Home > Steve > Go > Tools > Seki Table
Recently (as from Jan 1998) I have been frustrated by the confusion
generated by sentences like
"If Black 4 at 5 then White 5 at 17 and Black replies with 6 at
4."
Because of this I have started using a single tick mark to annotate
these sentences. Such a sentence is usually associated with a particular
diagram. I use "normal" numbers for points as numbered on the
diagram and I use numbers postfixed with a single tick (apostrophe) for
move numbers which differ from that shown on the diagram. The example
sentence then becomes:
"If Black 4' at 5 then White 5' at 17 and Black replies with 6' at
4."
Which I reckon is an improvement.
If you write about Go, why not use the same nomenclature?
For those of you with PCs running Windows. I can recommend Denis Lambot's Yago (v0.9) for printing out board diagrams for your records. It works with either Ishi format or SGF format files (I've only used it in Ishi mode), but it doesn't convert between them.
It is also good for entering games and editing and commenting them. It is quick and it recognises and correctly handles prisoners and variations.
After using Yago for nearly 9 months, I discovered the marking menu down "Properties". Very useful for putting numbers, letters or symbols on stones, and more importantly removing them if you change your mind.
All the bugs are minor.
It gets confusing if you have two games open at once. The mix of windows is wrong. It can be solved using the Window menu, but it is easier to only open one at a time.
Empty comments don't get deleted.
It is undocumented that you use the Delete key on the keyboard to delete either the last move of a branch or an entire variation.
It is not possible to delete added setup stones at the start of the game. Added setup stones part way through a game do not get saved.
It handles labels in SGF wrongly - each label is backwards and the files are thus incompatible with other SGF programs.
I've bought John Fairbairn & T.Mark Hall's DOS GoScorer
for playing through (professional) Ishi format games. It gives you hints
and up to 10 guesses where each move is. If you guess each move right
you get 10 points, on second guess 9 points etc.
(A Windows 95 version is now available.)
It seems a good way of making yourself study a game rather than just scrolling through it, and it is cheap (£5 when I bought it in 1997).
It suffers from being very fussy over the format of the games it reads, it won't read Yago Ishi format files for example. See Procishi below for my solution to that problem.
I've also bought Invincible and plan to enter the games in Ishi format. (After all, T.Mark reckons he got significantly stronger just entering all those Go Seigen and other professional games). After that I can replay them with GoScorer.
I've bought Thomas Wolf's DOS program for solving life and death problems in a board segment. It works well, although I often enter problems which are too hard for it on my 486/66. The interface is a bit clunky but easily liveable with.
I've tried v1.2 and v1.3. Each has advantages and disadvantages. I use the program frequently to assist with my monthly problem page.
Since I frequently need to renumber and process Ishi format *.GO
files whilst preparing these pages and since GoScorer is very fussy
about its *.GO files, I have written a windows application to process
these files. Basically this program massages an Ishi format file to make
it compatible with both Yago and GoScorer. It can also
renumber moves and variations which is useful for problem / answer
diagrams and for mal-recorded games. It can also promote variations
(exchange the main branch with a variation).
Version 3 (27dec98) includes a reprocess button.
Procishi is a Windows application written in Visual Basic 3 and tested on Windows 3.11.
This is a primitive DOS QBasic program to permit the rapid
entry of moves from a paper game record onto a PC. It needs to be used
as one part of an awkward process, but is easy once you get the hang of
it.
QBasic is supplied on Windows 95 CDs, but is not normally installed.
You'll need to have it installed to use Entergo.
Version 47 can create Ishi format files, HTML gif diagrams using a mixture of whole-point and half-point gifs and ASCII diagrams for emails. All can including lettered and numbered intersections with optional board coordinates. It can also read in the main branch of an Ishi format game for editing and can accept moves as a list of coordinates.
A brief description of the method I use (under Windows 3.11).
Use Yago to create the problem. Also use Yago to investigate the various answers and to make brief comments on the sequences.
Plan the page and use Procishi to ensure that the move numbers in
each diagram will be correct. Usually this involves the use of "Number
Vars from Move Number" and of "##nn" comments - (this won't mean a thing
to you unless you have read PROCISHI.TXT).
It is also likely that some variation will need promoting, so that the
main stream of the file is the correct response whilst the variations
are deviations from that. Again this is done with Procishi.
Use Yago again to mark the stones with numbers, symbols and letters as is required. Then for each diagram, use Yago's "Print to Clipboard" option.
Paste the clipboard into Paintshop, specifying an image size of 400
* 400. "Rectangular select" an area of the board approximately, then
double click the "Rectangular select" button and round all the numbers
to sensible values. It turns out that the Goban grid lines are all on
multiples of 20 pixels. CTRL-C to copy the selection. CTRL-V to paste
it. Use "Save as" to save the new image as a GIF file in the relevant
web upload directory.
Repeat as required.
The rest of it is now standard HTML editing.
I record some of my tournament games. When I do I like to transfer them to the PC for "posterity"! My records are often poor with numbers missing or duplicated, so it requires some fancy footwork to get it entered quickly before my memory of the games disappears.
I use Entergo to enter the moves, row by row (a19, b19, c19, ..., t19, a18, ..., t1) using just Enter for an empty point or typing the number for a move. Duplicates beep to warn you have a problem. Omit duplicates and note them on paper. Once the bulk of the board is entered, define locations of oddball stones (ko fights, throw-ins etc) and save the primitive Ishi file.
Use Yago to read the file, to add the players names etc and to add variations in the vicinity of duplicate stones to get duplicates (wrongly numbered) but in the file. Save the file.
Use a text editor (Notepad? I use PFE - a very good editor) to remove the VAR and ENDVAR bits so that the (wrongly numbered) moves are in sequence.
Now use Procishi to renumber the moves, and Yago to play through the game to check it is all ok. Save it from Yago to get the prisoner information added to the file.
This may sound involved but it works quite smoothly. For example, After the Coventry 97 tournament I was able to enter my games that evening, email one to be analysed the next morning and (praise to the BGA analysis service) I received the commented game back that evening, 26 hours after the game finished.
In the British Go Journal, issues 102 to 107, Richard Hunter wrote an excellent series of articles on counting liberties and the associated lifes, deaths and sekis that result. In issue 108 he started a series applying what had been expounded in the previous articles.
In order to understand these, you need to know what liberties are applicable. Based on R.H.'s original, I have created a table which I print out, fold and stick inside the back cover of each relevant BGJ. I can then unfold it to refer to whilst reading the articles.
With my browser settings, after printing the table on portrait A4 paper, I fold the left side up reach the table line between 'Type' and 'Description' and fold the right side down along the table line between 'Favourite' and 'Underdog'. This foldout table can be glued 1mm (or so) in from the right hand edge of the inside of the rear BGJ cover.
When discussing sekis, standard literature refers to the liberties which form a gap between the opposing groups as shared or mutual . Richard Hunter dislikes this since they are not 'shared', but are allocated to one side or the other. He uses inside which I am less keen on as it causes confusion when one or both groups have an eye. I have proposed border but that has raised other objections. For the moment I leave this table using RH's inside terminology.
The table itself is on a separate page to ease printing.
The EBGJ is slowly being placed on the BGA web site. There are EBGJ Writing Conventions.
Home >
Steve >
Go >
Tools >
Seki Table
British Go Association
Last updated 2004-08-10
This page is part of http://www.stocton.org/
Email: webmaster@stocton.org