13 February 2011

Oversight or Intentional?

In my previous post, Nakamura Sac Attack, I covered GM Hikaru Nakamura's qualification from the 2008 Chess Classic Mainz (CCM8) Chess960 FiNet Open to the 2009 (CCM9) Chess960 Rapid World Championship, where he beat GM Levon Aronian for chess960's most prestigious title. A few months before the FiNet event Nakamura won entry and expenses to the open by finishing first in the Internet Chess Club's (ICC) 2008 Chess960 Championship.

The ICC Chess960 Championship was held from 2007 to 2009, and in each year the winner won entry and expenses to Mainz (links to Chessclub.com):-

In 2009, just like Nakamura had done the previous year, Grischuk went on to win the FiNet Open, qualifying for the 2010 Chess960 Rapid World Championship, an event which was subsequently cancelled due to lack of sponsors. The games from the 2008 and 2009 ICC finals are available on the ICC site, but were played at a blitz tempo that renders many of them less interesting because of early blunders.

In the following game, the sixth and last game in Nakamura's final 2008 ICC match against GM Dmitry Andreikin (DSquared), I couldn't tell if the critical moves were an oversight or intentional. In start position NRBKNBQR (SP262), Andreikin opened 1.Nb3 and Nakamura (Smallville) replied 1...e5, arriving at the position shown in the first diagram.

Now White played 2.f4, with a double attack on the a- and e-Pawns. The a-Pawn is particularly sensitive, because Qxa7 attacks a Rook that can't be defended. Black played 2...Nb6, when 3.fxe5 left White a Pawn to the good. At this point Nakamura played for complications and open lines with 3...f6 4.Nf3 fxe5 5.Nxe5 d6 6.Nf3 Nf6. The game continued 7.d3 a5 8. c4 a4 9.Nbd4 d5, reaching the position shown in the second diagram.


Here White could have held the extra Pawn with several moves, where 10.cxd5 is one obvious choice, not necessarily the best. Instead he played for rapid development with 10.Bg5 dxc4 11.O-O-O, and the game continued with material equality. Was 1...e5 an oversight? What about 10.Bg5? Whatever the reasons for the two moves, Black went on to win the game for the fifth consecutive full point by the Black pieces. Here is the game's PGN, courtesy the ICC.

[Event "ICC tourney 865 (w22 3 1)"]
[Site "Internet Chess Club"]
[Date "2008.06.22"]
[Round "21"]
[White "DSquared"]
[Black "Smallville"]
[Result "0-1"]
[TimeControl "180+1"]
[SetUp "1"]
[FEN "nrbknbqr/pppppppp/8/8/8/8/PPPPPPPP/NRBKNBQR w KQkq - 0 1"]

1.Nb3 e5 2.f4 Nb6 3.fxe5 f6 4.Nf3 fxe5 5.Nxe5 d6 6.Nf3 Nf6 7.d3 a5 8. c4 a4 9.Nbd4 d5 10.Bg5 dxc4 11.O-O-O c3 12.Nb5 cxb2+ 13.Kxb2 Qd5 14. Nc3 Qd6 15.Qd4 Qxd4 16.Nxd4 Be7 17.e4 Ng4 18.Bxe7+ Kxe7 19.Re1 Rd8 20. Nf5+ Bxf5 21.exf5+ Kf8 22.Nb5 Rd7 23.Be2 Nf2 24.Rhf1 Nxd3+ 25.Bxd3 Rxd3 26.f6 Rd2+ 27.Ka1 gxf6 28.Rxf6+ Kg7 29.Rf3 Rxg2 30.Nxc7 Kh8 31.Ne8 Nd5 32.Re5 Nb4 33.Rf7 a3 34.Re6 Rg1+ {White resigns} 0-1

Note that the PGN lacks a 'Variant' tag, a glitch which did not cause a problem with either program that I used to step through it.

2 comments:

HarryO said...

Really appreciate you digging up those games! I have a few comments and a question as well.

Comment 1:
Us humans are habit based thinkers. If we have played a life time of chess, it is really hard to break out of the mental patterns that SP518 form in our minds. The two errors in the opening by the GM's illustrate the point:
1....e5?! Although not drastically bad, because the bishop is on the SP518 square, the mind says e5 must be ok. D5 was better. Later in the game, white plays 10. Bg5?! which destroys all his momentum. 10. Nb5 or even c5 is much better which is immediately attacking. Can you see how 10.Bg5 is so tempting because of the similarity with the SP518 version of Bg5?

Comment 2:
The point is that it is just a habit! If we were to play just 2000 Chess960 games (which is a drop in the ocean when you think of all the ten's of thousands of SP518 games that one single club player plays online). If we played these Chess960 games we would start to see the patterns of 960. The unique construct of the undefended a7/a2/b8/b1 squares can be easily remembered even with my pathetic memory.

Comment 3:
We feel intimidated by 960 even the GM's do. If you combine the 960 setup in this game with white's fears about Nakamura's amazing talents, you can see how Nak psychologically outwits white with the a-pawn push? Habit again! White feels under pressure because he thinks that he has the same advantage as he has in SP518 and thus must prove it. But there are so many factors and white is often under more pressure than black because black get's one extra move of information than white!

Comment 4:
All intellectual pursuits are pragmatic! Reducing the Chess960 gamut down to a smaller subset of starting positions does not address the essential point. Every position in Chess is pragmatic based and every position is not ideal. This is what Chess960 teaches us. There is no point in pretending that we can prepare specifics in the chess lab at home with a limited subset of opening positions, because we are just covering up our fears that we must face up to pragmatism! We can still do a lot of opening theory at home in the lab even with 960, but the approach is different and the approach itself is pragmatic! Pragmatic thinking is fun. It is just conditioning that makes it so.

Comment 5:
Chess960 is really hard mental work. You have to think from move one. Thus you have to be extra fit and have excellent endurance abilities because of the amount of effort compared to SP518 just to reach the mid-game! This is often a critical decider and also another reason why people are reluctant to take Chess960 up even the GM's.

Question 1:
Do you know how I can simply copy your game list into some chess960 app? Even Fritz 12 rejects the game list once it get's to castling for some reason.

It does not matter, because I can manually enter them.

Thanks again!
Cheers
Harry

Mark Weeks said...

Re 'Do you know how I can simply copy your game list into some chess960 app?', see the last paragraph in the post. Try adding the 'Variant' tag to the PGN headers. - Mark