When I play 5x5 with 24 komi, even if black kills white the score is going to say white won. Is it just understood black still wins? I guess i never noticed because if I loss as white I usually resigned. Is their no remedy for this?
Play 7x7
4 Likes
(assuming Japanese rules) So the simplest game might be Black plays, both pass, black has 24 points on the board, white has 24 komi. The game is a tie.
Is there an issue with that result? It shows up like that right?
If black and white both play 17 additional stones for example, and all the white stones are dead, then black has 17 prisoners + 7 territory (empty on the board), again 24 points to whites 24 komi, still a tie.
2 Likes
So why is komi so high if this issue arrives?
Are you not choosing the komi yourself?
I’m not sure I understand the question. If you want a “fair” komi, then you need to offset blacks advantage of the first move. If with best play there’s no way that white can make a living group, then the fair komi should be something like 24 if you want a tie with best play, slightly more like 24.5 if you want white to win, or slightly less like 23.5 if you want a black win.
If you don’t like draws then there possibly isn’t an exactly “fair” komi, but it’s supposed to give either player a reasonable chance to win, given the game is sufficiently deep that players can’t find optimal moves from the start and can both make mistakes.
That’s really all there is to it. Do whatever you want on 5x5. Or go to a larger board size which is deeper.
https://lifein19x19.com/viewtopic.php?f=15&t=17750
^^ fair komi
^^ ways to “balance” two player games. There’s other ways than komi if you do want to make 5x5 more interesting.
3 Likes
To be overly pedantic, the term “perfect komi” is often used for the “best play” definition. “Fair komi”, in contrast, takes the players’ abilities into account. For example, two 30k may reasonably pick a fair komi much less than 24 because Black is not skilled enough to consistently capture the entire board.
To be overly pedantic about you being overly pedantic, if it was a tournament situation, or even just a ranked game on a server like OGS, you wouldn’t always take the players ranks/abilities into account. You just have to go with the “perfect komi” if that’s what you want to call it instead because that just might be the rules. If anything “fair” would only make sense in casual situations then, and you could equally say that you’ll give me three stones if I buy you a beer and that seems fair.
I’m not sure I’d want to reserve the usage of the word fair for something that can’t even be worked out, and just depends on the time of day and on the opponent, but there we are.
(feel free to point out the distinction here also AI opinions on "fairest" komi by board size • Life In 19x19)
2 Likes
Here is an example. This tie is odd because white was utterly destroyed, and i he threw in less stones white would have won simply because of komi. How is that a win?
Because that was the rules that were agreed to.
It feels a bit philosophical really to debate whether something is really a win or not. Like if you never take the king off the board in chess with checkmate, you never capture the opponents king just give them no escape, is that really a win? It’s almost like a premature ending to the game. Is perpetual check really a draw? What if one side
the other side but they were saved by perpetual check or the 50 move rule etc.
The point is you can debate the rules, change the rules etc, but agree to play with a certain set of rules then a tie is a tie, a win is a win, a loss is a loss.
You can always choose to play with less komi.
I think you can have similar debates about 9x9 go with 4.5, 5.5, 6.5 komi etc it’ll change some board positions from wins to black to wins to white and you can debate why or why not they should be wins for one side or the other, but once you pick a komi you just have to adapt your strategy accordingly.
1 Like
They seem to have already acknowledged the distinction:
Firstly, from separate earlier studies, we know (probable) perfect komi for smaller sizes:
Code:SZ Japanese Chinese 3 8 9 4 0 2 5 24 25
1 Like
Not trying to debate, just trying to understand if thats the agreed upon komi because how is black suppose to win in this situation? If the komi is to get as close to 50/50 how is black overcoming a 24 komi on that board?
Looks like it’s been used interchangeably to me “fair” according to Katago and perfect.
I don’t really mean to debate either.
I guess what I’m saying is that in that specific situation if you count a draw as like winning 50/50, then the tie with 24 komi is as good as it gets.
If you don’t want draws, at least if the players play well enough you might never get 50/50 with any komi, it might go from 90% wins as black at 23.5 to 100% wins as white with 24.5.
I think the game needs a bit more depth, so like 7x7 like benjito me mentioned or 9x9 where players can reach certain winrates with evenly matched opponents and the right komi.
2 Likes
I think it’s simpler to use Chinese/AGA rules than Japanese rules for an “all-or-nothing” situation like this.
Playing with 24 komi on 5x5 with Chinese rules:
- If Black controls the board and White has not surviving stones, then Black wins by 1 point
- If White has at least one surviving stone, then White wins.
With Japanese rules it’s a lot less clear. Black might lose a few points by playing inside their own territory, and lose even though all White stones are dead.
Another popular version of 5x5 is: On the first move, Black is not allowed to play at the centre point. (From the second move on, normal rules apply). Then both players will manage to have living groups and a more “balanced” game of go ensues.
2 Likes
I think the crux here is that 5x5 is simply too shallow for komi to make much sense. The benefit of going first is so large that any attempt to fairly compensate it results in a very large komi.
As others mentioned, if this bothers you then either:
- agree with your opponent to use another komi on 5x5, even if it means black has the advantage;
- play on larger boards (at least 7x7).
3 Likes
Building off of @ArsenLapin1’s suggestions, what about setting 24.0 komi area scoring (Chinese, AGA, NZD, Tromp-Taylor, &c.), and using the swap-2 opening protocol? it then turns it into a sort of competitive tsumego match, where Slicer attempts to use 2 black and 1 white stone to create a tough tsumego problem for both sides?
For reference, swap-2 is an opening protocol used in Gomoku and some other games:
- Slicer plays 3 moves (ply)
- Chooser either picks a color and the game continues with white playing next, or plays two further moves (ply), in which case Slicer has to pick a color and the game continues with white playing next
I’m not convinced 5x5 or even 7x7 offers much that 9x9 doesn’t, though
1 Like
5x5 gets more interesting if you forbid opening on tengen. I think this site implies 3 komi is perfect in Chinese rules, and the play is complicated (note “Even Cho Chikun missed this one”):
http://erikvanderwerf.tengen.nl/5x5/5x5solved.html
My copy of KataGo agrees. With Japanese rules it seems to be calling it an even game with 1 komi (Japanese rules) and plays it out into a fun seki.
2 Likes
Thankyou, this is sort of what I was looking for, remedy to the situation. I want to get a 5x5 goban, I will also play capture go with my kids.
3 Likes