Problem of the day…

The problem of the day is trying to keep chess databases neat and tidy. A difficult task, thanklessly done by several individuals who work selflessly, tirelessly, and separately. Two very useful databases, available to all, are ICOfY and Mill Base. Today I was applying some recent updates to ICOfY and decided to delete twins inside SCID when I came across the following interesting position (interesting to me at least!):

(White to move – 46.??)

Contreras, Henry Esterban vs. Molina, Mojica Angel Valent 2013Now, the position is interesting because White has to find a critical move (there are actually two) which allows White to draw. Can you find it?

But there’s another reason the position is interesting – because FIDE’s game report for this tournament gives the game as follows:

[Event "VII FESTIVAL INTERNACIONAL DE AJEDREZ UNIVERSIDAD "] 
[Site "Aula Multiple Universidad Central sede centro"] 
[Date "2013.10.11"] 
[Round "4"] 
[Board "71"] 
[White "Contreras, Henry Esteban"] 
[Black "Molina, Angel"] 
[Result "0-1"] 
[ECO "A00"] 
[WhiteElo "1800"] 
[BlackElo "0"] 
[PlyCount "23"] 
[EventDate "2013.10.10"] 
[EventRounds "9"] 
[EventCountry "COL"] 
1. e4 c5 2. Nf3 d6 3. d4 cxd4 4. Nxd4 Nf6 5. Nc3 g6 
6. Be3 Bg7 7. f3 a6 8. Qd2 Nc6 9. O-O-O Nxd4 10. Bxd4 O-O 
11. g4 b5 12. h4 0-1

(Run an engine here, and you can reassure yourself that the game score must be truncated – there is no reason for White to resign at this point)

SCID’s twin detection deleted this version of the game for another containing the above position, ending with White resigning after 47 moves. Moreover, determining the exact identity of the Black player involves a measure of detective work – since the FIDE game score doesn’t give the FIDE id’s in the PGN.

See http://ratings.fide.com/view_pgn.phtml?code=87191 , for the recent 2013 tournament “VII FESTIVAL INTERNACIONAL DE AJEDREZ UNIVERSIDAD” held in Columbia.

Now, the fact that the tournament was held in Columbia strongly suggests that “Manas Molina, Angel” (ESP), FIDE id =  2254832 was NOT the player involved. A google search gives another possibility – that of “Molina Mojica, Angel Valentin” (COL), FIDE id = 4444426.

It’s understandable that different cultures might get confused with how to properly deal with a name from a foreign land – which name goes first, the paternal stacking of names, and other customs which might be unfamiliar. Even within one culture the names might become confused – whether Bobby Fischer needs a middle name, or if a Jr. should be added, etc. SCID, and other database software, all allow redundancy and remapping. But software fixes ultimately fall short, which is why hand cataloging is inevitably necessary.

So a word of thanks for all the catalogers out there, it’s a difficult job, which often goes unthanked.

But given such difficulties, one must wonder why FIDE would publish games used for ratings without explicit inclusion of FIDE id’s. I think all FIDE games should be notated with with PGN tags that clearly identify the players involved, as well as the names.

One more thought before I sign off. Ingo, the maintainer of ICOfY, managed to find a corrected tournament record with the complete game, and corrected names. It would be nice if FIDE could offer a corrected version of the tournament as well (while keeping the original scores for archival reference). I wonder where Ingo got the corrections. Searching on the web with the extended moves for the game fails to locate another source.

Maybe it’s inside a zip file served out by TWIC. But then, where does Mark get it from? And why wouldn’t FIDE get a corrected version (maybe they do, but then why wouldn’t they publish it?). Perhaps tournament directors have a personal relationship with Mark after all the years that he’s been doing it. Of course, I’m a little worried about such a system then. There would be a single source of failure. For instance, TWIC almost ceased operations not too long ago.

Consider Pitt – aka Pittsburgh Chess Archive. I can’t get any files from their ftp site. These are old files, but I needed to get the ’92 Tal Memorial Rapid Play scores the other day. Pitt had ’em, and the net is full of millions of links to it as well – all pointing to the Pittsburgh ftp site. Which means a million stale links as far as I determined. A formerly reliable source of info becomes a single point of failure.

The battle with entropy is never-ending (well, till the harps come a calling).

Leave a comment