26 December 2009

Christmas @ Schemingmind.com

I don't normally comment on my games while they are in progress, but this is worth an exception. Note the King on c1.


Don't try to give me advice on the game (SP732: RBKNQNBR), because we are well past the position shown in the diagram. I played 4.d4.

20 December 2009

Extravagant Openings in Chess960 (cont.)

Continuing with Extravagant Openings in Chess960, the previous post established that the most popular first move of the CCRL engines for White in my random start position (SP201: QNRBBKNR) was 1.Nf3. Chosen in exactly 1/3 of the 24 sample games on file, it gives the position shown in the diagram.


Start Position 201 • 1.Nf3

The same sort of analysis I made for White, respecting various positional principles, can be made for Black. This gives the following options for non-extravagant moves:

  • Three of the four central Pawns can advance two squares (1...c5, 1...d5, & 1...f5), the e-Pawn being excluded for a tactical reason;

  • all central Pawns, plus the b-Pawn (to develop the Queen), can advance one square; or

  • the Knights can jump to their natural development squares (1...Nc6 & 1...Nf6).

The following table shows which moves were chosen by the CCRL engines.

12 x 1...Nf6
  5 x 1...c5
  2 x 1...d6
  2 x 1...f5
  1 x 1...d5
  1 x 1...e6
  1 x 1...h5

In exactly half of the games starting 1.Nf3, the move 1...Nf6 was selected, and five of the other six moves were predicted by the previous analysis. The only exception is 1...h5, which was chosen by the same engine that played 1.h4 successfully on the first move. After 1.Nf3 h5, the game continued 2.c4 h4 3.d4 c5 4.d5 d6 5.Nc3 h3 6.g3 Bd7, an extravagant follow-up to an extravagant first move! It eventually ended in a draw.

While I was working on this post, I started to wonder why the advance of a central Pawn one square on the first move is considered dubious for White, but acceptable for Black. This is certainly true for traditional chess (SP518: RNBQKBNR) and probably true for the position at hand, SP201. It's a question for which I don't have a good answer.

19 December 2009

Extravagant Openings in Chess960

Earlier this month on my main blog Chess for All Ages, I introduced the concept of Extravagant Openings in chess. The reason wasn't to invent yet another synonym for 'unorthodox chess openings', of which I gave a comprehensive list in that same post. It was rather to have a term I could use for dissecting chess960 openings. In the follow-up post What Makes an Opening Extravagant?, I defined extravagant openings as those where a player 'speculates on certain values at the expense of other values' and gave a couple of well known examples from traditional chess.

Taking the concept from chess to chess960 might be easier to discuss by way of an example. I used the dice-rolling method described in A Database of Chess960 Start Positions, and selected a random start position. It is shown in the following diagram (SP201: QNRBBKNR).


Start Position 201

I then looked at the first moves played by the CCRL chess960 engines (see the link in the sidebar and my previous post Chess960 Opening Theory) in this position. Since the CCRL insists on playing chess960 with 'Book learning: Off for all engines', the engines play openings according to how various positional principles -- the center, open lines, piece activity, etc. -- are weighted by the software. This guarantees variety throughout the games.

A glance at the diagram is enough to see that in SP201 the first moves conforming to positional principles are similar to the choices in traditional chess (SP518: RNBQKBNR):

  • One of the four central Pawns can advance two squares (1.c4, 1.d4, 1.e4, & 1.f4), or

  • the Knights can jump to the same natural development squares as in SP518 (1.Nc3 & 1.Nf3).
What did the CCRL engines play? The 72 games with start position QNRBBKNR saw seven different first moves:-

24 x 1.Nf3
16 x 1.d4
14 x 1.c4
  7 x 1.f4
  6 x 1.d3
  3 x 1.Nc3
  2 x 1.h4

Five of the six moves predicted above are in the list, most notably the first four. Only 1.e4 is missing. There might be a tactical reason for not choosing 1.e4 (attacks by ...g6, ...g5, or ...Nf6?), but I don't see anything that is not easily parried.

Also in the list are two moves not predicted above: 1.d3 and 1.h4. Both moves could therefore be classified as extravagant openings. The move 1.d3 has the same drawback as 1.d3 or 1.e3 in traditional chess; it fails to contest the center. The move 1.h4, one of the worst first moves in SP518, looks just as bad in SP201; it fails to do anything useful except liberate the Rook on a1, the same Rook that White needs to castle O-O.

Also worth noting is that the six games starting 1.d3 resulted in a +3-2=1 edge for White, while 1.h4 resulted in +1-0=1. Just as in life, extravagance in chess isn't necessarily a doomed cause. We would all be a lot poorer if it were.

13 December 2009

Practical Issues for Hosting Chess960

Chess960 @ Chess.com isn't just about playing (see What Are the Odds? for a post on my first games there). It's also about community. The main forum at Chess.com for the topic is Forums > Chess960 and Other Variants, which is more active than most forums on the subject.

