 |
 |
Editor: Dr Tim Harding |
 |
 |
 |
 |
|
|
Editor: Dr Tim Harding |
|
|
|
|
© Chess Mail Ltd.
Last modified:
19 July 2010
When ChessBase first started producing correspondence chess database CDs (in 2000) we pointed out numerous errors on it and suggested that in future we should collaborate. They declined. Ever since we have been reviewing their CDs and pointing out errors in their treatment of historical games, but they persist in declining to take notice.
If you have already made the mistake of buying their CD, you may wish to make the corrections indicated below.
Tim Harding indicates some of the errors he has found:
There are 670,471 games on ChessBase's 2008 CD (compared with 588148 in 2006 and over 9000,000 on our CD). 74 of these 'games' are actually Openings Surveys.
Three games from the spurious “Munchhausen Memorial 1995” (just look at the name!) are included, as well as 125 games of the probably spurious “Pantos Mem 1999” and 66 of the equally dubious “Maturin Gambit-ch” (supposedly, but not, ICCF) 1999. The players in these are unknown to ICCF and don’t appear in any other events. I believe ChessBase, or another database compiler, has invented these events as a joke or to catch out copiers but I have been unable to identify who was responsible.
I mostly concentrated on looking at the pre-First World War games of which I have made a special study and in most cases have seen printed primary sources. UltraCorr2 does not include the source information and notes because I do not wish others to plagiarise my research before I am ready to produce a book on the subject, but eventually all these games will be published with full sources and annotations.
Some of the pre-1918 games on the ChessBase CD are NOT in fact corr games and others are modern games incorrectly dated, as in most cases I have pointed out before at least twice as the same mistakes continue to be repeated by ChessBase.
For example, it is well known that the first game with the Evans Gambit was played between Captain Evans and Alexander McDonnell in London, so why do they persist in including that game on their Corr-Cd instead of MegaBase where it belongs?
Wyvill-Lowenthal, St George’s is an obvious case where "CC" as an abbreviation for "chess club" has been misunderstood as meaning correspondence chess. Several St George’s Club and (London) West End club games etc. Also Blackmar-Hodges was an American OTB game.
Mnay of these cases involve Zukertort; I am sceptical that any of the Zukertort games on the CD are actually CC games though one or two might be.
For example “P. Bloch-Zukertort, 1862” is actually a Breslau-Berlin inter-club game 1863 where Zukertort was possibly involved, but maybe on the Breslau side?
Janowski-Leonhardt was played at Barmen 1905, a major international tournament OTB. Janowski did play some CC in the 1890s but not in the 1900s.
The well-known Clemens [or Clemenz] v Eisenschmidt game in the Evans Gambit was not postal; it was played in Dorpat, 1862.
Blackburne-Blanchard, 1891 is probably OTB also.
There are three games/fragments Tereschenko v “Linier” or “Lizmnieres”: all are variants on the same game against Baron J. de Lignieres. This indicates carelessness in editing by them and no real attempt to search for "fuzzy" duplicates.
Even with a player as important as Alexander Alekhine, they make no effort. Alekhine-Duhm, 1909 was an OTB game; STILL WRONG Black’s name may be wrongly spelled.
There are other mistakes over Alekhine’s games. The database includes two fragments, Alekhine-“Jukowski”, 1909, truncated at move 20; and a finish Alekhine-“Sukowski”, 1908. These are both actually from the same game versus V. Zhukovsky, from the 16 th Shakhmatny Obozrenie corr tourney of 1905-6 (which is also in the database) . I sorted this all out in 2003: see my book 50 Golden Chess Games. The game is annotated on the UltraCorr2CD.
In the 1970 English edition of Alekhine's Best Games, Black's name is given as "W. de Jonkovski". The date of the game is wrongly given there as 1908 (possibly A's fault). I have also seen Black's name given incorrectly as "Schachowski", but he was a different person (Shakhovskoi) who also played Alekhine at CC.
Another World Champion's CC career is also misrepresented. Steinitz v Liverpool Chess Club (0-1,23) was actually won by Steinitz with Black and 26 moves long.
Two games from the first World CC Championship Final have incorrect finishes on the ChessBase CD: Mitchell-Barda goes wrong at Black’s move 28 which should be 28…Kg8 (and 1-0 in 33 moves not 32). A move pair is also omitted by ChessBase in Napolitano-Cuadrado, spoiling the Italian GM’s nice finish. These errors by ChessBase particularly distress me because of all the hard work I put in many years ago editing the book of the early World Championships. The games are correct on UltraCorr2.
