Some remarks about the INGO rating system
-----------------------------------------
0. History: The INGO-Rating is the eldest chess rating system and
as I think the best one, too. It has been developped
by the German Anton Hößlinger in the fourties and became
modified for several times in the following years. It was
in use by the German Chess Federation until 1993. Still
today there is a quite similar rating system in England:
the British Grading System.
To keep the system comprehensible some simplifications
have been made for the special Xiangqi variant.
This Xiangqi version startet in 1990, first tournament to
be rated was the European Championships held in Germany.
Having his origin in western chess, points are given
in "chess" style: win is one point, draw half point
(in Chinese a win is 2 points, draw is one point).
1. Priciples: a) The "better" player the lower rating.
INGO range is from about 10 (world champion) up
to 250 (beginner), maximum rating should be 300.
b) Rating performance has to be calculated strictly
in chronological order of the tournaments.
c) The ratings will be updated after each tournament.
d) Each INGO rating gets an index that means the number of
tournaments that have been calculated for this player.
Example: a rating of 120-10 means an INGO of 120,
-------- resulting out of 10 rated tournaments.
2. Meanings: One point difference means one percent more or
less of expectancy range.
Example: If A has a rating of 110 and the average of A's
-------- opponents is also 110, the expectancy range is
exactly 50%; if the opponents' average is 100, expectancy
range is 10% less than 50%, means 40% (when playing 10
games A should gain 4 points otherwise his INGO rating
has to be changed).
3. Limits: Due to the linearity of the INGO system, the calculating
of expectancy ranges is limited to 50 points difference.
(Of course it is not possible that a player may gain
more than one point resp. less than 0 points per game..)
So we have to cut off in case of opponents' rating
differs more than 50 points from yours, when calculating
your opponents' average.
Example: A's INGO is 100, his opponents' 100,170,40,
-------- for calculation of average take 100,150,50,
so the average will be 100 and A should gain 50% that
means 1.5 points of the three games (three points in
"chinese" counting style).
4. Performance: Row "H" shows the rating performance of a player
for the calculated tournament.
5. New INGO: Abbreviation of INGO is "F" (F-old, F-new).
F is normally calculated as follows:
F-new = (F-old * k + H) / (k + 1), where
k means the expeansion coefficience:
6. Coefficience: The lower coefficience, the faster the changes!
Normally the expension coefficience is
k = 3; if your index is less than 10, make:
k = k - 1; if your index is less than 5, make:
k = k - 1; for every game less than 5 in the
current tournament, make:
k = k + 1; if the tournament's time schedule is
less than 1 hour per player (quickplay),
k will be doubled:
k = k + k (but herefore see 7.a Restrictions)
Example: your rating is 120-4 and you play 3 games
-------- in your 5th tournament, your coefficience is:
k = 3 (starting value)
- 1 (because less than 10 tournaments)
- 1 (because less than 5 tournaments)
+ 2 (2 games less than 5)
= 3!
If this tournament is a quickplay tournament,
k will now be doubled to 6!
New rule since 1997: only for overseas competitions of
-------------------- main interest, if time schedule is
more than one hour per player and game (World Cups f.e.):
in those tournaments k is decreased by 1 point, k = k - 1;
In any case minimum k has to be 1 - of course!
So the fastest rate can be 1:1, the normal one is 3:1.
7. Restrictions: games not to be rated are
a) fastplay tournaments (time schedule less than
30 minutes per player and game)
b) computer games
c) simultaneous games
d) forfeit games
e) correspondence games and similar competitions,
such as: internet, e-mail, phone, telex ..
f) tournaments which are sent in very late can only be
calculated if they are not elder than one year!
8. Specials: There are some special rules for calculating
of beginners or those players whose strength
increases quickly in a short time.
9. Chinese ELO: I tried to find out the best conversion to the
Chinese ELO rating (resp. vice-versa), but as we still
do not have more comparative items I propose
ELO = 2700 - INGO * 7 resp.
INGO = (2700 - ELO) / 7
10. Use of ELO: The use of existing chinese ELO-ratings is allowed
for players without INGO, but only if the normal
results seem to become irregular (if there are too
many players without INGO rating in the tournament..)
11. Other ELOs: Other Xiangqi Associations have different ELO rating
systems, f.e. Japan, Hongkong and Singapore. The proposed
formula as given above can not be used for these
national systems, the results can become quite irregular.
As far as I know in Taiwan a DAN system is used.
The British Grading system (from western chess, introduced
to Xiangqi by Patrick Donovan) is nowadays out of use.
Siegfried Huber
Ottenbohlstr. 12
D-88690 Muehlhofen, Germany
German Xiangqi Association
e-mail: Huber.Siegfried@t-online.de
quickplay / fulltime definition, if base time < minimum time:
a game of 30 move MUST guarantee the base time!
f.e. 45min + 30sec per move > 45 + 30 * 30 / 60 = 60min: full time!
f.e. 45min + 20sec per move > 45 + 20 * 30 = 45min: NOT full time!
|