Watch out for ****KRBN

As I pointed out in A Chess960 Catastrophe, 2700+ GMs can get caught by chess960 tactical surprises, so it's no wonder that lesser players get caught as well. The diagram is one move into a game that started with SP301 QNRBKRBN. The players might not have been super GMs, but they were both top-15 chess960 players at Schemingmind.com. I imagine that after White's 1.Ng3, Black reacted somewhat automatically with 1...Ng6. Symmetric opening play is, after all, a standard strategy for Black in many chess960 start positions.

Black must have been very surprised when his opponent replied 2.Nh5, attacking the Pawn on g7, a Pawn which cannot be defended. Even worse, the Knight move threatens smothered mate. Black answered 2...f6, along with the comment, 'Ugly starting position and I bungled it.' It's like giving odds of Pawn and move.

SP301 QNRBKRBN

What should Black have played in response to 1.Ng3? One idea is 1...f6, opening the diagonal for the Bg8. If then 2.Nh5, Black has 2...g6, and if White continues 3.Ng7+, the Knight is trapped after 3...Kf7. Another idea is 1...b6, with a counterattack on White's g2. The obvious move is 1...g6, preventing the White Knight from going to either f5 or h5, but I don't like this move because it does nothing for Black's development and takes the natural square from the Knight on h8.

A little thought convinced me that any position starting ****KRBN is susceptible to the same tactic. How many such positions are there? Since the remaining White Bishop must be placed on one of the two light squares, when the other three pieces can then be placed on any empty square, there are 2 x 3 x 2 = 12 positions. The same tactic is possible for any twin of those 12 positions, i.e. the positions starting NBRK****, although these have the additional possibility of castling 1...O-O-O.

How do the CCRL engines (see the right sidebar for a link) handle SP301? Of the 82 games currently archived there, White opened 1.Ng3 in 63 of them. The following table shows how many times Black responded with various moves:-

1...f6 x 34
1...f5 x 16

1...g6 x 5
1...b6 x 3
1...e6 x 3
1...e5 x 2

The move 1...f6 is the overwhelming favorite, followed by the surprising 1...f5. It's no surprise, however, that the unfortunate 1...Ng6 was not selected by any of the engines.

The Clock Is/Isn't Ticking

In last week's post on Naming Things, I issued a rain check because I was too busy researching the Sicilian Defense for a gaggle of correspondence games. This weekend I have to issue another rain check for the same reason. The games have all advanced to the range of 8-10 moves played and are all still in theory. That means the variations have been played many times before, even if it happens to be the first time I'm playing one in a correspondence game.

I use Chesslab.com as a database for opening research. It's well maintained and reliable, and serves my purposes very well. I know many players like to maintain their own databases of master level games, but I've never had much interest in doing so. It's not that I can't -- my professional background is in database management -- it's that I know how much effort database maintenance requires and I have other ways I like to spend my time.

As for my ongoing Sicilians, the game with the least theory on the current position has about 40 examples on Chesslab. The games with the most theory still have many thousands of examples, including hundreds of games played at the super grandmaster level (>2700). I'll probably reach move 15-20 in those games before I have to start relying on my own chess sense.

In contrast to those Sicilian games, I started two new chess960 correspondence games this past week. Neither game has reached the third move and both are already demanding deep analysis of candidate moves and their consequences. The event is the fifth round of a knockout tournament, so my opponents are both experienced chess960 players. One slip at an early stage will likely mean a lost game.

A few years ago I wrote a post about Differences Between Chess and Chess960, where everything I said still rings true. I might add that traditional chess covers only a tiny portion of the universe of interesting chess positions. The difference between chess and chess960 is like the difference between a small hobbyist telescope in your back yard and the Hubble space telescope.

I'll close this post with another plug for Chess960 Jungle. HarryO wrote this week about The chess clock - when it should start ticking. It's an important subject that's given short shrift in the implementation of online chess960. I imagine that most chess960 software developers are building on an existing implementation of traditional chess and reuse the same clock specifications that have been developed for the traditional game. The considerations for chess and chess960 are, however, a world apart in the opening phase.

