Screwy Documentation (and Scoring Procedures) on IGS
-
Posted to rec.games.go because email to IGS Admin is bouncing.
Concerning disagreement over life & death in ending positions:
===================================
(1) Cannot enter "done" until the game is properly scored ---
===================================
Usage: done
'done' is used to signify when players are done removing the dead stones,
and the game is to be scored. At the end of a game, 3 pass moves will
begin the scoring stage. After the third pass, this message will appear:
You can check your score with the score command, type 'done' when finished.
Then the prompt will change to: Enter Dead Groups:
After ALL dead stones have been removed, score the game by entering: done
Both players must enter 'done'. After 'done', the result cannot be changed.
_Be_ watchful for _live_ stones being removed. In case this were to happen,
enter: undo
===================================
(2) So Players disagree and try scoring according to what they think,
indicating that disagreement by repeated entries of "undo."
===================================
Usage: undo
NOTE: The undo command changed in January 2006, and the replacement is
'undoplease'. Usage: undoplease
If a player sends 'undo', the server replies with:
5 You cannot undo because you are not requested.
5 see help undoplease
'undoplease' works as follows:
The player must have 'toggle newundo on', otherwise you will see this
message "Opponent's client does not support undoplease."
Player A sends undoplease
Player B receives the message "24 *SYSTEM*: xxx requests undo."
Player B sends 'undo' (to agree), or 'noundo' (to decline).
If a 'noundo' is sent, the game will become a no undo game.
The 'teach' command has unlimited undo's.
See also: coords done match pass scoring teach
===================================
(3) After five repeat disagreements of "undo" IGS indicates game adjournment.
===================================
Usage: adjourn
'adjourn' is used by either player in a game to request the
adjournment of the game. The game will be saved. To 'adjourn'
the game, both players must enter: adjourn
Adjourned games can also be 'sealed'. See: help sealed
The opponent will then receive the following messages:
#> Your opponent requests an adjournment.
Use <adjourn> to adjourn, or <decline adjourn> to decline.
To 'adjourn' the game, the opponent must enter: adjourn
Both players will then see the following messages:
Game has been adjourned.
Game saved.
To be sure the current board position is saved, enter: save
To restart an adjourned game, use the 'load' command.
Adjourned games are 'stored'. Please read: help stored
For additional information, be sure to read: help escaping
See also: decline escaping load resign save sealed stored touch uptime
===================================
(4) There is no indication that the game was saved. (No "loook" available.)
===================================
Usage: look <whiteplayer-blackplayer>
'look' will display a 'stored' (adjourned) game. You must either know
the name of the game you wish to 'look' at, or use the 'stored' command
to determine the name of a 'stored' game.
To 'look' at a 'stored' game, enter: look whiteplayer-blackplayer
Example: look Tobby-Ivan
'look' will show the date and time when the game was left unfinished.
If the game was left unfinished due to an adjournment or disconnection,
'look' will show which player adjourned or disconnected first. If IGS
was restarted or rebooted, neither player will be listed as having
adjourned or disconnected from the game.
See also: adjourn load save stored touch
===================================
(5) There is no record of the game in "stored" nor anything in "results."
===================================
===================================
(6) The scuttlebut is that if cheaters don't like losing on IGS all they need
to do is disagree about scoring with repeated "undo" and then the game
result is thwarted. This happens rather often on IGS, apparently.
===================================
===================================
(7) It seems that no penalties accrue to a player who creates problems
by specious disagreement during scoring phase, and causes the
game to become adjourned (and lost to the bitbucket). No policy
statement that players who cause problems receive extra losses.
===================================
===================================
(8) IGS is the result of a lot of "research" and thousands of players with
playing experiences. Difficult to fathom why such shoddy procedures
persist to the present day at IGS. Only lack of listening could explain.
===================================
- regards
- jb
-----------------------------------------------------------------
Norwegian Authorities Are Covering Up Muslim Rapes
http://www.wvwnews.net/story.php?id=294
-----------------------------------------------------------------
Michael Alford
that need to be addressed, and I didn't want anyone bitching about
"having" to read jb's post to get to my offering, so jb's post is below.
While I personally do not like several of the recent changes to IGS,
there is really no problem with the new undo command. Both the glGo and
Panda-IGS clients support it. If you do not use either of these clients,
there is a link in the motd leading you to explanations of the new undo
and stored commands, and how to use the new undo with an older client.
