Xref: utzoo sci.lang:2133 sci.psychology:150 rec.games.go:400
Path: utzoo!mnetor!uunet!mcvax!nikhefh!t68
From: t68@nikhefh.hep.nl (Jos Vermaseren)
Newsgroups: sci.lang,sci.psychology,rec.games.go
Subject: Re: kyu?
Message-ID: <445@nikhefh.hep.nl>
Date: 31 Mar 88 10:02:59 GMT
References: <1257@PT.CS.CMU.EDU> <1046@cod.NOSC.MIL> <1021@bingvaxu.cc.binghamton.edu>
Reply-To: t68@nikhefh.hep.nl (Jos Vermaseren)
Organization: Nikhef-H, Amsterdam (the Netherlands).
Lines: 84

The ratings of go know several systems. The most used system in the West is 
the Japanese amateur ranking which goes in dans and kyus. The higher ones
dan ranking the stronger one is. Kyu's are negative dans but there is no
zero in this scale. So 1 kyu comes one step under 1 dan.
In the amateur ranking the difference in strength between two players
can be given in terms of the number of handicap stones that are needed to
play an evenly matched game. This number is also the number of dan or kyu
grades that they differ in 'official' strength.
In practice there can be large fluctuations as people have 'angst gegners',
off days etc. Also it turns out that the handicap is a little meager and
white wins most handicap games. This can be attributed to several sources:
1:	Players tend to overrate themselves. In many countries the dan grades
	are awarded by a classification committee. For the kyu's such a thing
	may not exist. People like to be 'almost 1-st dan' or 1 kyu. Etc.
2:	The system may not be linear.
3:	The following was recognized in (I believe) Germany many years ago:
	1 stone handicap gives black on the average only half a stone more
	on the board. Two stones handicap gives him actually only 1.5 stones
	more etc. So maybe black gets not enough handicap.
The following solution for point three was made:
To be accurate we need half dan grades. This gives the
'class' system in which the zero was taken to correspond to the
strongest player. It is not known whether this is very accurate. Anyway,
1-st dan corresponds to the classes 19 and 18. 19 is called weak 1 dan and
18 is called strong 1 dan. 17 and 16 is two dan, etc.
The handicap is now determined via h = (class(a)-class(b))/2 Fractions are
rounded up but the compensation is then 5.5 points komi. This means that
it is always black who gives komi. The handicap between a two dan and a one dan
can vary this way from 1 stone ( 17 vs 18 ) to a full two stones ( 16 vs 19 ).
In practise this works very well, although some white players may complain
that now they have to work so hard.
In the past years this system has lost quite some ground. The Germans have
abandoned it, because getting the dan grades right is already no simple task.
Only the Dutch are still using it but then only for internal tournaments.
Still it has many advantages. Often one knows that a player is getting stronger
but the promotion of a full dan doesn't seem justified. It is still possible
to give him official recognition with a promotion to 'strong .. dan'.
It also makes the handicap games much more interesting.

Lately (in the past few years) the Hungarians have tried to set up an ELO
rating for go in which people coeld earn or loose points according to
results in tournaments. To do this properly is a formidable task that can
only be expected of somebody if he gets paid for it. In that case it is 
possible to complain is things start lagging behind. The go scene in Europe
doesn't have these financial means, so as could be expected the results are
very irregular. This system suffers also under the visits of people who's
strength is not well known, or who visit on their vacation one or two
tournaments and then go back to their own continent, never to return.

Nowadays the dan grades in Europ start to become a little unified.
For years there could be up to two stones difference between players with
the same official strength. As people meet each other more frequently in
international tournaments things get corrected, sometimes people get demoted,
and nowadays Germany, the Netherlands, France and England have their
classification basically in agreement. Also the Poles are getting tuned to
this level. Till two years ago the Russians and the East Germans were still
off by one or two dan grades but also this is improving. It seems that their
early classification was based on comments of very polite japanese visitors.

In Japan the amateur rankings are an anarchaic mess. There is no system in
it and each club has its own dan grades. If one goes there as a Dutch 4 dan
it can happen that one day one has a tough time agains 3 and four dans, while
the next day in another club one gains even points while playing as a six dan.
This must make the organization of a good tournament very difficult.

The dan grades in the US are very often one or two dan grades off compared
to Europe. Occasional visitors to European tournaments have found that it
very difficult to win any games in the early rounds if they use their US
strength. There is no problem with this, as each country may set up its own
system and the main thing that counts is its internal consistency. Only when
individuals join in oversea competition one may need a 'personal reevaluation'
on a case by case basis.

The stronger amateur players in the West are about six dan. A player who would
be a seven dan player would probably qualify to become a professional (European
scale wrt the seven dan). The professionals have also dan grades but their
differences in strength are probably about 1/4 of the differences for the
amateurs. Maybe the 1/4 is nowadays even an overestimate because a 9 dan prof
will have a very hard time giving a 1 dan prof 3 stones.

I hope this gives some information about the go classifications.

Jos Vermaseren
T68@nikhefh.hep.nl (or T68@nikhefh.uucp)