Some "Beasts" which hurt the brain, but do not challenge Rulesets

The permanent location of this page is

These "beasts" are positions which are curiosities. Other positions -- which give different results/scores according to which ruleset are used are discussed separately. Here we will look at two types of position:

In this collection, I have made little attempt to provide an explanation for each Diagram -- please check with the original article.

References to the publication "Go" are to the journal "Go" of the Nederlandse Go Bond, which published a sporadic Series entitled "Goochelrubriek" {Jaap K Blom tells me that Ger agreed that this is to be interpreted, in English, as "Weichicraft" -- 'Goochelen' is wizardry and a 'goochelaar' an illusionist}. As far as I know, all of these were written by Ger Hungerink. I am especially grateful to Theo van Ees, who took the trouble to find copies of the original journals, and send them to me.

Hans Mulder ( kindly sent me many useful comments on 18 October 1996. I have used, and slightly changed, his comments -- I mark these with "[HM]" -- all faults are my own.

Item 1: Go 17/4 p10 "Ger Hungerink Interview" by Annabelle Bousquet

[HF : Apparently, this diagram may refer to the Goochelrubriek in Go 16/2, where it was first published as the winning entry to a "problem" posed in Go 16/1 -- it is a Black group with 72 "false" eyes, with no Black stones in atari.[HM] ]

Item 2 : Go 25/2 Ger Hungerink

A problem first posed in Go 17/5 in which the first and last row should consist entirely of alternate White and Black stones, with as many rows in between being the same! [HM]

Item 3 : Go 14/1 Ger Hungerink : 28 Two-eyed groups

Possibly a first attempt to find the maximum number of two-eyed groups -- see below -- 31 two-eyed groups.

Item 4 : Go 14/2 Ger Hungerink -- 31 Two-eyed groups

Maximum number of two-eyed groups? See below.

Item 5 : Go 14/3 Ger Hungerink -- 31 two-eyed groups

see above

Item 6 : Go 14/3 Ger Hungerink -- Multiple Seki

A search for the largest number of groups in Seki -- see below.

Item 7 : Go 14/4 Ger Hungerink -- Multiple Seki

129 groups. The max number of groups with 2 eyes is 31 -- see above; max number of groups without any eyes is 126. The max number of groups with one eye is 42 -- see below

Item 8 : Go 14/4 Ger Hungerink -- This can be substituted for the Top Right Corner in the above.

Item 9 : Go 14/5 Ger Hungerink -- Multiple Seki -- one eyed groups

Item 10 : Go 14/5 Ger Hungerink --

129 groups. The max number of groups with 2 eyes is 31; max number of groups without any eyes is 126. The max number of groups with one eye is 42 -- see below

Item 11 : Go 14/5 Ger Hungerink -- These can be substituted for the Top Left and Bottom Right Corners in the above.

Item 12 : A huge capture which still does not give life

On 24 October 1996, I was sent the following diagram by Clive Hunt from Johannesburg ( who told me that he was given it by Victor Chow, but does not know where Victor got it from.

I said "The diagram looks vaguely familiar to me -- does anyone recognise it from one of the Classics?", and then, on 12 Jan 1999, I got the following information from Dimas Cabré i Chacón (

I have made a little research and I discovered that this position is a simplification of a Go problem in the classical book Gokyo Shumyo [HF Note: Japanese original: in 1811 by Hayashi Gembi -- check out the section on Classical Go problems on my main Go page]. In fact, in the original problem White (colors are reversed in the original problem) manages to capture 72 stones (not outright) and still is not able to make two eyes. I'm sending you an SGF file of that position with its initial setup and with the main variation of the "solution".

On 02/08/2000, I got a message from Jean-Pierre Vesinet , saying:
In your excellent web page you attribute the following 16-point nakade problem to the classical book Gokyo Shumyo (1811).

This beautiful problem is a century older: it is exactly the same (symmetry excepted) as the problem #74 of Igo-Hatsuyo-ron, written by Inoue Dosetsu Inseki in 1713.

I checked in the Chinese version Fayanglun of the Hatsuyo-ron. It can be found on the Internet (without solutions, in GB codes) at

{HF note: to view this wonderful page, and many others in Chinese, Japanese, or Korean, you may like to use NJStar Communicator from NJStar, where you can find other useful resources, such as Wordprocessors (with built-in dictionaries!)}

It is Black to play. Black plays at B17, capturing the White stone. White plays atari at B16, and will now capture 16 stones (rather than 15), but will still die. Here is the SGF file.

Does anyone know of a capture of (a single group of) more than 16 stones, which is still not enough to give life to the group that makes the capture?

Item 13 : Go 15/1 & 17/3 Ger Hungerink -- another Seki

Item 14 : Go 15/4 Ger Hungerink -- White cannot make a live Group, no matter how many free moves

Item 15 : Go 17/3 & 17/4 Ger Hungerink -- White 1 is the worst possible move!

This is "White To Play -- find the worst move"! The group in the lower left is dead if White does not play at 1. It seems as if Black dies if White plays anywhere else but 1! [HM]

Item 16 : Go 17/3 & 17/4 Ger Hungerink -- Position with White to play, after move 290 in the above

There is an error in the diagram below (thanks to R.W. Crowl for finding this) -- in the bottom right corner 4-5 from corner -- there is a black stone where there should be a white one -- the black and white spirals should be unbroken! I will fix this in the next edition.

[HF : The White stones live, with sente, but Black lives! White wins if receiving komi!]

What is more interesting is that if White had not played the single White stone -- White 1, in the above Diagram -- then White could kill all the Black stones thus: White plays one point below White 1 (putting the whole Black group in atari -- if Black captures then White can play atari again, and all the Black stones die! White 1 really was a bad move.[HM]

Item 17 : Go 15/1 & 15/2 Ger Hungerink

Black Plays at 1 -- what is the result?

White takes the ko, but the ladder does not work for Black. Black's ko threat is to play one exchange in the ladder and then re-take the ko. White is now in a ladder, and does the same. Black's ladder is 4 steps longer than White's, but Black has to play 129 before White plays 132, otherwise White wins by playing 128 at 129. That is why Black does not take the ko at moves 119, 121, and 123. [HM]

Many Thanks to Fred Hansen of the Andrew Consortium for the gifs which he created from files consisting of X's and O's, using their tools. Thanks also to Fred, Hans Mulder, and Nick Wedd, for pointing out several errors, and generally improving the quality of the annotations.

last Updated 2007/07/20

This page maintained by (sorry please remove ".NOSPAM" from my address), whose WWW page is Harry Fearnley's Home Page