It is now almost impossible to cheat on IGS. Even if you pass and go to
the scoring phase, and your opponent repeatedly removes your live stones
forcing you to undo until the system adjourns the game, which will
happen, the game is not lost, it does not disappear. IGS has a team of
professional players, I don't know how many there are or whom they may
be, but these pros adjudicate adjourned games. Nothing is lost. Stats
are updated. All adjourned and "odd" games are now adjudicated. It may
take several days for your stats to update, but no adjourned games are
lost, they are adjudicated. People have noticed and remarked on this on
the server. I haven't seen a single complaint about this process :)
Cheaters and other offensive people do not have a free rein. If you are
a subscriber, you can submit a complaint about another account to Panda
Patrol and have it immediately investigated. I am not a subscriber, but
I know from personal experience and the experience of others that an
email to the admins about behavioral problems gets attention. Cheaters
and behavior problem accounts are not tolerated and end up disabled.
This is a bit different from the historic "ban" oft discussed here,
access to these newer accounts is not denied, you can log in, but they
don't work properly, and the account holder is advised that he must
re-register, full operating ability is not restored, as in the past. I'm
told that having to go through the process of re-registering takes care
of a lot of the behavior problems. While cheating and behavior problems
are not common, IGS, at the ripe old age of 15 is pretty staid and
proper these days :), they do happen occasionally, and are dealt with
when they do occur.
Cheers,
Michael
-
Michael Alford <ma...@aracnet.com> wrote:
> While I personally do not like several of the recent changes to IGS,
> there is really no problem with the new undo command. Both the glGo
> and Panda-IGS clients support it. If you do not use either of these clients,
> there is a link in the motd leading you to explanations of the new undo
> and stored commands, and how to use the new undo with an older client.
A bit squirrelly already. Documentation for "undo" is identical to
documentation for "undoplease." What, for the reader, distinguishes
"undo" from "undoplease"? Client "support" appears to be nothing
other than setting "toggle newundo on" because these are merely
pass-through commands. Here's another -toggle- which does not
display in "stats" so users haven't any idea of all toggles being made
available for settings. Are we to understand that "toggle newundo off"
applies only for games and not during the scoring phase?
> It is now almost impossible to cheat on IGS. Even if you pass and go
> to the scoring phase, and your opponent repeatedly removes your live
> stones forcing you to undo until the system adjourns the game, which
> will happen, the game is not lost, it does not disappear. IGS has a team
> of professional players, I don't know how many there are or whom they
> may be, but these pros adjudicate adjourned games. Nothing is lost.
> Stats are updated. All adjourned and "odd" games are now adjudicated.
> It may take several days for your stats to update, but no adjourned games
> are lost, they are adjudicated. People have noticed and remarked on this
> on the server. I haven't seen a single complaint about this process :)
The game has been declared "adjourned" however in documentation
for "adjourn" it says that game has been "saved." One feature on glGo
can even "save" the game on the server. Even so there is no "stored"
game to "look" at once it is "saved" by the forced "adjourn" procedure.
Says right there: ---> " Adjourned games are 'stored'. "
Now "IGS has a team of professional players..." But suppose, after
five "undo" efforts (maybe difficulty communicating with other player)
players get together and communicate some more. They could each
examine their "stored" game with "look" (if this were made feasible).
The players discuss game scoring and agree how it should be scored.
They "load" & score. Game result recorded. Professionals not bothered.
Pay professionals, if you will, but let them use their time effectively.
There is an incentive for players to settle their own dispute if IGS has
in place this policy: "A game sent for forced arbitration to PandaPatrol
will penalize losing player 2 games, who incorrectly disputed results,
and reward winning player 2 games, who correctly disputed results."
Losing players, incorrectly disputing game results, also need some
education or the incorrect scoring procedure is likely to reoccur. The
message could be sent to that player, an email, the game record, the
ascii of final position and some words of explanation in that player's
native language. Then the losing player who has incorrectly disputed
is also informed of the 2 game loss as a consequence of being stupid.
Eventually the losing players will learn how to properly score and then
become winning players. Here "winning" is defined as "not bothering
the professionals unnecessarily at PandaPatrol adjudication HQs."
> Cheaters and other offensive people do not have a free rein. If you
> are a subscriber, you can submit a complaint about another account
> to Panda Patrol and have it immediately investigated. I am not a subscriber,
> but I know from personal experience and the experience of others that
> an email to the admins about behavioral problems gets attention.
