hello, can anyone enlighten me how this ranking system works exactly?

i find it rather anoying that if i win 3/4 of mine games the change in mine rating is zero.
in the faq it says its based on the chess rating system, but i believe in real life you always gain ranking even against a much lower ranked opponent like perhaps 2 points but never zero.

does anyone know a detailed way how this is handled

furbster: I know how frustrating that is, but you are lucky because your opponent has proven to be quite reasonable in the past. Perhaps just inexperienced here. I'm sure it will work out.

emmett: I'm rejecting all of the ones he has marked (all mine). I have messaged twice in the game stating none of them are dead. it might be over soon though as i have had the first pass now, hopefully he will do the same and the game might end.

furbster: What stones are you rejecting?

Have you tried messaging your opponent directly, explaining how it is obvious that it is your game, and explain go etiquette a little?

Is there any solution for games like this? I've rejected the same set of dead stones 3 times now. Go 9x9 (furbster vs. FrancescoLR)

onigoroshi: lots of things aren't handled correctly in the .sgf's.
See http://brainking.com/en/ReadBug?bgi=977 for a few of the bugs I've identified.
As well as:
http://brainking.com/en/ReadBug?bgi=1067
and:
http://brainking.com/en/ReadBug?bgi=1082

It seems that handicap stones are not handled correctly in the .sgf export code...

hexkid: Okay, I get it, I didn't realize what I was agreeing to, being a newbie.... Thanks for enlightening me!

jurek: Will he do this to all of the other hundreds of games that were improperly scored, too?
If I were him, I'd say no :)
I think that unless he's asked to change a specific game result all the results will be kept as they are now. It's very tough for a computer to determine the correct score in many games; that's why the players need to mark dead stones and let the computer determine the score from there. If the stones are incorrectly marked dead, the score is incorrectly calculated.

Binabik: it shouldn't be possible for this to happen, am I right?
Well ... you accepted the score.
If you had declined the dead stones the game would go on ... your opponent could mark the stones differently on the next move ... or maybe she'd mark them the same and you decline again ... and again ... and again ... and again :)
The end of game needs to be reviewed by Fencer. There have been quite a few threads about that already.

fakarten: Thank you fakarten and Hexkid for the explanation... you're right, it shouldn't be possible for this to happen, am I right?

hexkid: Will he do this to all of the other hundreds of games that were improperly scored, too?

Binabik: Send a message to Fencer and he'll make it right (once he returns from vacation).

Binabik: Because your opponent incorrectly marked all your stones as dead, and you accepted it.

Hi, could someone tell me why I lost this game?
Go 9x9 (Binabik vs. gene)

joshi tm: Yes, he can go directly to the database and set the game finished with a specific winner (if I'm right Black wins by 1.5 points -- that would terminate the tournament and give you 2nd place).

I'm sure he reads this board, but send him a message.

hexkid: Can't Fencer do anything about that?

joshi tm: Maybe you can *bump* one of these threads of the "Feature Request" board :)

Hey i'm having trouble in this game here Go 13x13 (joshi tm - faith) . My opponent is losing the game and when we both pass, she makes an offer that i must refuse, because she doesn't mark her own stones dead. What should I do?

Would it be in poor taste to equate some of our go and shogi BK ratings into levels of kyu and dan rankings? Is this at all feasible? I know that only the extremely skilled make it to dan levels in go, yet I've always wondered what kyu level I play at. Is this worth discussing on Brainking?

Thank you for your detailed responses. Wow!

emmett: As far as I know, using a komi ("handicaps") is standard, though the values vary.

Sensei's Library has a lot to read about komi among many other Go related articles.

emmett: This is taken from http://senseis.xmp.net/?Komi :
"With the notable exception of the Oteai, almost all tournaments nowadays, both amateur and professional, use komi. But this has not always been so. In fact, komi was rarely used in professional tournaments before 1937, and its gradual introduction into professional play was not without controversy.

