Chess Mail logo   J. H. Blackburne   Howard Staunton   Paul Morphy  
The Chess History
Research Zone
Dr Tim Harding  
Editor: Dr Tim
Harding
  Tim's correspondence chess history

More problems with Steinitz games

This week's article deals with some games played by Steinitz from 1888 onwards against opponents other than Lasker. In these our research, and Thomas Niessen's Chess Suite, have detected discrepancies between what was actually played and what has appeared in certain books and databases.

There is no PDF file this time. The full set of 25 corrected games (which includes some minor games from 1882 onwards) may be found in our annotated PGN file. The more serious cases are discussed in this HTML page. Please note that, with the exception of two examples from Cologne 1898, these are not in my book Steinitz in London, which details numerous cases of mistakes in databases dealing with Steinitz's career in Europe up to 1883.

We begin with Steinitz's first match in defence of the World Championship, against Chigorin in Havana (1889). We found discrepancies in three games but only one is serious and affects ChessBase's record of the match. The following position arose in Game 17 after Chigorin's move 26 f3.

Here Steinitz played 26...Qc2, as shown in the April 1889 number of his International Chess Magazine. Pickard had the correct score but many ChessBase databases have a wrong move, 26...Qa2, returning to the correct position after 27 Ne3 Qb3.

We found no fewer than seven problem games in Steinitz's second title defence match, against Gunsberg in New York (December 1890 to January 1891). ChessBase have really butchered the record of this contest.

They have the wrong final move in the first and fourth games and slightly truncate the second game. They have an incorrect 55th move for White in the 9th game (restoring the correct position at move 58) and in the 11th game they add an extra move 29 a6 which was mentioned in Steinitz's notes but not played. In Game 19 they have an incorrect final move for Black.

Here is the worst case. In Game 13 they go wrong at Black's 35th move by wrongly transcribing descriptive notation, as in one of the Lasker cases in our first article.

Here Gunsberg played 35...Re7 but ChessBase have 35...Re2 and the rest of the game is therefore wrong.

In his standard work Schachmeister Steinitz, Ludwig Bachmann made several mistakes about the ex-champion's games against the Russian master Schiffers. Their game from Hastings 1895 (number 792 in his collection) was also mistakenly included as number 837, purporting to be one of their match games in 1896. Bachmann omitted two games of the match and the sequence of the games he did include is incorrect. In the third game of the match (see the PGN file) most databases have an incorrect move 56 for White.

One of Steinitz's games with Chigorin in the 1896 St. Petersburg tournament is wrong in every database I have seen, although the score was correct in Hoffer's column for The Field, Chigorin's column in Novoye Vremya, and in Bachmann's book. The problems begin after Steinitz's move 57...Ke7-d8 (see diagram). For the correct continuation see the PGN file.

There are also some relatively minor discrepancies shown in various games included in the PGN file. The most interesting case is the third game, against Michaelis. This is not in ChessBase while many sources, including Bachmann's book and chessgames.com, currently lack the finish which we found.

We end with one truly horrific case where ChessBase has no excuse for a really bad mistake. The correct score can be seen both in Richard Forster's biography of Amos Burn and Vlastimil Fiala's Cologne 1898 tournament book.

Amos Burn won the tournament and considered his win against Steinitz to be the best game of his career. At one point near the end Steinitz set a devilish trap, but Burn spotted and evaded it. With the incorrect move order given by ChessBase, Black would have sprung the trap.

Steinitz here played 55...h5 whereas ChessBase give 55...a5 first and ...h5 next move. But after 55...a5? White would win by 56 e5 which would be a blunder in reply to the h-pawn advance (56...fxe5 57 fxe5 Bxe5! 58 Rxe5 Rh6 mate!).

There is also a query in this game about which Black rook captured first on b6 at move 39, but with some primary sources unavailable during the pandemic, I have not been able to settle this question. See the PGN file for details.

Return to the series introduction page.