Biggest Free Chess Database! (Over 6.9 Million Games)

Try FICS, the link I provided on the previous page.

Oh, there's also twic, the London Chess Centre. They give free updates every week, or at least they used to. I haven't checked in there for a while.

TWIC is still going on, it only goes to 2012 though and costs money to get the full package and has blitz/rapid games, but is still a good source. FICS is just internet games, I want rated classical games which people play over the board, just like in all other databases. it looks like there is really no pure classical game database out there and it would require some research/work to compile one.. meh.

Take another look at the FICS database. Filter for games played by players over 2600, and then filter for classical time controls.

Indeed the names are written GMxy so no wonder I couldn´t find anything lol. Games go back to 2008, but blitz/rapid games are marked as standard so no good either.

Whaddya expect for nothing, a rubber biscuit?

https://www.youtube.com/watch?v=jYyBZE0kBtE

I'm a bit surprised there isn't a crowd sourced project to create a high quality database for free. With good guidelines to normalize the data and headers, it seems like a project that would do well, and hopefully not have a single point of failure.

After downloading the database I checked for some strange games that had been pointed out to me earlier. The position after 1. e4 e5 2. Nf3 a6 3. Bb5 (some kind of a botched Ruy Lopez where white gambits the bishop) occurs on chess.com's Opening Explorer, and 10 times on AepliBase. Luckily for white, black never captures the bishop in these games. Suspicious!

18:35 ChessGameDB $ fgrep "1. e4 e5 2. Nf3 a6 3. Bb5" AepliBase.pgn
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 d6 5. Bxc6+ bxc6 6. d4 exd4 7. Nxd4
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 O-O
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 d6 5. Bxc6+ bxc6 6. d4 exd4 7. Nxd4
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 d6 5. Bxc6+ bxc6 6. d4 exd4 7. Nxd4
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.
1. e4 e5 2. Nf3 a6 3. Bb5 Nc6 4. Ba4 Nf6 5. O-O Be7 6. Re1 b5 7. Bb3 d6 8.

I should add, having this database is pretty cool and I plan to do some work on it. Just pointing out a problem in hopes of helping to improve it.

Yes, basically great work.

Clearly one of the sources for AepliBase (and chess.com Opening Explorer!) has some very dodgy games. I believe some arranged draws between pro players are as daft as this.

 Works very well.  Thank you very much.

Don't want to spoil the moment, but anything that is free lacks the motivation of update. Nice one time project which isn't ready to continue due to the fact that creator won't receive any benefits taking it further. Not speaking about many duplicates, index errors, computer games and many other fails professional chess databases look for. If you for instance compare it to the Opening Master Chess Databases which have been updating since 2004 regularly (and for commercial benefit) you will see the difference. Although I support free lunch, humans are not set to deliver anything for free more than 3 months. That is 90 days. All it matters is quality. And people here after trying so many things will end up with the quality product. It just saves time. The Opening Master is not perfect, it has flaws, but so far the best seen.

Greetings, Alexander H.

Founder of Opening Master

Hi Alexander, I wrote a mail to Opening Master support asking some questions about your 26 million game DB and I have still heard nothing back.

Hi Ed, apologies if the email gets missed, we reply 99% emails, sometimes it never comes or gets busted in between. Please write me PM or let me now the full name so we can search your email. Or just ask here. What were your questions? Happy to answer all.

Alexander

thanks for the free database access.

If I wanted to, I could just import the database into itself, and I'd have the new biggest free chess database - 13.8 million games! ;)

And that wouldn't be any more ridiculous than what the OP has done. His database is riddled with duplicate games...

Just for fun, I decided to run the AepliBase database through PGN Extract to separate out the duplicate games from the unique games. Bear in mind that PGN Extract determines duplicate games by looking ONLY at the moves of a game; It doesn't look at names, dates, or locations. Because of this, there are probably a small number of identical games that were flagged as duplicates, but that were actually different games played by different people at different times. However, I'd guess that these occurrences aren't too numerous.

So here are the numbers I got:

Number of unique games: 3,487,293

Number of duplicate games: 3,500,504

(There were 6,987,818 total games in AepliBase.pgn; 21 games somehow got dropped during the PGN Extract operations.)

So, factoring in the probable small number of false-positive duplicate games mentioned above, I'd guess that there are still probably nearly as many duplicate games in the database as there are unique games.

Thank you!

In a sense two duplicates of a game tell you nothing more than one game. They are rare as well, because it is exceptional for people to stop in exactly the same place.

?