> Cheaters and behavior problem accounts are not tolerated and end up
> disabled. This is a bit different from the historic "ban" oft discussed here,
> access to these newer accounts is not denied, you can log in, but they
> don't work properly, and the account holder is advised that he must
> re-register, full operating ability is not restored, as in the past. I'm
> told that having to go through the process of re-registering takes care
> of a lot of the behavior problems. While cheating and behavior problems
> are not common, IGS, at the ripe old age of 15 is pretty staid and
> proper these days :), they do happen occasionally, and are dealt with
> when they do occur.
Would an -honest- difference of opinion during scoring comprise
"cheating"? What if that player's behavior were repetitive? Having a
penalty for improper scoring efforts would tend to enforce discipline
among the players, provide a disincentive to "cheating" attempts and
reward players who apply proper scoring procedures in that context.
What occurs on KGS, or other "alternative" servers, in scoring disputes?
- regards
- jb
--------------------------------------------------------------
Iacocca bashes Bush in new book
http://www.detnews.com/apps/pbcs.dll/article?AID=/20070412/POLITICS/704120383
--------------------------------------------------------------
Robert Jasiek
wrote:
>It is now almost impossible to cheat on IGS.
It is nice to hear that many years after my suggestions in that
direction something is done finally.
>pros adjudicate adjourned games [...] I haven't seen a single complaint
>about this process
Here are some: It is a waste of time of the professionals. Instead
simpler rules of play could be used. - It leaves an awful feeling that
games are decided arbitrarily instead of only due to the players'
skills. Do the professionals at least comment on their decisions so
that one can verify whether they have made mistakes?
--
robert jasiek
Bantari
> It leaves an awful feeling that
> games are decided arbitrarily instead of only due to the players'
> skills.
Would that include skills at "cheating"? After all, if the rules of
IGS allow such undo and redo, and whatever, manipulating them is also
a skill within the rules itself, no? I recall a dispute in which you
claimed that a skill at passing is a valid strategic skill within the
applicable rules (and you even tried toclaim a win because the
opponent passed once too many times?) How is that different from the
skill at "undo" if the rules allow it? So, if the opponent manages to
confuse or fool you with undos into clinking the wrong button, or
typing the wrong command, this should be also a valid strategy to play
for a win.
One could say that "cheating" is not possible on the ne... since if
the rules of the server allow you to do something, it is within the
rules, and thus not cheating at all, but strategy instead. :)
I know, I know, the rules are all flawed, but still...
This is funny. :)
________________
-Bantari
http://www.bantari.net/
tweet
> On Mon, 16 Apr 2007 08:48:20 -0700, Michael Alford <ma...@aracnet.com>
> wrote:
>> It is now almost impossible to cheat on IGS.
>
> It is nice to hear that many years after my suggestions in that
> direction something is done finally.
>
Actually, the person responsible for the implementation and
idea on IGS does not read or understand English and has never
read rec.games.go.
Robert Jasiek
>Would that include skills at "cheating"?
no
>After all, if the rules of
>IGS allow such undo and redo,
These are not rules of play, not even tournament rules, but go server
command nonsense (some say "features").
>I recall a dispute in which you
>claimed that a skill at passing is a valid strategic skill within the
>applicable rules (and you even tried toclaim a win because the
>opponent passed once too many times?) How is that different from the
>skill at "undo" if the rules allow it?
"pass" is part of the rules of play. "undo" is not.
--
robert jasiek
Bantari
> These are not rules of play, not even tournament rules, but go server
> command nonsense (some say "features").
Rules are rules are rules... On a servers they include more than just
where you can and cannot play. One could argue that "pass" is a simple
communication of an intent. So is "undo" and the rest of the
"commands". Real-life tournament rules specify how to deal with
"undo" (its not allowed), while IGS rules have a different
specification and a different protocol of dealing with it. But its
still a "rule" by which you have to abide.
______________
-Bantari
http://www.bantari.net/
T Mark Hall
<tw...@pandanet.net> writes
Best wishes.
--
T Mark Hall
Honorary Vice-President, British Go Association
http://www.gogod.demon.co.uk/index.htm
http://www.gogod.demon.co.uk/NewInGo/NewInGo.htm
-
"tweet" <tw...@pandanet.net> wrote:
> Actually, the person responsible for the implementation and
> idea on IGS does not read or understand English and has never
> read rec.games.go.
