4.2 Who Scored against Whom?

The fundamental data for rating are the results of all games, which is contained in the cross-tables.

Many of the cross-tables sent in for rating contained questionable data. Though most contradictions are recognized by the rating program, fixing is an annoying task wasting much time. To mention only a few examples (some of them have been fixed meanwhile):

• One player simultaneously playing against two opponents:
  • London, 1999-08-26, round 2, `Hermann Kleier' and `Reyk Lettenmayer' against `David Spence'

  • N˙mwegen, 2000-03-18, round 4, `George Franka' and `Sander Rinzema' against `Wim Snelleman'

  • Wageningen, 2000-05-07, round 2, 1st, `David Bouten' and `Rutger Rietveld' against `Mark van der Linden'

  • Wageningen, 2000-05-07, round 4, ere, `Jeroen Otto' and `Sander Rinzema' against `Hans Geertsma'

  • Kiev, 2000-12-03, round 5, `Vladislav Krasnij' and `Masha Klimenko' against `Anna V. Klimenko'

• Contradiction between the standing printout during tournament and the Bombbrief.
  • London, 1998-08-29, round 8, `Abel Smits' against `Johan van de Ven'

• A player and his opponent won/lost the same game:
  • Maastricht, 1999-08-14, round 1, `Frank Poppelaars' against `Wim Verle˙sdonk'

  • Maastricht, 1999-08-14, round 1, `Timmy Slagboom' against `Wilfried van Hoof'

  • Wageningen, 2000-05-07, round 6, ere, `Ruud Koopmans' against `Frank Poppelaars'

• Incomplete cross-tables:
  • Leuven 1999-10-16

  • Hoogeveen 1999-10-17

• SP files were broken
  • N˙mwegen, 2nd class, 2000-02-27

• SP files contained invalid data fields. Had to input data from html-files.
  • Maastricht, 2000-08-05

• Scoring of 8 points in a tournament, which was scored 6-1-0
  • Arnhem, 2000-12-23

• Player was late for his game. It was scored for the opponent. Though correct for the tournament ranking, this is plain wrong for rating.
  • London, 2000-08-25, `Hermann Kleier' against `Iwan Bralts'

• Some cross tables seems to be sorted by standings. But there are minor differences from the computed rankings. This may (or may not) point to an error.

Just to give you an idea, how nerve-wracking the data entry can be, I put forward the example, how I disentangled the tournament of N˙mwegen, 2nd class, 2000-02-27 (see above). The SP-files were broken and after typing in the data manually, I had to use the redundancy in order to fix the broken data:

The fax [and Bombbrief 18, page 15] stated that in the first round `Niek van Beuningen' won against `Jochem Gord˙n'. But `Jochem Gord˙n' played against `Anes Tesnjak' and `Anes Tesnjak' played against `Jochem Gord˙n'. On the other hand `Lars Janssen' and `Kees Lewiszong' played against `Joel Schiel', but `Joel Schiel' played against `Kees Lewiszong'. Therefore I paired `Niek van Beuningen' and `Lars Janssen' for the first round and let `Niek van Beuningen' win the match. Nice puzzle, wasn't it?

What worries me most is that all these results have been obtained with SP.