A couple of threads on the forum caught my attention over the past few weeks. The first was Chess960 Tournaments: Variety. Although it addressed the strategy of running chess960 tournaments at Chess.com, the question applies to all such tournaments.

In chess960 tournaments, how often is the same starting position played? (A) The same starting position is used for all games of all rounds of the tournament; (B) The same starting position is used for all games of the same round of the tournament, but for every new round of the tournament, there is a new starting position; (C) [...]

It appears from comments to the thread that Chess.com uses strategy '(A)'. I also know from past investigations that the premier annual chess960 event, Chess Classic Mainz (see Chess960 @ Chess Classic Mainz), uses '(B)' for its chess960 open. Tournaments at Schemingmind.com (see Pyramids and Dropouts for background) use a strategy that the Chess.com author called '(D) There is a new starting position for every single game of the chess 960 tournament'. Which strategy is best?

The second thread was Subversion of the Spirit of Chess960, which addressed an issue that is more a problem for chess960 than for traditional chess.

Chess.com allows a player to cancel a game within a small number of moves. This seems like a reasonable policy for normal chess, but it has peculiar implications for chess960: Any player can start a bunch of games and selectively choose only those starting positions which match a given set of criteria.

Cancellation is particularly annoying in chess960 because, as I pointed out in Differences Between Chess and Chess960, the first few moves of chess960 require a lot more thought (i.e. work) than the first few moves of traditional chess. I've already had a few games where, just when I had played the first moves and was starting to appreciate the position, the game was abruptly cancelled. If this happens to players who are just starting chess960, it could deter them from continuing to play.

12 December 2009

What Are the Odds?

Since posting about Chess960 @ Chess.com, I've played a handful of correspondence games at the site ('online chess' they call it, as opposed to 'live chess'). By an unusual coincidence, my first two games started with the same position (SP638: RNKQRBBN) and I had White in both games.

A little thought convinced me that 1.e4 was a good first move. It opens diagonals for the Queen and light-squared Bishop, just as in traditional chess, with the added bonus that the e-Pawn is defended by the central Rook. My opponents must have used similar reasoning because they both played 1...e5, to which I answered 2.Nc3. That brought the games to the diagrammed position.


1.e4 e5 2.Nc3

In the first game, my opponent played 2...f6, activating the other Bishop, and I followed suit with 3.f3. Now, however, he (she?; Chess.com users are mainly anonymous) overplayed his (her?) position with 3...d5. This let me snatch the Bishop pair with 4.exd5 Bxd5 5.Nxd5 Qxd5, then menace the Queen with 6.Bd3. I thought I already the better game, but my opponent blundered a few moves later and resigned on the 16th move.

In the second game, my opponent followed my lead by playing 2...Nc6. This let me interfere with his normal development by playing 3.Bc4, when 3...f6 is not possible. He copied me again with 3...Bc5, but after 4.Ng3 g6, I forced his Bishop off the active diagonal with 5.Na4. He retreated with 5...Be7, then after 6.f3, tried to copy my plan with 6...Na5. Following the idea I used in the first game, I played a better retreat with 7.Bd3, leaving e2 for the Queen. Now after 7...f6 8.Qe2 Nf7, I was able to castle 9.O-O-O and again felt I had the better game.

06 December 2009

Four Cornered Bishops

In Attention to the Chess960 Center, I gave a position from the Nakamura - Aronian match, Chess960 World Championship, Mainz 2009, where the players paid attention to the center starting with the game's first moves (1.e4 e5). The positions in the diagram are from the second game of the same match.

Comparing the first diagram (SP451, BRNNKQRB) with the second, an inexperienced chess960 player might get the impression that the players have been moving their Pawns at random. There is, in fact, an instructive chunk of logic behind the moves leading to the second diagram.


One of the main features of this start position is that the four Bishops all start in the corners. Another feature is that all Bishops are protected by their respective Rooks.

The Bishops in the corners usually dictate a positional duel at the beginning of a game, since both players are forced to develop their Bishops on the long diagonals, where they come into contact immediately. The player who is the first to open a long diagonal can exchange Bishops as soon as the opponent opens the same diagonal. Balancing this, the decision to open either diagonal first leaves the opponent the opportunity to open the other diagonal first.

In the current game, White chose to open the a1-h8 diagonal with 1.b3. If Black plays 1...g6, White can exchange Bishops and Black will no longer be able to castle O-O. Black followed a similar strategy on the a8-h1 diagonal with 1...b5, but grabbed more space than White and left the possibility of ...Nc8-b6.

Now White continued the duel on the diagonals by playing 2.f3. This prepares a subsequent g2-g3 without allowing the exchange of Bishops on h1, thereby preserving the option to castle O-O. Black followed an identical strategy with 2...f6.

Now White made an overt play for space in the center with 3.d4, preparing Nc1-d3 and leading to the second diagram. He might not have noticed that his third move closed the a1-h8 diagonal, giving Black the chance to play the surprising 3...f5, advancing the f-Pawn for the second consecutive move. Consequently, Black's dark squared Bishop will have more freedom on its long diagonal than White's light squared Bishop will have.

