09 July 2009

Unclear : Some Huebner Analysis

For the next example of Unclear Positions, I looked for a game from the 1980 Korchnoi - Huebner Candidates Final, hoping to find notes by either Korchnoi or Huebner. Although all the games from the match were included in Informant 30, they were anotated by Matanovic rather than by the players. Like many analysts, Matanovic did not use the '∞' (unclear) symbol in his notes.

There were no other games annotated by Korchnoi (did he ever write for Informant?), but there was a single game annotated by Huebner, and it had several instances of the requisite '∞' symbol. Two of them stemmed from analysis of the position shown in the diagram.

Huebner's Informant annotations were always deep, often on a par with the notes that players like me produce today using software like Fritz. Huebner did his work before computers were used for that sort of thing. I'm sure of this because the current game was played when the Apple II was at its peak popularity. Around that time I bet a friend that, although I had never played against his Apple II chess program, I could checkmate it in less than 20 moves. I won the bet.

1980 Tilburg (Inf. 30/270)
Kavalek, Lubomir

Huebner, Robert
(After 12...Bf8-e7)
[FEN "r2qk2r/2pbbppp/p4n2/1p2R3/3N4/1P6/1PP2PPP/RNBQ2K1 w kq - 0 13"]

Huebner played 13.Qe2, adding to the pin on the Black Bishop and preventing White from castling O-O. His notes to the move investigated the alternative 13.Qe1 with the same idea. I would play 13.Qe2 without giving it much thought, but that's the difference between a world class grandmaster and a run-of-the-mill master. After 13.Qe1, Huebner gave

  • 13...Kf8 14.Nf5 (14.Bg5 Bd6 15.Re2 Bxh2+) 14...Bxf5 15.Rxf5 Qd7 16.Rf3 Bd6 , and

  • 13...c5!? 14.Rxc5 (14.Nf3 Bg4 with the idea 15...Ra7) 14...O-O 15.Re5 Bd6 16.Re2 Bg4 17.f3 Bc5 18.c3

In other words, he rejected 13.Qe1, because the variations were less clear than his game continuation. After 13.Qe2, the game continued 13...c5. Here Huebner gave 13...Ng4 14.Rd5 O-O 15.Nc6, and 13...Kf8 14.Bg5. After the further 14.Nf3, he gave 14.Rxc5 O-O 15.Nc6 (15.Re5 Bd6 16.Re3 Ng4 with advantage to Black) 15...Bxc6 16.Rxc6 Re8 with compensation for the material.

Note how the game variations parallel the variations in the note to 13.Qe1. This is the mark of a diligent analyst, investigating the difference between two very similar moves. To play through the complete game see...

Robert Huebner vs Lubomir Kavalek, 4th Interpolis 1980
http://www.chessgames.com/perl/chessgame?gid=1462518

...on Chessgames.com. Although there is no kibitzing on the game, it is worthy of attention. Why else would Huebner have chosen to annotate it?

2 comments:

Michael Goeller said...

I think his notes on Qe1 are more evidence that Huebner typically over-annotated his games. You ask why he annotated this one? Because of the Exchange sac at the end and nice attack. That's the much more interesting part of the game. Why bother with Qe1 at all? He should spend more time on the sac.

Tom Chivers said...

The differences between 13.Qe2 and 13.Qe1 are yet another example of the kind of minutiae Grandmasters deal in daily, but which even with annotations such as these remain somewhat mysterious to ordinary mortals such as myself.

Trying to read between the lines, so to speak, I believe the difference between the two moves is as follows. On e2 the queen covers g4; therefore after 13.Qe1, 13...Kf8 is the best reply for black because the move white would now like to play -14.Bg5- can be met by 14...Bd6 15.Re2 Bxh2+ intending ..Ng4+ and ..Qxg5. This line doesn't defend for black, however, if the queen is on e2, since ..Ng4+ can be met by Qxg4.

However, having the queen on e1 does have an advantage, in that it allows the retreat of the white rook to e2. That is why the game defence of 13...c5 is not as good against 13.Qe1, because white can play 14.Rxc5 and then extricate his rook via e5 to e2.

That leaves the question, which is better: 13.Qe1 Kf8, or 13.Qe2 c5 ---- ? Perhaps Hubner answers this question in his analysis by (say) giving 13.Qe2 a ! or by rating the position after 14.Nf3 as += rather than the unclear positions 13.Qe1 ---- I don't know. If he doesn't, then he is assuming his Grandmaster audience will be able to read between the lines and judge which move was the better bet. From the snippet given on your post Mark, I actually can't judge. Incidentally, I believe the Rybka 3 book gives 13.Qe1 as slightly superior to 13.Qe2, if I've understood it correctly.

Picking apart these sorts of things is a constant challenge for a weak player such as myself. Even in wordy annotations such as those published in New In Chess, these kinds of details are often simply implied in the analysis ---- a variation is bracketed in one line, promoted as the mainline after a different move, and it is up to the reader to work out the differences. Whatever strength player you have to be to read such analysis and just get it, it is stronger than I. I struggle, typically rather slowly, to do so each time.

Michael's comment is an interesting one. I am not sure I agree. There are different styles of analysis: Huebner's is at one extreme, a deep, concrete striving for the truth, no matter how complicated and unwieldy. There are advantages and disadvantages to that style when compared with its opposite. Let's take Ray Keene's style as its extreme opposite, i.e., an annotational style which points out the positional and strategical features, highlight the human drama, and on the concrete side sprinkles on a handful of two move or three move tactics, at the most. How would Ray Keene have annotated 13.Qe2, if at all? "White doubles on the e-file to prevent castling and pile on the pressure." Something like that. So what? I could tell you that.

So I think it probably was worth analyzing 13.Qe1 ---- but does Huebner's analysis stand up to scrutiny? A quick, preliminary look with Rybka suggests that in fact he may have missed some interesting details. So would a more exhaustive analysis have been better? Or would something like "13.Qe1 with the idea of meeting 13...c5 with 14.Rxc5 and then Re5-e2" have been a reasonable middle-ground?