Scoring number (chess)

from Wikipedia, the free encyclopedia

Scoring numbers in chess are intended to evaluate the skill level of chess players.

Rating of chess players

In order to make the playing strength of chess players - mainly club players - comparable, the players are assigned ratings. These valuation numbers are either recalculated after each competition or valuation tournament or they are recalculated and published in regular periods - e.g. every quarter.

The Elo number is used at international level. This also serves as the basis for awarding international titles such as grandmasters . The World Chess Federation's Elo numbers have only been calculated and published since 1970. For chess tournaments before this time, however, rating figures were also calculated retrospectively on various occasions. At the national level, there are many country-specific systems for historical and financial reasons. In West Germany there was the Ingo rating since 1948 and the NWZ (National Rating Number) in the GDR . Both systems were replaced by the DWZ (German valuation number) in 1993 after reunification . The Swiss Chess Federation calculates a guide number for its members , which is published every two months in the so-called guide list . In England there is the so-called ECF Grading of the English Chess Federation. In the USA, the USCF rating is widespread among amateurs and is calculated similarly to the Glicko system.

More about the individual ratings:

A historical Elo number is an Elo number that was calculated for chess masters who were active before the introduction of modern ratings. It should help to compare the skill levels of previous and current chess masters. Such was calculated by Jeff Sonas for chess tournaments from 1840 to 2004.

Mathematical backgrounds

The systems differ in their complexity of the processes required for their calculation. Since integrals had to be solved for the calculation of expected values, older systems are based on their approximations with the help of linear functions (cf. Ingo system, Harkness system). Only the use of EDP systems made it possible to use non-linear approximations such as logistic functions (cf. Elo number, NWZ, DWZ). The NWZ introduced in 1972 in the DRR was already calculated on a mainframe computer. Modern systems, like the Glicko system, can no longer be calculated without a computer. This is also due to the large number of additional parameters. If in the first systems only one rating number was kept for each player, which indicated his current level of play and possibly also his age in the calculation, in the Glicko system it is also the time intervals between the games and other parameters that indicate, for example, how is a reliable rating number.

Example of a calculation

A simplifying example should clarify the principle of a rating number. It is believed that a higher rating means a higher skill level.

Assume that player A has a higher rating than player B. Player A is therefore stronger than player B. Now A and B play against each other and the game is to be evaluated. He can give three different game results.

  • A wins the game : This result was expected. A's rating number increases by a small amount, B's rating number is reduced by a small amount.
  • B wins the game : this result is a big surprise. B is rewarded for the unexpected performance with a large increase in his rating number. The rating of A is significantly reduced.
  • The game ends in a draw : This is also a (small) surprise. B's rating increases slightly, but not as much as if the game was won. The rating of A is reduced slightly.

The amount by which a rating changes depends largely on the difference in rating between players A and B. However, other factors can be added and depend heavily on the system used.

Relationship between the ratings

A FIDE Elo rating is usually greater than a DWZ, but this difference also depends heavily on the player's DWZ. From 2300 it is rather marginal, below that it gets bigger and bigger, in the range of a DWZ from 1900 to 2000 it was usually around 100 points before the lowering of the Elo, but becomes smaller due to the constant lowering of the Elo, because now also worse tournament results be included, but earlier only the good ones. Nevertheless, there are also major differences among individual players with DWZ and Elo, which were caused by the lowering of the limits.

A specific formula for converting Elo to DWZ and vice versa cannot be given. Both numbers are basically calculated according to the same principles. The differences come about because DWZ and Elo numbers are based on different databases. The Elo rating tends to only include higher quality tournaments (international) or games against stronger opponents.

As a rule, Elo-rated tournaments are only played after a certain performance level has been reached. In the DWZ, however, benefits from youth and childhood are also included. It is not foreseeable whether national rating systems will become superfluous as a result of the gradual lowering of the lower Elo limit.

Tournament rating number (TWZ)

The tournament rating number (TWZ) is the term used within a specific tournament - e.g. B. to create the starting ranking list (for the Swiss system ) - rating number used. This becomes necessary if not all participants have ratings from the same system. In Germany it is customary to use the higher valuation number when both numbers are available. National ratings from other countries or other associations (e.g. the professional chess association in the Berlin-Brandenburg company sports association ) can also be taken into account here.

Web links