After three moves each, the contours of the subsequent struggle have been defined and the game is launched. The PGN for the full game, courtesy Chesstigers.de, is:

[Event "CCM9 - Chess960 Rapid WCh"]
[Site "Mainz"]
[Date "2009.07.30"]
[Round "8.1"]
[White "Aronian, Levon"]
[Black "Nakamura, Hikaru"]
[Result "0-1"]
[WhiteElo "2800"]
[BlackElo "2777"]
[Variant "chess 960"]
[SetUp "1"]
[FEN "brnnkqrb/pppppppp/8/8/8/8/PPPPPPPP/BRNNKQRB w GBgb - 0 1"]
[Source "Chess Tigers"]

1.b3 b5 2.f3 f6 3.d4 f5 4.Nd3 g6 5.Qf2 Bf6 6.g3 Qh6 7.e3 Ne6 8.Qe2 Nb6 9.Nc3 a6 10.Nc5 Nxc5 11.dxc5 Qg7 12.Qd3 b4 13.cxb6 cxb6 14.Nd5 Bxa1 15. Nc7+ Kf7 16.O-O Bc3 17.f4 Bxh1 18.Kxh1 Qf6 19.Rbd1 Qc6+ 20.Qd5+ Qxd5+ 21. Rxd5 Rb7 22.Nxa6 Ra8 23.Nxb4 Bxb4 24.a4 Bc5 25.Re1 e6 26.Rd3 d5 0-1

Nakamura appears to have outplayed Aronian already in the opening.

05 December 2009

Attention to the Chess960 Center

In More from Mainz 2009 I quoted GM Grischuk saying,
The first year [at Mainz] I was playing like g4/b4, but in order to play like this successfully you have to be either Aronian or Nakamura. They look to be the only two persons who do it successfully. Kamsky does it but he tries more like f4/c4, but they are g4/b4. It works for them, but for me it was just terrible. Since then I try to play a more central approach at the starting stage.

After playing chess960 for almost a year and a half, I have some idea what Grischuk is saying. There are two distinct, fundamental ways to treat a chess960 opening. The first way is to follow traditional chess opening principles, of which one of the most important is to pay attention to the center. The second way is to pay less attention to the center, but by taking into account the specific start position, to emphasize the rapid development of the pieces to good squares, even if this means making early moves like g4 or b4.

Grischuk didn't pick the names Aronian and Nakamura at random; they have been among the most successful competitors at recent Mainz events. Aronian won the main chess960 event at Mainz on several occasions (see Chess960 @ Chess Classic Mainz for a list of his accomplishments), although he finished second to Nakamura in the most recent tournament (see CCM9: Nakamura, Grischuk, and Rybka).

This year the two 2700 GMs played two games in the preliminary stage of the closed Rapid World Championship and then met in the final stage. Nakamura lost to Aronian +0-1=1 in the preliminaries, but won +3-0=1 in the finals.

Most of the six games between the two players were more like the g4/b4 variety, but one game used more classical principles in the attention paid to the center. The following diagram shows a position from game three of the final match.

Chess960 World Championship, Mainz 2009
Aronian, L.

Nakamura, H.
(After 8...Bh6-g7)

At first glance the position might be mistaken for the Philidor Defense (1.e4 e5 2.Nf3 d6) in traditional chess, but the Queens in the corners, the Kings on the c-file, and the position of the Rooks, especially the Rook on d3, are a dead giveaway that the position is chess960 (SP666, RNKRBBNQ). White continued 9.Be3, defending the attacked d-Pawn a third time and letting the Rook on d3 retreat if necessary. The PGN for the full game, courtesy Chesstigers.de, is:

[Event "CCM9 - Chess960 Rapid WCh"]
[Site "Mainz"]
[Date "2009.07.30"]
[Round "9.1"]
[White "Nakamura, Hikaru"]
[Black "Aronian, Levon"]
[Result "1-0"]
[WhiteElo "2777"]
[BlackElo "2800"]
[Variant "chess 960"]
[SetUp "1"]
[FEN "rnkrbbnq/pppppppp/8/8/8/8/PPPPPPPP/RNKRBBNQ w DAda - 0 1"]
[Source "Chess Tigers"]

1.e4 e5 2.Nf3 g6 3.Nc3 c6 4.g3 d6 5.d4 Bd7 6.Rd3 Na6 7.a3 Bh6+ 8.Bd2 Bg7 9.Be3 exd4 10.Bxd4 Nc5 11.Rd2 Nf6 12.Ng5 Rf8 13.O-O-O O-O-O 14.e5 Ng4 15.f4 Ne6 16.Nxe6 Bxe6 17.h3 dxe5 18.Bxa7 Nf6 19.Ba6 e4 20.Qg1 Rxd2 21. Bxb7+ Kxb7 22.Qb6+ 1-0

The ratings (WhiteElo and BlackElo) are special chess960 ratings calculated by Chess Tigers. In my next post, I'll show a position from another game between the two players where less attention was given to the center.