Though that "person responsible" expects other people who
are responsible to be forwarding (and translating, if necessary) any
relevant suggestions for improvement so that IGS remains competitive.
------------------------------------------------------------
> Robert Jasiek <jas...@snafu.de> wrote:
>> These are not rules of play, not even tournament rules,
>> but go server command nonsense (some say "features").
One might also say that filling unnecessary dame is feature nonsense.
"Bantari" <ban...@yahoo.com> wrote:
> Rules are rules are rules...
This, yet, from the same poster who pronounced:
" This is not a code. " ( Tues, Apr 10 2007 )
http://groups.google.com/group/rec.games.go/msg/43aa97b22cf226f3
> On a servers they include more than just where you can and cannot play.
> One could argue that "pass" is a simple communication of an intent.
Yes, "pass" communicates an intent, however it is also a directed
element for game play to signal a phase transition in game activity.
> So is "undo" and the rest of the "commands". Real-life tournament
> rules specify how to deal with "undo" (its not allowed), while IGS rules
> have a different specification and a different protocol of dealing with it.
> But its still a "rule" by which you have to abide.
Earlier we were discussing "rules in general" and not merely a
particular person's notion of Rules. Possible to incorporate "undo"
into a Ruleset, say with the proviso that the "undo" costs One Stone
and/or time on the clock. If we're speaking of "real-life tournament
rules" then of course we would consider what people are capable
of doing in real-life. An "undo" may just as well be a component of
"tournament rules" say for a scholastic endeavor among the kiddies.
Some Chess Tournaments demand "touch move" and others do not.
Some Chess rules require backing up the position if King Check was
unnoticed, or in the context of an illegal move. Usually the "undo" is
fenced-off as unplayable by those who wish not to entertain this sort
of discussion. Yet that's a convenience which particularizes a RuleSet.
Placing a stone inexactly on a wrong point and sliding a little bit, etc.
This introduction of "new undo" at IGS also suggests that those who
do -not- select for "toggle newundo off" are -inviting- "undo" requests.
Even if a player never honors an "undo" request, the decision to allow
incoming "undo" requests amounts to tacit admission of being more than
a mere procedure. Whether playing conventions are rules or features are
finer points. Apparently all required for something to be a "rule" is a
mention in some blurbsheet describing events at that social gathering.
- regards
- jb
----------------------------------------------------------
* Bntr ntrprtr (- prjct cde -)
1 > 2 Thnk bt t nd tr t ndrstnd wht t sys
3 <, ~(mndlssl rdng nd r-rdng smthng mlln tms)
5 grnt dpr lvl f ndrstndng 7 ## ~(ffhnd sspcn)
----------------------------------------------------------
Bantari
> This, yet, from the same poster who pronounced:
>
> " This is not a code. " ( Tues, Apr 10 2007 )http://groups.google.com/group/rec.games.go/msg/43aa97b22cf226f3
And what does this have to do with anything? I asked you for a CODE
and you provided a totally crappy pseudo-code which did not even
perform to specs.
In any case - I actually happen to agree with both you and Robert on
this particular issue. I am poking at him because I am curious how he
will defend his position here. From my previous discussions with him
I was under the impression that he considered as strict "rules" more
than just the principles guiding placement of stones. As I racall, he
defended such things as skill at passing, for example, as being part
of the rules, and thus application thereof a strategic skill. I was
also under the impression that other things, such as when to press the
clock, what to bring to the game and what not, how to resign, and so
on are also considered by him in the same category. In general, all
actions around the game environment which are explicitly allowed or
not are "rules", and skillful application of them part of the game.
Thus - I am wondering why he made a jump from the above position to a
statement that undo on IGS, which is specifically governed by the IGS
"rules", is suddenly not a "rule", but only a "feature". What is the
difference, for him here. Or maybe I misunderstood his previous
statements. If so, then nothing will get explained if nobody asks.
I would not be surprised if he claimed that IGS rules are "flawed".
But his statement that the ability to "undo" and when you can apply it
is not a "rule" puzzles me.
_____________
-Bantari
http://www.bantari.net/
Robert Jasiek
>he defended such things as skill at passing, for example, as being part
>of the rules,
No.
>and thus application thereof a strategic skill.
