9 x 9 Go Is A Real Game - Part I

British Go Journal No. 66. November 1985. Page 22.

Francis Roads

Beginners walk into a Go club and see everybody playing 19 x 19 Go, so naturally they want to do likewise. It is often very had to persuade them that they will learn a whole lot more about the game playing six 9 x 9 games in the time that they would have spent playng just one on the full board.

In order to raise the status of the 9 x 9 game, I wanted to write some articles for this journal - but where to find a supply of game scores? I decided to invite ten freinds to visit my house for an all-play-all 9 x 9 tournamnet. All the games were recorded, so I now have a stock of 55 games for commentary (Does anybody want any?).

The range of strengths was 6dan to 4kyu, so a handicapping system was needed. We borrowed the system used bythe Dutch for their 13 x 13 tournamnets. There was no reason for thinking it would work especially well, but in the event Black won 26 games and White 27, with two jigos (draws). This is rather more equitable a result than you get with the normal full board handicapping system, which tends to favour White.

The way the system works is to assume that komi is nine points, and that a one stone handicap represents a difference of three dan or kyu grades. So for an even game White gets nine komi; with one grade difference the stronger player (as white) gets only six points komi; with two grades it is three and with three grades, zero.

Theoretcally, for a four grade difference we now need two handicap stones with six komi for white, but in order to reduce the need to place handicap stones, the players preferred to play an even game with black receiving three points komi.

Five grades difference is a two stone handicap (using the 3-3 points) with three komi for white, six is two stones with no komi, seven is two stone with three komi for black and so on. If you regard five stones as the maximum sensible handicap, you can accomodate a difference of 16 grades in playing strength.

Now to the first game. I have chosen it not because I was playing, but because it raises some interesting questions about the endgame.


Black: Francis Roads, 3d
White: Jim Barty, 4d
Komi: 6 points
The game-file in SGF format.

EBGJ has 'Figure' and 'Diagram' where BGJ had 'Diagram' and 'Figure' respectively.

Figure 1 (1-5)








Diagram 1








Fig 1. I was surprised to see white 4, this looks like aji-keshi (committing yourself too early), coming at the start of the game. After 5, white is welcome to push through at A if he wishes, as in Dia 1. Black 3 turns out to be a profitable sacrifice.

Figure 2 (6-14)








Fig 2. After white has established himself in the lower corner with 8, it becomes worthwhile preventing the sequence of Dia 1 with 9, which also forces white to defend his corner with 10. I could have separated the two white groups with 11, but playing this large territorial move instead seemed to me to be enough to win, even though White was able to link his groups at 14. This concludes the middle-game.

Figure 3 (15-28)








Diagram 2








Fig 3. Black makes various sente endgame plays up to 23. White is right not to answer this move at once, but to play 24. Why? Well if black ignores 24 to capture at 26, white can jump to B and black loses far more than he gains in the lower left. So black answers at 25, leaving white free to go back to 26.

If black were to ignore white 24, Dia 2 shows a possible outcome. black captures some stones on the lower side, but white is threatening both C and D and is several points ahead (remember komi).

Figure 4








Fig 4. There are three remaining endgame plays at A, B and C or D.

[Start] Can you work out the final moves before turning to the end of the game on page 27?


This article is from the British Go Journal Issue 66
which is one of a series of back issues now available on the web.





Last updated Tue Jan 01 2013. If you have any comments, please email the webmaster on web-master AT britgo DOT org.