9x9 and the tactic of Black creating Ko threats

I was contemplating in game how few times I’ve ended up in Ko while playing 9x9 and this partially has to do with my level, I suspect, and partially because the board size is small enough that unless you were deliberately trying to create a Ko it wouldn’t be likely to come up.

However, I was recently in a game where there was a Ko, early in the game, and I realized that as Black, I was gaining an advantage with each capture. To understand why this is you have to realize that White normally starts with Komi, a number of stones captured before the game even starts. With every stone exchange from a Ko fight, the Komi’s significance is reduced. As the captures started to end up into double digits, the Komi became less than half the captured stones. Still significant, but with each Ko capture less so.

Has anyone postulated that Ko threats are advantageous to Black because of this? I’ve never seen any mention it before but Go is deeply researched so I can’t be the first to think about this.

Any links to articles about this concept or comments would be welcomed.

But in the end, with each exchange of the ko, both black and white gain one prisoner, so in the end the difference between the captured stones remains the same. If I’m 2 points ahead of you, and I recapture the ko, I’m now 3 points ahead, but then you play a threat, I respond, and you recapture, which brings it back to me being 2 points ahead (and out of ko-threats, dang it!).

The ratio between the number of prisoners and the number of komi doesn’t matter, it’s all about the difference between black and white. Giving both players 100 points “for free” doesn’t change the score.

And as a side note: this also does not depend on the board size, although I feel that komi should be higher on 9x9 than on 19x19, since the first move on a small board has more impact than on a large board.


Yeah, that makes sense. If the result for a Komi 6.5 game is White 106.5 to Black 100, it’s the same as White 16.5 to Black 10.

The proportion doesn’t really matter.

1 Like

Detailed article on Ko: http://home.snafu.de/jasiek/ko.pdf

Ko calculations: http://gooften.net/2016/02/01/sente-gains-nothing-part-3-ko-calculations/#more-1926

The two free sources I know of. Now if you could get your hands on “Mathematical Go” (1994)…


There’s a lecture on youtube by Berlekamp about the things he writes in ‘Mathematical Go’


Now looking for that on YouTube wouldn’t even have occured to me. Thanks for the link. :slight_smile:

Horrible title though.

1 Like

I don’t remember how I stumbled upon it either, to be honest…