of ko is "a shape in which the players can alternately
> capture and recapture one opposing stone. There are three such
> shapes: this one in the middle
>
> # O
> # a b O where there is either a white
> # O stone at a or else a black at b
>
> and one each in the corners and on the sides away from the corners.
> I would suggest that stones and intersections other than those shown
> are not part of the ko shape. If you want it larger, why not the whole
> board?
Matthew Woodcraft points out that there are other, larger ko shapes.
Any _nakade_ sequence entails playing out the stones in a ko shape.
The "alternate capture" quality more generally refers to whether the
colors of stones, claims to attached (temporary) territory, also would
alternate in the sequences which simply the position toward scoring.
Robert Jasiek has published webpages devoted to ko discussions,
attempting a ko taxomony or enumeration of various types of ko.
The ko references an unsettled position or unsettled shape, as by
contrast to seki or life which Go Players regard as already decided.
I read Article 7.2's reference to "that particular ko capture" to mean
not only the physical location of the ko but also the type of capture,
i.e. whether it was 1-stone, 2-stone, 3-stone, 4-stone capture, etc.,
as well as the specific shape in which the ko capture occurs. Once
simplifying a ko capturing sequence down to exchanges of 1-stone
then the position cycling must stop within ten moves for a double ko
seki, as previously described: capture, capture, ko-pass, ko-pass,
capture, capture, ko-pass, ko-pass, capture, capture, ( STOP ) or,
depending on the overall complexity, position cycling must stop
within a finite number of ko-passes. Each player may exercise
the privilege of a ko-pass only once for "that particular ko capture."
In general it might seem that ko-pass applies chiefly to single stone
cycling, however I am reading Article 7.2 more generally than that.
> But I think you may be mistaken in your claim that
>> If the answer is "no"
>> then there is no need to consider the problems of an "infinite
>> hypothetical sequence" which would preclude finite analysis.
> Infinite hypothetical sequences exist with no ko at all. The place to
> look is L&D example 6 (chosei).
Long life entails two 2-stone kos along the side of the board,
just as with 2-stone kos in a double-seki. The identification and/or
recognition of ko is the question here, its definition; whether it is to
mean simply a region where stones are being exchanged repeatedly
in an unsettled condition. Examples for Article 12 under Article 9.3
rule long life as "no result" (third example in Diagram 18). Example 6
considers a ruling in context of Article 7.2. In any case White is alive
however if White plays first White can kill Black. The ruling which has
Black pre-emptorily dead allows White to kill without playing that extra
stone. If repetition by SSK or PSK were the guide here, White's play
with the single stone on the right side, after the two 2-stone captures,
would seem to violate repetition. J1989 has declared the situation
too difficult for SSK or PSK application, owing to the favorable effect
for Black. Yet Article 7.2 responds by awarding a favorable effect to
White. If Black is going to lose the game by that amount, then Black
cannot accept the Article 7.2 ruling and instead opts for Article 9.3,
which compels White to play another stone for the inside-5 shape
if White can afford to do so, and thereby avoid a "no result" ruling.
If White cannot afford to play another stone, they accept "no result."
Because the decision by White one way or the other is so close
to a tie anyway, annullment is no major disappointment for each.
URL is at: http://www.cs.cmu.edu/~wjh/go/rules/Japanese.html .
The question chiefly concerns the scope of SSK or PSK in Go.
A prisoner count difference did not change in the "whole-board
position" concept, so White's throw-in safety on the right-side of
the situation appears to repeat "whole-board position." Actually,
this was an end of one "double ko cycle" in the long-life condition.
Ing Rules characterize Black as "disturber" so one pass intervenes
before Black can restart a sequence, which -begins- from disturbing.
Or, Ing Rules state that the disturber cannot continue a recyling ko.
In that one pass by Black, White can play the necessary stone for
the inside-5 shape. The players could also decide that the situation
is seki, since they are the parties to "confirmation and agreement,"
however White here would not consider seki a favorable decision.
Well, the White stones may remain on the board, so anytime White
can play to remove the Black stones. That was the criterion.
I don't know which ruleset to which you refer that mentions
anything about "infinite hypothetical sequences."
- regards
- jb
-------------------------------------------------------------------
Ing's SST Laws of Wei-Chi 1991
http://www.cs.cmu.edu/~wjh/go/rules/SST.html
-------------------------------------------------------------------