In traditional chess you're basically on autopilot for the first few moves because you've seen the start position thousands of times. In chess960, you're on your own from the very first move. It is fundamentally unfair to let White consider the first move without using clock time, then start the clock for Black as soon as White has moved. You could say that the unfairness applies to all players 50% of the time, but if you're playing a chess960 game for an important prize and you happen to have Black, that 50% argument doesn't help you. I could say a lot more about this last topic, but the Sicilians beckon. Maybe next week...

Naming Things

No time for a real post this weekend. I started a dozen new correspondence games during the past week, most of them in traditional chess (SP518 RNBQKBNR), and am wading in opening database research. All of the SP518 games are for the third (semifinal) stage of a multi-stage event, so my opponents are good players.

All but one of my SP518 games as Black started 1.e4, which is also the move I chose in my games as White. In my 'White' games, all but one of my opponents responded 1...c5, which is also my preferred move. That means all but two of my games started as Sicilians (1.e4 c5), and another game transposed to a Sicilian after a half dozen moves. There's no doubt about it. Chess960 opening play is a *lot* more interesting than traditional chess. The difference is between making your own path or following in the footsteps of others.

In the absence of a more substantive post on my part, I'll recommend taking a look at Chess960 Jungle, where HarryO has been naming things: Naming the Bishop Pairs and Naming the Knight Pairs. It's a good idea that merits careful consideration.

Before reading HarryO's suggestions, I had already decided that the best way to classify chess960 openings is the start positions of the Bishops. There are 16 Bishop combinations, with 60 positions per combination. The positions with two corner Bishops (B******B) have much in common with each other, as do the positions with the Bishops in traditional positions (**B**B**). For those who insist on cataloging first moves in all 960 start positions, ECO becomes 'Encyclopedia of Chess960 Openings', with volumes A through P (the 16th letter of the alphabet).

Now I have to get back to researching Sicilians. I hope I'll have time for a more original post next week. In the meantime, don't miss the Military Knights.

CCM 2003

After writing the post on Svidler - Leko, Mainz 2003, I took another look at the links for the year 2003 on Chess960 @ Chess Classic Mainz (CCM). That was the year when the CCM organizers introduced the basic format that was kept until the end of the annual series in 2011 (as reported in No Place for Chess960) -- the winner of a 'World Championship' match played the winner of an open tournament in another match held the following year.

Before the 2003 match, GM Svidler gave an interview where he discussed winning the CCM Chess960 Open in 2002, Thanks to Twins I am back in the Top-10.

Q: Can you benefit from chess960 in a normal chess game? A: So far I haven’t played enough chess960 to draw any definite conclusions, but it can certainly help develop intuition and tactical alertness, since you can’t rely on knowledge any more and have to improvise in every game.

After the first two games of the match, both draws, he said,

We could not entertain the people that much today, but there were some funny moments in the first game, because we were not sure how to castle in a certain position. We asked the arbiter after the game and he showed us how to do it. But I will promise you: we will learn!

I can't imagine that happening in a high level chess960 match today. After winning the match +2-1=5, Svidler said somewhat modestly,

I was outplayed in the opening in all games, but as soon as I started getting the pieces off the back rank, I played good chess.

The games were played at a time control of 25 minutes per game plus 10 seconds per move (G/25+10), while 'Starting positions will be known only about five minutes prior to the start of the games.' Here are some quotes from the report on the open (G/20+5), won by Aronian (2. Chess960 Open):

Zvjaginsev: "I have never played Chess960 before. It is quite unusual to think from move one on, because you have to play basic chess. Some positions were quite difficult today, e.g. in the third round in which I played a very sharp game. My opponent could have drawn but I was lucky. But don't ask me to recall the initial positions because I have no idea!"

I.Sokolov: "If I would have to play 100 rated games a year in this chess variant, I would go completely mad! It is so different, but I have to admit that I like playing chess960 occasionally. It will not replace classical chess, but it is good alternative. I am very tired now, because you have to start thinking on move one! More time would be better for the quality of the games. I would probably think about the initial position for at least 20 minutes. I don't know yet what is more important: developing the pieces or trying to occupy the centre quickly."

Aronian (after the last game): "I don't really like playing chess960, because I have to think too much in the initial position. I am not used to use my head in the opening. I have to admit that I was very lucky in the last round, the position was probably losing. There were some odd positions in this tournament, some of them were very hard to play with black. But in normal chess it is not always fun to play with the black pieces as well. I think it is important to open lines for the Bishops, but you know, every position is different and difficult."