Yes.
>I was
>also under the impression that other things, such as when to press the
>clock, what to bring to the game and what not, how to resign, and so
>on are also considered by him in the same category.
They are tournament rules - pass belongs to the rules of play.
>In general, all
>actions around the game environment which are explicitly allowed or
>not are "rules", and skillful application of them part of the game.
Like switching off one's PC to disconnect from a server?
>I would not be surprised if he claimed that IGS rules are "flawed".
I shall not surprise you:)
--
robert jasiek
-
> jazzerci...@hotmail.com (-) wrote:
>> This, yet, from the same poster who pronounced:
>> " This is not a code. " ( Tues, Apr 10 2007 )
"Bantari" <ban...@yahoo.com> wrote:
> And what does this have to do with anything? I asked you for a
> CODE and you provided a totally crappy pseudo-code which did
> not even perform to specs.
Eh? Is my function to entertain you? Do you compensate to
me in some way on services performed by me for your behalf ?
CODE I specified merely echoed some performance specs "Bantari"
had suggested as conducting lint-free awareness and quality control.
If it's "totally crappy" then, well (deep question), surely we know why.
How many people do you supervise in your department, "Bantari" ?
> ... I am wondering why [Robert] made a jump from the above position
> to a statement that undo on IGS, which is specifically governed by the
> IGS "rules", is suddenly not a "rule", but only a "feature". What is the
> difference, for him here. Or maybe I misunderstood his previous
> statements. If so, then nothing will get explained if nobody asks.
Robert characterizes a "playing rule" as operative under all
circumstances, as distinct from a "tournament rule" which is highly
context dependent. I'm uncertain whether I can share his optimism.
Is there some Eiffel Tower Standard for Reality by which everything
else may be measured, quantified, and compared? Or are there
components of game theory, even game analysis, which resist a sword
of dimensionalization and remain forever -incomparable- as notions?
Is anything purely context-free (and/or context-independent) or
anything purely context-bound (and/or context-dependent) ? If one
could say everything is connected by one undifferentiated synergistic
whole then one could just as well assert everything is disconnected.
The disconnections connect on both sides of the disconnectors, and
connections are broken by the joins which stich up such connections.
Does not matter whether we can put it together if it's already together.
> I would not be surprised if he claimed that IGS rules are "flawed".
> But his statement that the ability to "undo" and when you can apply it
> is not a "rule" puzzles me.
He's saying server rules are not "playing rules", perhaps. Could make
up anything you please as a "tournament rule" -- even rules governing
use of cell phones and bubble-gum -- which are not "playing rules."
- regards
- jb
-----------------------------------------------------------------
Silence - A Nemesis More Deadly Than Bullets
http://www.rense.com/general76/silence.htm
-----------------------------------------------------------------
Harry Sigerson
> One could say that "cheating" is not possible on the ne... since if
> the rules of the server allow you to do something, it is within the
> rules, and thus not cheating at all, but strategy instead. :)
>
If two players are in a 'live' game - tangible stones and goban
- and they had a dispute they'd settle it one way or another or just
abandon the game.
If such a live game was being played in a competition and the
same dispute arose, they'd refer their dilemma to the referee or
tournament director. He in the interest of the competition
progressing timeously would make a decision - good, bad or
indifferent that's it. Nobody gets shot.
It is to be hoped that he'd come to the 100% correct answer; but
he might not get it that perfect. Nevertheless the players would have
to accept the determination - for the reason stated.
In playing a game online, subliminally, players have the idea
that the Go client is omniscient, in the sense that they each cannot
influence what it does/'decides'. This is both the 'benefit' and the
'penalty' of being able to play someone 'live' who is located on the
other side of the globe.
All you can hope for, as with any computer application, is that
with each version more and more rule anomalies are resolved.
This might be a neverending too-ing and for-ing operation but it
still has to come down to the same thing as in those first two
tangible, stones-in-the-mitt examples.
In an online-game between two players, in that a game is 'lost'
that maybe shouldn't have been, perhaps there is a little ego
bruising in the process - that's not a call for a Glock 19
resolution.
In an online-competition game, the players will have recourse to
arbitration as defined in the tournament rules. Here again the
particular 'Scient' being appealed to, may not be as Omni as either
he nor the player could wish for but that's the deal in the tangibles
so why not the same in the 'intangibles'?
Harry.