The usual komi in Japan was for some time 4.5 points (specified as 4 points with White winning jigo)[4], although even lower values were commonly used early on. In 1955 the Oza became the first tournament to adopt 5.5. Today the standard komi in Japan is 5.5 points, although the Nihon Ki-in decided to change to 6.5 in September 2002.[1] Korea also used to use 5.5, but is already in the process of switching to 6.5.[2]. The usual komi in China was formerly 5.5, but 7.5 is now standard.[3] The Ing rules also have a komi of 7.5, specified as 8 points with Black winning jigo. Western countries often used to follow Japanese practice in using a komi of 5.5 points, but tournaments with komi set at 6 or 6.5 are not uncommon. The New Zealand rules specify a komi of 7. The American Go Association changed komi from 5.5 to 7.5 in August 2004, effective 2005."

I have played go for awhile and have not seen automatic handicaps in white's favor like we have here on BK. Is it de rigeur to feature such handicaps? Do tournament players in Japan or Korea play with these handicaps? (By handicaps, I mean the 5.5 points at the end of our 19x19 games in white's favor, etc...)

Fencer: Thank you.

kaluza: I've just fixed it. However, it was only a visual bug, the real scores were counted correctly.

kaluza: So I see. Sorry Fencer (but it serves as a reminder ).

hexkid: I already reported it on bugtracker a month ago:

http://brainking.com/en/ReadBug?bgi=976

kaluza: You're right. The score is wrongly calculated by the server.

I reported it as bug in the Bug tracker.

When BrainKing calculates points in Go, it does not include green areas (areas captured on the last move).

Examples:

Go 9x9 (kaluza vs. milionovej kluk)
In this game, BrainKing says that black has 15.5 points, but it should be 19.5.

Go 13x13 (Serg vs. kaluza)
Here, BrainKing says that black has 47.5 points, but it's really 61.5.

There's one thing I don't understand: Why can just one player mark dead stones???
It should of course be like that:

1. One player passes
2. The other player passes too
3. The first player should be asked to mark dead stones
4. The other player should either accept the marked stones (that would finish the game) or give the chance to mark dead stones himself.
6. If the first player doesn't accept the dead stones marked by his opponent the game should continue WIT the possibility to pass again.

Ok, I know this setup could lead to a never ending game with both players not accepting the score but you could limit this sequens to maybe 2 times then the game would automatically send to the Go administrator to say who's the winner.

volant: yes i know that but when one player just keeps playing cause it gaves him action points it gets very annoying

volant: and a player should pass when every move he/she makes decreases his/her score.

imsoaddicted: it ends when both players pass their turn. 

could someone explain to me how do you know when a game is finished
and isnt it pointless for someone to continue to lay stones when the game was over 50 moves ago

gringo: The game seems to have been tweaked (by Fencer?). Even though it shows Black with 7 points and White with 9.5 points, the winner is Black; which is correct.

Black wins by 11.5 points.

Something went wrong wih the calculation of this game: Go 13x13 (gringo vs. jitkasafa)

Being lazy I didnt't mark the dead stones as I thought I would have won anyway as the interface shows more points for me. What can I do now?

Chaosu: Try changing the text size in your browser. Open the View / Text Size menu and choose different sizes until, hopefully, it looks right.

hexkid: Thanks!

fakarten: The SGF viewer/editor I like best is Kiseido's. Unfortunately, unlike MultiGo, it doesn't rotate or mirror the board (at least not my old version 2.6.12).

Can anyone recommend a SGF-viewer which will let me rotate and mirror the board?

I find it very confusing to review a game from a different perspective than I played it, but I guess that can be trained.

Fencer: I think there are two different issues here.

The first is that the board orientation is different based on being black or white. For most games on this site (chess, checkers, backgammon, etc), board orientation is of prime importance, but for games like go, it is mostly irrelevant to the game. Most SGF viewers have 1A on the lower left-hand corner. This would account for the rotating that ikkentobi is seeing. Personally, I really don't mind, and I don't see this as a problem.

The second is slightly goofy.
This site has a very good run-down of the SGF file format:
http://www.red-bean.com/sgf/
In particular, this page: http://www.red-bean.com/sgf/go.html
has listed the directionality of the axes slightly different. "aa" is displayed as the upper-left corner of the board (even though that corresponds to 13a or 19a or 9a, depending on the board size) and then continues across and down. So in essence, the vertical number column is inverted, with regards to how it's displayed (a=19, b=18, ..., s=1). This discrepancy would account for the mirroring that ikkentobi is seeing. Again, this isn't *that* big of an issue, but it is inconsistent with the SGF spec.


