Life In 19x19

Hi Jeromie,

the player with more kings of their color on the board wins.

I see post 8, now.

But the bottom line is still determined by the number of kings.

EdLee wrote:Do you mean any pawns and territories have no effect on the scoring to determine who wins ?

Example: W has ( 100 points, 5 kings ), B has ( 10 points, 6 kings ) -- B wins ?

EdLee, yes, a player's score consists of their number of kings only. In your example, if both players have passed in succession, Black wins by one point, but White should have filled their territories with kings before passing, which would have won them the game by about 89 points.

EdLee wrote:It occurs to me the komi is a mystery ( pure pawns ? pure kings ? mix of the two ? )
If pawns have zero effect on the scoring, then it's meaningless to have pawns in the komi.

Pawns have indeed zero effect on the scoring, so komi can be thought of as a number of kings.

I assume its value should be about the same as it is under territory scoring, i.e. about 6.5 points. While territory scoring counts territory plus prisoners, Kingo effectively counts territory (minus group tax) plus kings, and it's not clear to me that there should be fewer kings (before filling territories with them) in a game of Kingo than there are prisoners in a regular game.

(Of course, we could just say that scoring in Kingo is territory plus kings, which is probably more practical and gets rid of the implicit group tax. I kind of wanted to keep things simple at first, but I will change it if people think it's better this way. Also, as I said, my personal preference is to have integer komi and solve ties with the button.)

lightvector wrote:

$$B What is the result?
$$ ------------------
$$ | . W X . O a W W . |
$$ | B W X . . X X W . |
$$ | B W X X X X W . W |
$$ | B W W W W W . W W |
$$ | B B B B B W . W X |
$$ | . . B . B W . W . |
$$ | O O . B B W . W B |
$$ | X . O B . B B B B |
$$ | . O O . B . . . . |
$$ ------------------

Click Here To Show Diagram Code

[go]$$B What is the result?
$$ ------------------
$$ | . W X . O a W W . |
$$ | B W X . . X X W . |
$$ | B W X X X X W . W |
$$ | B W W W W W . W W |
$$ | B B B B B W . W X |
$$ | . . B . B W . W . |
$$ | O O . B B W . W B |
$$ | X . O B . B B B B |
$$ | . O O . B . . . . |
$$ ------------------[/go]

Interesting. It seems at a deserves serious consideration.

luigi wrote:

lightvector wrote:

$$B What is the result?
$$ ------------------
$$ | b W X . O a W W . |
$$ | B W X . . X X W . |
$$ | B W X X X X W . W |
$$ | B W W W W W . W W |
$$ | B B B B B W . W X |
$$ | . . B . B W . W c |
$$ | O O d B B W e W B |
$$ | X . O B . B B B B |
$$ | . O O . B . . . . |
$$ ------------------

Click Here To Show Diagram Code

[go]$$B What is the result?
$$ ------------------
$$ | b W X . O a W W . |
$$ | B W X . . X X W . |
$$ | B W X X X X W . W |
$$ | B W W W W W . W W |
$$ | B B B B B W . W X |
$$ | . . B . B W . W c |
$$ | O O d B B W e W B |
$$ | X . O B . B B B B |
$$ | . O O . B . . . . |
$$ ------------------[/go]

Interesting. It seems at a deserves serious consideration.

Changed the quote to add b,c,d,e
To me it seems there are 5 endgame moves left:
b is a usual dame that is always worth a point(2 points gote) in area/stone scoring
c is worth 3 points gote as opposed to 4 pt gote with usual area scoring, because the black pawn isnt worth anything if it stays on the board(capturing pawns is worth 1pt less than we are used to in area scoring)
a and d are also worth point(a 2points gote, d 1 pt gote), cause by making(real) eyes with the dead group you force more nakade during taking out, that have to be pawns.
e is a usual andgame move worth 2 point sente or 4 point gote, depending on whether white answers(here he maybe should, because this is the biggest move, so its gonna be 2pt sente).
Is my analysis correct?

My suggestion for best endgame would thus be(4 and 5 are actually miai, I believe):

Click Here To Show Diagram Code

[go]$$B Best play?
$$ ------------------
$$ | 5 W X . O 4 W W . |
$$ | B W X . . X X W . |
$$ | B W X X X X W . W |
$$ | B W W W W W . W W |
$$ | B B B B B W . W X |
$$ | . . B . B W 2 W 3 |
$$ | O O 6 B B W 1 W B |
$$ | X . O B . B B B B |
$$ | . O O . B . . . . |
$$ ------------------[/go]

where 1 to 5 are kings counting it 36:37 for white(area score would be 41:40 for black, but black has 3 pawns in his area(one on the right and 2 after taking out the lower left and white has just 1. Additionally subtract 2 points each for group tax)). 2 at 3 also seems to lead to a 1 point win for white

Yeah, I think so. I found it intriguing both that adding "death in gote" moves could be worth points, and that despite the area-scoring-like nature of the game, you could have moves of miai value 0.5.