The game Fejer-Chalupetzky has the colours wrong, year wrong and is truncated; a pity as this was a fascinating struggle. (In fact Chalupetzky won with White in 39 moves.)
Menke-Bode: CB has year 1811 for this game but it's obviously wrong; there was hardly any correspondence chess between individuals before 1840, and Menke was one of Keres's early postal opponents. Most likely this game is from the 1930s or immediate post-WW2. ChessBase corr CDs give the date of this game implausibly as 1819. We are still trying to trace the source which was probably later 19th or 20th century. The 4...Qh4 line wasn't played in 1819l ChessBase says "USA Queen" which could mean APCT Queen.
Wrong finishes from some early inter-city games are wrong in Pagni’s books too, but not sure who has copied from whom…
The game Wroclaw [Breslau in those days] v Hamburg, 1840 is truncated at move 37, when the great fire of Hamburg caused a cessation of play. It was completed a few years later (47 moves) as can be seen on my CD.
The finish of the game Archdall-Ellis, 1872 is wrong. The primary source, notes by Zukertort in City of London Chess Magazine vol2, july 1875, p182, clearly shows Black played 31…Kt to B sq (i.e. 31…Nf8) as on UltraCorr2 not 31…Nc7 as in ChessBase database. This is not an isolated instance?
The 10-move London-Athens game of 1897 (which Pagni has put in one of his books) is almost certainly spurious; maybe it was played between English and Greek individuals.
Amusingly in another fragment ChessBase has White as "Hell" instead of Hull chess club (v “Notingham” [sic]).
There are mistakes with more recent games too. They include the first 23 moves of my game against Joe Ryan played in an OTB league game in Dublin in 1990 – not postal! A game by the correspondence master Trevor Thomas (ENG) has him as ‘Thomas Trevor’.
There are some positives. Amelung-Nolcken, 1860, is one of several games new to me on the ChessBase 2008 CD, said to be from the Baltische Schachblätter. There were 48 games (mostly new to me) played up to 1914 from this source (in all 63 on the CD). I found no other pre-1914 games on the ChessBase CD which were both new and genuine, i.e. there were lots of wrong names, truncated games, OTB games etc. but no correct and genuine games that I did not already have.
Unfortunately no exact source information is provided in some of the other Baltische Schachblätter cases. As yet I have not seen the hard copy primary source. There does seem to have been rather more effort put into compiling this collection than the 2006 version which had very few games new to me, except some recent ones from Eastern Europe.
Most of the games new to me are from 2005-8 and appear to have been supplied to ChessBase by ICCF national federations.
Here are the most recently discovered mistakes which require correction if they are in your database from ChessBase.
The 50-move game Barcza v "Egon R. Madsen" from "olm2 corr4852" is actually Barcza v Bertil Wikstrom from CCPOL final-1 bd 2, 1949.
Gutop v “Kance, Edmunds” is actually Gutop v Konca, Damian from the 11 th CC Olympiad, 1987.
Shephard v “Bielczyk,Jacek” was actually Shephard v Bieluczyk,Bogdan
Dubois v “Snider,John” supposedly 10 th olympiad was actually Dubois v Snyder, Robert but truncated (the ChessBase version goes wrong at Black’s 38th move).
Two games "D'Engel v Desmarest" and "Humbert v D'Engel" from France, dated 1989, must have been late 19th century, probably 1889. These were players who regularly competed in French postal toiurneys, although D'Engel (Imre Engel) was actually Hungarian.
Here are some more mistakes that have persisted from previous ChessBase Corr-CDs without correction, despite there being poiinte dout in my previous reviews, in the Chess Mail printed magazine when I was running it and on this website.
Evans,W v Saint-Amour (who is A modern Canadian player); this is evidently an OTB game of Capt Evans v St. Amant! STILL WRONG
Mende-Demian, probably 1966 (they are modern Romanian players). STILL WRONG (ChessBase says 1866!)
Frank Parr-Wheatcroft, 1938 was OTB. STILL WRONG
So, almost certainly. was Fairhurst-Alexander, 1928. STILL WRONG
A Mrs Gilbert-Berry game from 1875 can hardly involved Canadian CC-GM F. John Berry who is still alive; it was William J. Berry of the US. STILL WRONG
“Couts-Dzedd”, 1879 is really Wm. Coates-Max Judd, 1877-9. STILL WRONG
This is not a complete list of mistakes.