rokirovka: **((no subject))**
(2009-01-07 22:10) [#5606]

rokirovka: I suppose this is as good a place as any to express my frustration at the lack of a good algebraic notation for go moves. Kifu and diagrams with numbered stones may be fine for recording games, but I have two problems with them.

First, it is cumbersome to have to have a diagram to record any move or sequence of moves. In chess, I can look at one initial diagram, then read a line of text such as "Here, White plays 1.Ng5 g6 2.Qg4," and I know exactly what this means without the text moves having to be marked in the diagram. But in kifu or numbered stone notation in go, either the initial diagram is cluttered with the three extra stones, making it harder to see the initial position, or a second diagram is needed just to describe the three new moves. It is true that one can use the initial diagram with the letters a,b,c marking the points where the next three moves are to be played. But this gets confusing when one wants to describe multiple possible continuations of several moves each from the initial diagram. In chess, one can write "Here, the game can continue 1.Ng5 g6 2.Qg4, 1.d4 h6 2.Qd3 g5 3.Nxg5, or 1.Bd3 g6 2.h4 f5," and I can look at the initial diagram and easily follow exactly what the text is saying. But on a go diagram of the initial position with lower-case letters marking the points where the next stones are to be played, describing three variations with three or four moves each becomes very confusing to follow on a single diagram. I know if I tried that system for the above chess example, I would get lost with letters a,b,c,d,e,f,g,h marking the squares g5,g6,g4,d4,h6,d3,h4,f5.

Second, when I calculate go moves in my head, I like to have a name for each point that each stone will be played on. It helps me tremendously to visualize a sequence of moves in my head in advance, if I can say to myself, "Black plays A2, then White plays (B)1. Now Black E1 fails to White C1, and Black C1 fails to White E1. So Black F2 threatening F1 with a straight five in the corner and life. Then White F1, Black G2 threatening G1, White G1, and now Black E1, White C1, Black H1 capturing F1 and G1, and next either G1 or H2 with a second eye and life for the Black group." Unfortunately, I had to put B in parentheses above because this text editor automatically translates B followed by a number as a numbered black stone. (This example is from a variation that I added to the solution page for Beginner Exercise 13 in the life & death problems.) I don't need to look at a diagram with numbered stones or lettered points to follow this analysis; I can follow it in my head while looking at the initial position on the board. With an algebraic notation I can do this, but with kifu I cannot.

The problem with the existing algebraic notation for go is that while the letters and numbers A,B,C,D,E,F,G,H,J,K and 1,2,3,4,5,6,7,8,9,10 are natural to identify with points on the board in one's mind, the letters T,S,R,Q,P,O,N,M,L and numbers 19,18,17,16,15,14,13,12,11 are much less natural, especially since one is counting and alphabetizing backwards as the points go away from the corner and sides, the opposite of the A,B,C... and 1,2,3... points. The double-number and double-letter notations are even harder to keep clear in one's head when there are more than two or three moves. For example, if I memorize a joseki as Black 3-4, White 5-3, Black 7-3, White 5-5, Black 4-6, White 8-4, the numbers are already blurring together indistinctly in my head. It's much easier to memorize Black C4, White E3, Black G3, White E5, Black D6, White H4. That I can visualize in my head. The problem is that Black R16, White P17, Black N17, White P15, Black Q14, White M16 is a blur again with the bigger numbers and letters in the middle of the alphabet and everything going backwards.

Thus, I personally find it much easier to follow an algebraic notation where all four corners are marked as A,B,C,D,E,F,G,H,J and 1,2,3,4,5,6,7,8,9 and asterisks are used to mark the right side and the top side. K and 10 of course never need asterisks. Thus I can notate the above joseki in the top right corner as Black C*4*, White E*3*, Black G*3*, White E*5*, Black D*6*, White H*4*. I say the moves in my head as "Black C-star-4-star, White E-star-3-star," etc. As an added bonus, I can instantly recognize a tenuki in a different part of the board because the asterisks are different. If I am analyzing a position in only one corner of the board, I can leave out the "stars" as I note the moves in my head as I calculate.

Here is an example of my notation recording the beginning moves of a game I like, Cho Chikun vs. Kato Masao, 7th Kisei, Strongest Players' Final, Game 2, 16 December 1982. See how far you can follow the opening moves in your head as you read them. (Again, I unfortunately had to put B in parentheses when immediately followed by a number so that this text editor did not translate it as a numbered black stone.)

1. C*4* D4*

2. D*3 C3

3. E*3* C*5

4. C*7 E*5

5. F*4 D*8

6. C*4 B*4

7. D*7 F*5

8. G*4 E*7

9. E*8 C*8

10. F*7 E*6

11. E*9 C*10

12. B*3 D*10

13. B*5 C*6

14. C6* F4*

(tenuki!)

15.(B)4* C3*

16. C9* J*3*

17. F4 C6

18. H*6 H3

19. K3 E*4

20. E*3 K4

21. J*3 G5

22. D4 C4

23. D6 C7

24. G4 H4

25. H5 G6

26. J5 G2

27. D3 E2

28. D7 D8

29. E8 D2

30. E9 F7

31. C8 (B)8

32. C9 C*6*

33. D5

fractic: **audouard coordinates**
(2009-01-07 22:14) [#5607]

Kifus work fine for me as a notation. Its very readable if there are no more then 10 moves in a diagram and very compact when recording an entire game.

Anyway you might be interested in this: Audouard Coordinates.