1

I noticed that in the freely available TWIC databases (and perhaps in other databases as well) there are sometimes different notations for the same move. Just now I am looking at the Ruy Lopez where in some games white plays 6. Re1 while in others it is Rfe1 (which are of course the same move). Incidentally white scores much better with "Rfe1" ;-) Anyway having different notations for the same move makes reading the statistics (white win / draw/ black win) much harder.

That's why I was wondering how to get rid of such problems in the database. Also what other ways are there to clean up the database (e.g. removing duplicate games, ....)?

user1583209
  • 20,735
  • 3
  • 41
  • 97
  • Only Re1 is "correct" by most standards. I would expect that loading such a database into programs like ChessBase would automatically merge all Rfe1 games into the Re1 tree, and merge the scores as well. – TMM Sep 21 '17 at 09:42
  • 1
    I found even Kd3 Rxd3 sequence in one twic :-) – hoacin Dec 14 '17 at 08:33

1 Answers1

1

PGN files aren't databases, they're just lists of games. It's an exchange format, not a database format.

You can load them into a real database, like Scid or Chessbase. Those have their own internal representation of chess games that is separate from PGN. If the same move can be written in several ways, they'll all be interpreted the same way by the database software when it reads the PGN.

Database software is probably better for doing analysis than working with raw PGN files too.

Alternatively, if you really want PGN files, import them into a Scid database and then export them to PGN again, that should make them consistent.

RemcoGerlich
  • 27,974
  • 80
  • 122