SuperGM pgn collection

[Norm Pollock]

I have added another pgn collection. This one involves "Super GrandMasters" against each other. Keep in mind that in many of the games, one or both of the players are not at the peak of their playing careers. There are 21,000+ games involving 115 players. Full details are in the included readme file.

direct link:

http://www.mediafire.com/?3yhc5mzyniz

link through my download page:

http://www.hoflink.com/~npollock/chess.html

-Norm

[Thomas McBurney]

G'day Norm,

You probably want to fix the following game (Game No. 13356)...

Code: Select all

[Event "Novosibirsk"]
[Site "Novosibirsk RUS"]
[Date "1993.??.??"]
[Round "?"]
[White "Beliavsky, Alexander G"]
[Black "Akopian, Vladimir E"]
[Result "1-0"]
[WhiteElo "2655"]
[BlackElo "2655"]
[ECO "D18"]
[EventDate "1993.??.??"]

1.d4 d5 2.c4 c6 3.Nf3 Nf6 4.Nc3 dxc4 5.a4 Bf5 6.e3 e6 7.Bxc4 Bb4 8.O-O
Nbd7 9.Qc2 Bg6 10.e4 Bxc3 11.bxc3 Nxe4 12.Ba3 Qc7 13.Rfe1 O-O-O 14.Qb2
Rhe8 15.a5 e5 16.Rab1 c5 17.Bf1 f6 18.Ra1 Nd6 19.Nd2 Kb8 20.Rec1 Ka8 21.
Nb3 c4 22.Bxd6 Qxd6 23.Bxc4 a6 24.Bf1 Rc8 25.Nd2 Rc6 26.g3 Bf7 27.Rcb1 Rb8
28.Ra4 exd4 29.Rxd4 Qc7 30.Qb4 Ne5 31.f4 Ng4 32.Re1 Rc5 33.Bg2 Nh6 34.Qb6
Qxb6 35.axb6 Rb5 36.Rd6 Ng8 37.Ra1 Rf8 38.c4 Rb2 39.c5 Ne7 40.Rxa6+ Kb8
41.Ra7 Nc6 42.Rd7 1-0


The move '9. Qc2' is a queen suicide. It appears to be a simple data entry error and the move probably should be '9. Qe2'. You don't want people using this game to build their opening book.

Cheers,
Tom.

[Thomas McBurney]

I just found that game No. 13387 is the correct game. So you actually want to delete that game from your PGN file.

Cheers,
Tom.

[Norm Pollock]

I just found that game No. 13387 is the correct game. So you actually want to delete that game from your PGN file.

Cheers,
Tom.

That's a great find. How did you do it?

I traced the source of the error to another collection. These errors are usually widespread and all I can do is correct them in my pgns.

Speaking of errors in pgn files, the most amusing one I ever found was the one where Johann Fischer's name had evolved into Joan Fischer probably because someone assumed that he was Bobby's sister. Fortunately there is http://www.365chess.com/ where these things can be sorted out.

[Thomas McBurney]

That's a great find. How did you do it?

I traced the source of the error to another collection. These errors are usually widespread and all I can do is correct them in my pgns.

G'day Norm,

This game was actually a recent topic here -> http://www.open-aurec.com/wbforum/viewtopic.php?f=2&t=50195

Gavoita played the queen sacrifice move from opening book which was traced to a PGN file used to create the book. I looked at my database which I use to create Kanguruhs opening book and also found that move there. In my case it was traced to a PGN called elo2600.pgn I downloaded from someones website. Fortunately my opening book tool can tell me who played that move, which I was then able to find in the PGN file.

I suspected that this was a wide spread issue, so I searched your PGN file and found it, confirming my suspicions.

I think there is probably more than one game with data entry errors, so I think it would be good if someone developed a tool to find these games.

Cheers,
Tom.

[F. Bluemers]

Good idea Tom
We won't stop you

Problem with those tools is the time they take checking many games containing more moves.
I have played a bit with the 'blunder checking' from Ed's protools.
I guess one can do the same with the new openingsbook manager of Glass.
Both tools would need you to build or convert the pgn file(s).

Best
Fonzy

[Thomas McBurney]

Good idea Tom
We won't stop you

Haha, thats funny because I was thinking that this is a job for me.

I could probably incorporate it in my Game Analyser utility. If I have a crack at it, I will only look for obvious errors such as this queen sacrifice. It will need to be fast, so deep searches for each position analysed would be out of the question.

Cheers,
Tom.

[Norm Pollock]

I recommend that a minimum of 2 occurrences be used for any opening move based on human pgn games. This will avoid the faux move from becoming part of the book, although statistics of that game will be used for prior moves.

The most outrageous type of recording error is a false result after a decisive conclusion, ie checkmate or stalemate. My estimate is that this occurs about 1 in 10,000 games in your average pgn file. I use "joined" by Andreas Stable to detect this type of error, and then I manually correct the game.