Stone scoring converting

[Krama]

We all know that if the number of groups is not the same stone scoring method yields a different result from modern area or territory scoring.

I was thinking about converting score from stone scoring to area scoring (since both don't count captured stones)

Is there any formula as such that would enable us to easily transform from one result to the other?

Formula should probably compare the number of groups of both colors and somehow implement this.

[RobertJasiek]

Beware of complicated sekis.

[Krama]

Without going into sekis (I didn't go into that) I think I figured how to convert it to our japanese/chinese score system.

Once the game is over (even dame is filled) we keep playing until we make 2 eyes for each group.

Keep in mind black has to place 8 stones in their territory.

Then if player A has let's say 4 groups in the board and player B has 2 groups, player B has to pay the difference by playing 4 stones in their territory (2 stones per extra group opponent has).

Then you also need to pay the difference in prisoner count. So if player A captured (not during the game but after the dame phase where you fill in to make 2 eyes) 18 prisoners in total and player B captured 11 then player B has to play 7 stones (18-11) in their territory.

My suggestion is after the dame is filled to simply keep the order of playing and capture opponent stones in your territory and then pay komi, group difference, and prisoner difference.

Now without messing up the play order the last player that can't make a move lost. If opponent can't play any other move after that (without ruining 2 eyes of any group) then that player won by 1 point.

Example: White can't play any move in their own territory without taking one of 2 eyes of one of their groups, white lost.
If black then can't do the same, black wins by 1 point. If black can still play let's say 3 moves then black wins by 4.

It is kinda complicated and I am still playing with it but if you are interested simply download finished games from go4go or something similar and play out yourself.

The last game I checked it with was 8 komi game where black wins by 1 point.

[xed_over]

Is there any formula as such that would enable us to easily transform from one result to the other?

Simply fill in all the eyes (like Ing fill-in counting).

[Krama]

Is there any formula as such that would enable us to easily transform from one result to the other?

Simply fill in all the eyes (like Ing fill-in counting).

Not sure if I follow.

Also I just discovered that if black can't make any moves (black lost) you count the number of moves white can still make and reduce it by 2 to get a true score. No idea why it's -2 if white wins and +1 if black wins

[RobertJasiek]

Krama, your conversion does not produce the intended result. If you truely convert, then you do not have territory scoring (like in Japanese rules) any more but you have area scoring (like in Chinese rules). If you do not truely convert (omit the condition that White passes last), then you modify territory scoring by introducing pass-fights (fights about not passing first).

http://senseis.xmp.net/?PassFight
http://senseis.xmp.net/?PrisonerCountingForStoneScoring
http://home.snafu.de/jasiek/rules.html

[Krama]

Krama, your conversion does not produce the intended result. If you truely convert, then you do not have territory scoring (like in Japanese rules) any more but you have area scoring (like in Chinese rules). If you do not truely convert (omit the condition that White passes last), then you modify territory scoring by introducing pass-fights (fights about not passing first).

http://senseis.xmp.net/?PassFight
http://senseis.xmp.net/?PrisonerCountingForStoneScoring
http://home.snafu.de/jasiek/rules.html

I download the games from go4go that are finished and have a score set then I use my method and it's always correct.

This method works however if the players used a chinese ruleset not japanese one.

[RobertJasiek]

See AGA rules for your method. Like Chinese rules, they use area scoring so it works.