Along with the two main events, both Svidler and Leko gave 20 game simuls, with 20 different start positions each. A real novelty was a game Chess Tigers against the Rest of the World, featuring SP017 BNQBNRKR and one pair of moves per day. I haven't tried to replay the game from the many comments in that last link, but it might be possible.

The most remarkable document to emerge from CCM 2003 was written by Mr.CCM himself, Hans-Walter Schmitt: A pleading for Fischer’s ideas - say "Yes" to Chess960. It starts,

"Experts are playing for experts": With breakneck speed, the two world champions Ponomariov and Anand are playing the 21 opening moves of the Najdorf variation in the Sicilian Defense. The time used so far is 16 seconds on one side, 21 seconds on the other side – one observer in the tournament hall sighs protestingly "The organiser should make them play slower in the opening – there's no chance of enjoying the game". Despite grandmasters Yussupov's and Lobron's commentary via headphone, despite a video screen 4 times 5 meters in size as in a cinema, despite the directly visible players on stage, the ordinary chessplayer sitting comfortably in his chair understands nothing at all.

And later,

What happens to chessplayers who have to subdivide their time differently after having set up a family or having started a time-consuming job? They stop playing chess because with abandoning only one thing, they gain an enormous amount of time which they can spend on other, more important aspects of life. Furthermore, it is not beneficial for their image or even psychologically unacceptable not to present themselves successful in their favorite hobby. Most often, they are defeated easily by people with enough spare time, just because they have the time to continue studying more and more theory. Thus, they are not found in the club-structure of today anymore. The organised chess, especially chess played in clubs, generally has nothing to offer for socially successful people with not enough leisure time.

There are many other chess960 discussion topics in this visionary document and I would like to return to it in a future post.

Svidler - Leko, Mainz 2003

In my previous post, The Path of Bliss?, I mentioned a game by GM Peter Svidler from the recent 2011 European Championship where Svidler opened with 1.g3. The blog Chess960 Jungle expanded on this with Chess960-SP080: The Buckingham Palace Defense, where HarryO wrote, 'This amazing 1.g3 game looks like the players grew up playing chess960'. I don't know about the player of the Black pieces who, after seeing the move 1.g3, 'sat there, trying hard to think, and simply wasn't able to find a decent reply', but Svidler was a multiple winner at Chess Classic Mainz (CCM).

According to my page Chess960 @ Chess Classic Mainz, the super-strong Russian GM won the 1st Chess960 Open in 2002, where he earned the right to a match against the reigning chess960 World Champion Peter Leko in 2003. He won that match, followed up with match wins against Aronian in 2004 and Almasi in 2005, then lost to Aronian in 2006. After the lost match, he appears to have hung up his chess960 spurs and did not compete in CCM again.

The first match against GM Leko was a close affair. Svidler had White in the first game and the first three games were drawn. In the fourth game, Leko won with Black; there was another draw; Svidler won with Black to level the score; and there was another draw. The last game saw Svidler playing with White in a game with start position BNNRQKRB (SP051).

The players started 1.d4 d5 2.g3 g6 3.O-O O-O 4.b3 Nc6 5.e3 b6, leading to the first diagram, where the start position is still discernible. Among the special characterisitics of the position are the four Bishops placed in the corners, and the right to castle O-O on the first move. In the first five moves, both players have released their Bishops and castled O-O. The game continued 6.c4 dxc4 7.bxc4 Nd6 8.Nd2 Na5 9.Bxa8 Rxa8, as shown in the second diagram. One pair of Bishops has already been swapped off, a frequent occurrence in start positions where the Bishops are in the corners.

In the second diagram, White appears to be in trouble. The c-Pawn is attacked twice and defended only once. If 10.Qe2 to defend the Pawn, then 10...Qa4 attacks it again, when there are no more pieces to aid in the defense.

Svidler solved the problem with 10.Ncb3, sacrificing the Pawn. After 10...Naxc4 11.Nxc4 Nxc4, he continued 12.Rc1 Qa4 13.Qe2 b5 14.Nc5 Qa5 15.Bc3 Qa3, developing his forces on every move while Black was obliged to play defensively. Now there came 16.Rc2, threatening the terrible 17.Rb1 with a Queen trap. Leko played 16...b4, letting White recover the Pawn. White's position was better than Black's and Svidler won in 63 moves to become the second chess960 World Champion after Leko.

