Ingo number

The Ingo number (also called INGO or INGO number ) is a former rating number in chess with which the playing strengths of individual players could be assessed and compared.

General

The inventor of the Ingo numbers is Anton Hößlinger . He developed the process and named it after his place of residence, Ingo lstadt . From 1947 this system was used in West Germany until it was replaced by the German rating number in 1991/92 . In Chinese chess ( Xiangqi ), the Ingo rating is still used in Germany.

Manfred Hollack (Hessen), Hermann Markgraf (Hamburg), Hans Rammin (Berlin), Heinz Wilms and Karl-Heinz Glenz (North Rhine-Westphalia), Eduard von Wolff (Lower Saxony), Reinhard Cherubim (English ) helped introduce the Ingo system Tournaments) and Georg Müller (Rhineland / Palatinate).

The Ingo numbers of the individual players were published in "Ingo-Spiegel". This was published once a year from 1957 to 1964, 1966, 1967 and from 1975 to 1991. Hermann Markgraf was head of the so-called “Ingo-Zentrale” from 1960 until (until his death) March 16, 1974. His successor was Karl-Heinz Glenz (until 1994). In 1974 the DSB Congress created the “Ingo-Elo-Zentrale”. The head was Karl-Heinz Glenz, deputy Manfred Hollack.

A player with a low Ingo number is better than a player with a high Ingo number. Most club players have a skill level between 100 and 190. The Ingo scale does not end at 0, a player with 0 Ingo points has a corresponding skill level of 2840 Elo points and can theoretically of course increase even further. With a difference of 25 points, the average number of points to be expected according to the Ingo system is 75% (compared to 76% for the estimate using the Elo system, conversion see below). With 50 Ingo points difference, the point expectation according to Ingo is 100% (according to Elo approx. 91%). The reason for this difference is the linear calculation of profit expectations in the Ingo system. This was necessary to enable an effective manual evaluation.

calculation

The Ingo performance (tournament success number, so-called "half number", therefore H) was calculated from the average Ingo number of the opponents and the average points scored against these opponents using the formula: ${\ displaystyle H}$ ${\ displaystyle {\ bar {G}}}$ ${\ displaystyle {\ bar {S}}}$

{\ displaystyle H = {\ overline {G}} - \ left (100 \ cdot {\ overline {S}} - 50 \ right) = {\ frac {1} {n}} \ sum _ {i} ^ { n} I_ {i} - \ left (100 \ cdot {\ frac {1} {n}} \ sum _ {i} ^ {n} s_ {i} -50 \ right)}

The new ingo rating was then calculated as follows: ${\ displaystyle I _ {\ mathrm {new}}}$

{\ displaystyle I _ {\ mathrm {new}} = {\ frac {H \ cdot n + I _ {\ mathrm {old}} \ cdot E} {n + E}}}

I _old : alter Ingo value

H : Ingo performance (tournament success number)

n : number of games played

E : Development coefficient (this was based on the age of the player to be evaluated: if this player was younger than twenty years it was 10, if it was between 20 and 25 it was 15, and from the age of 25 it was 20)

The idea of the development coefficient, which was based on the age of the player to be evaluated, was based on experience that younger players fluctuate more strongly in their performance, so that their previous Ingo number should be weighted less. The term development coefficient came up later with the introduction of the Elo and DMZ system. The common term was "factor".

One advantage of the Ingo value number system is that the point expectation P can be estimated very easily from one's own evaluation number W and the evaluation number of the opponent G.

{\ displaystyle P = 0 {,} 5 + {\ frac {GW} {100}}}

Example: A player with an Ingo number of 130 plays against another with a number of 160. The result is an expectation of 0.5 + 1.6 - 1.3 = 0.8 = 80%. This means that with 100 games on average, the better player can be expected to score 80 points.

Each one more Ingo point distance thus corresponds to one percentage point in the result estimate. From this formula the above formula for the tournament performance can be derived (it is a maximum likelihood estimator ).

Conversions

Elo

With the introduction of the Elo numbers, Reinhard Cherubim, Manfred Hollack and Arpad Elo determined a linear conversion formula with which the Elo number can be determined from the Ingo number : ${\ displaystyle I}$ ${\ displaystyle R}$

{\ displaystyle R = 2840-8 \ cdot I}

With the following formula, an Ingo number can be calculated from an Elo number : ${\ displaystyle R}$ ${\ displaystyle I}$

{\ displaystyle I = 355 - {\ frac {R} {8}}}

DWZ

When converting the Ingo numbers into the DWZ , which uses the same scale as the Elo numbers, it turned out that (minor) local differences in the meaning had emerged. Since almost all games only take place between players from one region, the scales have slipped slightly over the decades. For this reason, the specified Elo formula was not used for the conversion, but different normalization factors were used from state to state. However, the differences were not excessive, so that the given formula is always a good guide.

literature

Karl-Heinz Glenz: Ingo system . In: Manfred van Fondern (Ed.): Lexicon for Chess Friends , Verlag CJ Bucher, Lucerne / Frankfurt am Main 1980, ISBN 3-7658-0308-1 , pages 140-141.

Web links

Everything about Ingo numbers at the German Chess Federation - with Ingo brochures by Anton Hößlinger and an archive of Ingo Spiegel