I guess it would be a better problem you replaced one of the white kings at the bottom right with a black king, so that black's good play can win the game.

luigi wrote:(Of course, we could just say that scoring in Kingo is territory plus kings, which is probably more practical and gets rid of the implicit group tax. I kind of wanted to keep things simple at first, but I will change it if people think it's better this way. Also, as I said, my personal preference is to have integer komi and solve ties with the button.)

I've finally decided to make this change. Having to fill territories with kings seems impractical.

I've edited my original post to reflect this.

luigi wrote:

luigi wrote:(Of course, we could just say that scoring in Kingo is territory plus kings, which is probably more practical and gets rid of the implicit group tax. I kind of wanted to keep things simple at first, but I will change it if people think it's better this way. Also, as I said, my personal preference is to have integer komi and solve ties with the button.)

I've finally decided to make this change. Having to fill territories with kings seems impractical.

I've edited my original post to reflect this.

Click Here To Show Diagram Code

[go]$$B
$$ +----------------
$$ | . B O . . O B .
$$ | B B O . . O B .
$$ | . B O O O O B .
$$ | B B X X X X X .
$$ | . . . . . . . .[/go]

How many points for black in this corner?

HermanHiddema wrote:

$$B
$$ +----------------
$$ | . B O . . O B .
$$ | B B O . . O B .
$$ | . B O O O O B .
$$ | B B X X X X X .
$$ | . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ +----------------
$$ | . B O . . O B .
$$ | B B O . . O B .
$$ | . B O O O O B .
$$ | B B X X X X X .
$$ | . . . . . . . .[/go]

How many points for black in this corner?

20 points, it seems. Black's first three plays in White's eyespace need to be pawns.

I actually prefer the original rules for this variant. I don't think the group tax is onerous, and it seems natural because of situations like Herman's example where players are forced to capture groups that would normally be removed as dead.

Also, I think it makes an interesting strategic tradeoff: do I play for early secure territory so I can start playing kings or build a large moyo in the hope I can end up with fewer groups?

luigi wrote:

HermanHiddema wrote:

$$B
$$ +----------------
$$ | . B O . . O B .
$$ | B B O . . O B .
$$ | . B O O O O B .
$$ | B B X X X X X .
$$ | . . . . . . . .

Click Here To Show Diagram Code

[go]$$B
$$ +----------------
$$ | . B O . . O B .
$$ | B B O . . O B .
$$ | . B O O O O B .
$$ | B B X X X X X .
$$ | . . . . . . . .[/go]

How many points for black in this corner?

20 points, it seems. Black's first three plays in White's eyespace need to be pawns.

Yes, so you need to be careful in specifying what is or isn't territory.

How about this situation:

Click Here To Show Diagram Code

[go]$$B
$$ +----------------
$$ | . B O O a O B .
$$ | B B O . O B B .
$$ | . B O O X X . .
$$ | B B X X X . . .
$$ | . . . . . . . .[/go]

In isolation, when there are no ko-threats, black can play a king at ''a''. But if there are unremovable ko-threats, he can't.

Perhaps it should required to capture all dead stones, and only after that we count territory?

I also like the old rules, they are nice and clear. Assinging Points to certain stones on the board doesnt seem to fit with territory scoring. I am with Jeromie on this one.

jeromie, HermanHiddema, Schachus, thanks for your feedback. Those are good points indeed.

Even in regular Go, it's not so easy for me to choose between stone scoring and area scoring. I currently prefer area scoring, but not by much. As a game designer, I usually strive for short rules and conceptual simplicity.

In Kingo, it's dawning on me that the case against counting territories is stronger than in Go. The implicit group tax probably doesn't stand out as much given that there are situations of the type Herman points out where some territory points other than the two eyes have to be excluded from counting as well.

I'm now inclined to revert to the original rules, but I'll wait a couple of days in case someone wants to share their perspective as well.

HermanHiddema wrote:Perhaps it should required to capture all dead stones, and only after that we count territory?

I'd never consider defining dead groups in the rules, so the idea would be to let players decide which groups should be removed at the end of the game. The owner of a dead group will never agree with having it directly removed if they know that capturing it requires the opponent to place some pawns in what would otherwise be territory for them.

Of course, a territory of a player's color would be strictly defined as a set of connected points with no adjacencies to pieces of the opposite color, regardless of ko situations and such.

I think it would make sense to move this thread to the "Other versions of turn-based games" forum. There are many posts about Go variants there and the overwhelming majority of posts here are about "actual Go".