I also have a wishlist pertaining to the end of the game :)

When the final + move is made and the game is over, the SGF file can have some added tokens to properly show the score. These tokens TW and TB are used to indicate white's and black's (respectively) territory. This also automatically marks dead stones, so nothing extra is needed for that.
The syntax is (as poorly described on http://www.red-bean.com/sgf/go.html): TW[point1][point2][point3]...[point4]
and doesn't include the player's own played stones.
For example, in this game: Go 9x9 (JB007 vs. tenuki)
The output would be:
TW[aa][ba][ca][bb][cb][ac][bc][ad][af][bf][ag][bg][ah][bh][ai][bi][ci][di]
TB[fa][ga][ha][ia][fb][gb][hb][ib][fc][gc][hc][ic][hd][id][he][ie][hf][if][ig]

Fencer: Here ya go:
Game 1: Go (Ceiter vs. ikkentobi) - on my screen this game has A1 in the upper right corner. The resulting BK-SGF is vertically mirrored.

Game 2: Go (ikkentobi vs. stiveletti) - This is an iteresting example. First of all this one throughs an error "Invalid board location "pass" on line 134". Pass should be an empty node like ;B[] or a tt like ;B[tt].
Back to the coordinates. In this game A1 is on my screen at the lower left corner which is right. But the resulting BK-SGF is horizontal mirrored.

FYI:
All SGF-editors that I know have A1 at the lower left corner!
The SGF notation has [aa] at the upper left corner (see http://www.red-bean.com/sgf/go.html)
That means a black move shown by a SGF-editor at A1 should be written ;B[as]

Hope that helps
Tobias

ikkentobi: Mmmm, I don't fully understand what you mean about the coordinates. Please give me an example of a game which generates wrong SGF file and why.
Player names are included. PB and PW tags.

Fencer: Thanx for putting the SGF link. Its a good start but up to now it seems that the code doesn't take care of the changing coordinates directions at BK. So some games are ok, others are mirrored or rotated (actually I don't know why BK does this swapping of coordinate directions).
Like it is now you cannot simply bind the SGF to the coordinates in BK.

Another wish: it would be great to include the players names in the SGF file name.

thx
Tobias

ikkentobi: Looks good, thank you :-)

what would be immensely helpful for teaching is a "load SGF file and start at move 16" for example. I have several people that once I point out a key mistake in the game, we start from there again... currently, we just take turns to get there... manually.. but the above feature would be very cool ;)

Fencer: Here is an example:

(;SZ[19]HA[0]KM[5.5]
PB[Ceiter]
PW[ikkentobi]
;B[dp];W[pd]
;B[cd];W[pq]
;B[ec];W[cj]
;B[ch];W[cm]
;B[co];W[jq]
;B[qf];W[pi]
;B[qc];W[qd]
;B[pc];W[od]
;B[rd];W[re]
)


The file on top was created with my proggy.

(; - it starts with an open bracket and a semicolon

SZ - Bordsize
HA - Handicap
KM - Komi
PB - Black Players Name
PW - White Players Name

After the game properties comes the move list.
Every move is seperated by a semicolon.
You could write:

;B[dp]
;W[pd]
;etc.

B stands for Black, W for white
The small letters are the bord coordinates:
aa would be the upper left corner and
ss the lower right on a 19x19 bord.

) - the file ends with a closing bracket

For more information see:
http://www.red-bean.com/sgf/

Fencer: I wrote a script to transform the move list into an SGF file. It needs a few twists, but as I don't use it often, I haven't twisted it. You can see the PHP source or use it to convert any public game to SGF by going to http://hexkid.info/BK/ParseGame.php?g=##### where ##### is the game number.

I've also integrated this functionality into BrainKing itself with the help of a Greasemonkey script, accessible at the user scripting fellowship :)

ikkentobi: It doesn't work for me. Each time I copy and paste the moves, it shows an error message that something is wrong with the move list.
Anyway, I just wanted to use it to generate a sample SGF file because I didn't fully understand the format description. Can you provide a sample game please?