The Path of Bliss?

The site ChessInTranslation.com appeared in the first half of last year and immediately became a big hit with fans of top level chess. I mentioned it on my main blog in July 2010 (see ChessInTranslation.com) and have been following it ever since. A recent post translating GM Shipov's comments on a game from the 2011 European Championship at Aix-les-Bains, France, sounded like a promotion for chess960: The massacre of the innocents.

Kids and Hulking Brutes: To catch your opponent out in an opening line, to arrive for a game and, without making even a single move of your own at the board, to get a won position – that's the dream of any chess player. And if, in former times, such victories were won through the blood, sweat and tears of long home analysis – moreover, without complete confidence in the quality of the analysis, and knowing that your opponent would sometimes find strong refutations at the board – the situation has now changed.

Nowadays you need to choose the sharpest of lines, where you've looked at the complications with a computer, while your opponent hasn't. And the game's in the bag!

It's as if there's a fight between young school pupils, but from behind the back of one of the kids an older hulking brute towers up and lands crushing blows against his defenceless young opponent.

According to Shipov's notes, Black(!) unleashed a novelty on move 9, followed preparation for a few moves ('14…Be6!; This was the first move in the game on which [Black] spent even a little time thinking'), and accepted his opponent's resignation on the 19th move. I could comment on the notion that 'the dream of any chess player' is 'to get a won position without making even a single move of your own at the board', but I'll leave that to scholastic chess coaches and other proponents of chess in the schools. Instead, I'll just point out that when Shipov asks,

What is there for us, kids, to do in the current traumatic situation? There are only two options.
1. Prepare hard with a computer around the whole perimeter of the opening repertoire.
2. Avoid topical and sharp lines, taking Svidler's approach.

then answers,

There's also a third, purely hypothetical, possibility: to reach the level of Kasparov and Kramnik, i.e. to become a hulking brute yourself! To learn to play an unfamiliar position at the same strength as a decent computer program.

he is overlooking another possibility that isn't hypothetical and doesn't require becoming a 'hulking brute'. Readers of this blog will know exactly what I mean.

A reference at the bottom of that post points to another Chess In Translation post that had appeared a few days earlier and that I had also selected for mention on this chess960 blog: The future belongs to 1.g3!. Here's Shipov, again on a game from Aix-les-Bains:-

I've just returned to the hotel after watching the start of the fourth round. I'm in a state of shock. At the tables I said hello to the five-time Russian Champion Peter Svidler, then I had a look at his game – and was dumbfounded. The move 1.g3 had been made on the board, and his opponent – the Greek grandmaster Mastrovasilis – was spending a long time studying the position and, it seems, couldn't believe his own eyes. No doubt he'd been preparing from yesterday evening to lunch today for the normal 1.e4, had unearthed a couple of deep saving resources – and then he gets an incredible surprise like that at the board. He sat there, trying hard to think, and simply wasn't able to find a decent reply.

Here's the game on Chessgames.com: Peter Svidler vs Athanasios Mastrovasilis; 12th European Individual Championship 2011. Later Shipov added,

[Svidler] had the urge to simply play chess – to leave the beaten paths and recall his youth. After all, in our childhood all of us, out of ignorance, played mischievous setups in which you had to start thinking from the very first move. And everyone will confirm that was the happiest period of their life. So why not try to recover at least a fraction of that carefree past?

That truly is happiness – not to have to cram up on a ton of variations in the morning before the next round, not to have to study new games, not to have to sit at the computer expecting miracles from chess engines.

If that 'truly is happiness', then chess960 is bliss.

Parallel Games II

In my previous post, Parallel Games, I used a pair of games from Schemingmind.com's Chess960 League - Season One, so it's natural that I follow up with two games from Season Two. The opponents are again both top-10 players on the site and, as luck would have it, I've already had the opportunity to play both of them, finishing 0-2.

The players were assigned start position SP287 NRKNRQBB, which reminded me of the position discussed in NR****Q* & *Q****RN. The difference is that in the current games we have NR****B* instead of NR****Q*, where the Bishop instead of the Queen is positioned to attack the undefended a-Pawn on the diagonal. The top diagram shows the position after 1.f4 f5 2.Bxa7 Bxa2 3.g3 g6 4.Nb3 Bxb1 5.Kxb1 Nb6 6.Bxb8 Kxb8 in the first game; the bottom diagram shows the position after 1.f4 Nb6 2.g3 f5 3.Nb3 g6 4.e4 fxe4 5.Bxe4 e5 in the second. I show one extra move pair in the first game to close the sequence of tactical recaptures.

Both players chose to open 1.f4, a move which attacks the a-Pawn with the Bishop on g1. In the game with NR****Q*, Black had the option of sacrificing the a-Pawn in return for a gain of time by encircling the Queen. In the present position, Black doesn't have this option, because the Bishop on a7 attacks the Rook on b8, threatening to win the exchange. In place of the Pawn sacrifice, Black has the possibility of copying White's tactic with 1...f5. Now the Bishop on g8 attacks the Pawn on a2, threatening to win the Rook on b1.

It's instructive that in one game, Black copied the tactic to win the exchange, while in the other Black declined by blocking the diagonal with ...Nb6. This apparently forced White to do the same. In the first game both players forfeited the castling option by capturing the Bishop with the King. In the second game you might guess that the players would castle O-O-O; in fact, only White castled and, in the other direction, by O-O. Here are the PGN scores of both games, again courtesy of SchemingMind.com.

[Event "Chess960: Chess960 League"]
[Site "SchemingMind.com"]
[Date "2006.05.10"]
[Round "-"]
[White "Tyler"]
[Black "dmichael"]
[Result "1/2-1/2"]
[SetUp "1"]
[FEN "nrknrqbb/pppppppp/8/8/8/8/PPPPPPPP/NRKNRQBB w KQkq - 0 1"]

1.f4 f5 2.Bxa7 Bxa2 3.g3 g6 4.Nb3 Bxb1 5.Kxb1 Nb6 6.Bxb8 Kxb8 7.d4 e6 8.c3 d5 9.e3 Nf7 10.Nf2 Nd6 11.Kc2 Qh6 12.h4 Kc8 13.Nc5 Kd8 14. Bf3 Ke7 15.Qh1 Nbc4 16.b3 Na3+ 17.Kd3 Qf8 18.Ra1 Ra8 19.h5 Bg7 20. hxg6 hxg6 21.g4 c6 22.Kd2 b6 23.Na4 Ndc4+ 24.Ke2 fxg4 25.Nxg4 Qf5 26. Ra2 g5 27.Qc1 Qb1 28.Qxb1 Nxb1 29.Kd3 b5 30.bxc4 Rxa4 31.Rxa4 bxa4 32. Bd1 dxc4+ 33.Kc2 Na3+ 34.Kb2 Nb5 35.Bxa4 gxf4 36.exf4 1/2-1/2

[Event "Chess960: Chess960 League"]
[Site "SchemingMind.com"]
[Date "2006.04.29"]
[Round "-"]
[White "dmichael"]
[Black "Tyler"]
[Result "0-1"]
[SetUp "1"]
[FEN "nrknrqbb/pppppppp/8/8/8/8/PPPPPPPP/NRKNRQBB w KQkq - 0 1"]

1.f4 Nb6 2.g3 f5 3.Nb3 g6 4.e4 fxe4 5.Bxe4 e5 6.Bc5 d6 7.Bxb6 axb6 8.Qh3+ Ne6 9.fxe5 Bg7 10.Nd4 c6 11.Nxe6 Rxe6 12.exd6 Qxd6 13.Bd3 Qd7 14.Ne3 b5 15.a3 Kc7 16.Qg4 Qe7 17.Qf4+ Be5 18.Qg4 Rd8 19.O-O h5 20. Qf3 h4 21.g4 Bd4 22.Qf4+ Kc8 23.Rfe1 Bc5 24.Qf3 Qg5 25.b4 Rf8 26.Qh3 Bb6 27.Bf1 Rd8 28.Kh1 Rxd2 0-1

Since the games ended with different results, it's natural to look for the reason in the early moves. Is it a coincidence that the draw resulted from the game where two pieces -- a Rook and a Bishop -- disappeared from the board in the first few moves?