Tuesday, June 30, 2009

DKos Chess Tournament, Round 3

Not too much to add to the commentary in the Flash player for this one. I really like Black's 12th move, not because it was hard to find so much as just an interesting, tight cluster of mutually supported pieces and pawns.

The commentary to moves 15 and 17 show that this wasn't a stress-free game... I had faith in my ability to lose these types of advantages. I think my decision not to capture on h3 was practically best, no matter what Rybka thinks. White's remaining queen, bishop and rook make it look like nastiness could ensue against g7.

Current standing: 1.5/3

Sunday, June 28, 2009

DKos Chess Tournament, Round 2

It's very rare for me to attribute the primary cause of a loss to my not trying, especially in a serious game. It was clearly the main factor in the following game.

It's interesting to see how the flash player handles variations. I have to make more use of this widget! For this game I'm going to let the widget do my work, and skip .jpg images of key positions.

1. e4 e5 2. f4 Bc5 3. Nf3 d6 4. Bc4 Nf6

This is as far as my "1% repertoire" continues. I've finished a lean repertoire database that contains only positions I've encountered in at least 1% of my games. Up until this point I haven't studied King's Gambit lines much...they're not encountered frequently, and opponents don't typically play main lines. 4.Bc4 is the main move that I encounter here, and the only one in my 1% repertoire (another example of how, at lower levels, sidelines are actually main lines). It's not specifically addressed in Marin's Beating the Open Games, but is likely to transpose to the main line...

5. Nc3 Nc6 6. d3

...as it does here. However, in my database of 7000+ personal games, I've only encountered this tabiya twice! So, until now, I've been justified in not studying the main lines of this opening.


...and here I don't recall my "repertoire" move 6...Bg4, which was played in the other two games. Now that I've finally encountered this game in a serious game for the first time, I'll spend some time reading over Marin's chapter on this opening and map out a main line. I intend to flesh out my "1%" repertoire by mapping out one main line for each variation.

Marin actually has some analysis of 6...a6 in the footnotes. The idea is to preserve the bishop against Na4. This is a common moves in openings such as the Guioco Pianissimo, but here it's costing a tempo that could be used for developing. Some of the lines I'll be investigating are included in the widget's annotations.

7. fxe5 (7. Rf1!? Marin) ... dxe5 8. Bg5 h6 9.Bxf6 Qxf6 10. Rf1

Essentially a novelty (one unrated game in my main database). Marin gives 10. Nd5.


Understandable, but not addressing the need to develop. 10... O-O is preferred by Rybka, who doesn't seem to mind castling into a minority attack. However, I would be inclined towards queenside castling.

I like 10... Bg4, e.g. 11. h3 Bxf3 12.Qxf3 (12. Rxf3 Qh4+ =/+) 12... Qxf3 13. Rxf3 Nd4 14. Bxf7+ Kd7 15. Rf2 Raf8 16.Rc1 Be7 =/+. However, because this line is still somewhat crazy tactically, 10... Be6 may be a better practical choice, e.g. 11. Bxe6 (11. Nd4?? Qh4+ -+) 11... Qxe6= (Rybka also likes 11... fxe6!?, which I find interesting...the doubled pawns take away key squares from White's knights).

11. Bxf7+?! -/+

Looks more dangerous than it is. At this point, however, I felt I had missed something and from this point on stopped trying--which is not like me at all.

11... Kxf7 12. Nxe5+ Ke8?! +/=

This move can't be explained. 12... Kg8!? is the obvious choice, even if it wasn't clear at the time whether Bxf7+ was "!" or "?". 13. Nxc6 is then probably the best chance for White. I would probably have played 13...Qxc6, deflecting the Q off of an important diagonal. Rybka prefers 13... bxc6 =/+.

However, 13. Nf7? would be a mistake: 13... Qxh2! -+ I think this is an example of where opening study can be helpful, not because knowledge of an exact sequence of moves would have helped, but because knowledge of a typical motif would help find the right move here. A Queen and Bishop ravaging the White kingside and creating mating threats appears as a common theme in the King's Gambit.

Even if Black didn't find this killer move after 13.Nf7?, a move such as 13... Qg6 would leave Black with a slight edge, e.g. 14. Nxh8 Kxh8. With the bishop pair offsetting a rook and two pawns, Rybka and I prefer Black. Black has an advantage in development; the white king is still stuck in the center; and White appears weak on the dark squares.

13. Qh5+ g6 14. Nxg6 +/- Black resigns??

Black resigned without even trying to find a solution. It would not have been hard, however, to find the following resource: 14...Bg4!, which would have allowed Black to play on. 15. Qxg4 Rg8 looks like it leaves Black a manageable game (B vs. three pawns). Rybka at this point initially evaluated the position as a modest advantage for White, but as I explored some variations it found a good continuation that leaves White with a distinct advantage: 16. Qh5! Rxg6 (16... Qxg6? 17. Qxc5) 17. d4!, e.g.17...Bxd4 18. O-O-O Qe619. Nd5 Kd7 20. c3!

Which is besides the main point: there was no need for Black to resign in this position.

Friday, June 19, 2009

23 lines

I was thinking about how little opening theory can be applied to my games, so I conducted a little exercise with Chessbase and a database of about 7000 of my games (most from ICC). I constructed an opening tree consisting only of positions that I would expect to face as White in 1% or more of my games.

The result is shockingly small: 23 variations, with none longer than 6 moves.

Because I'm in the middle of an online tournament, I don't want to reveal too much of my repertoire, but just breaking down the percentages after 1.e4:

1...e5 44.6% (My mainest of main lines continues past move 5 just under 1% of the time!)
1...c5 16.0%
1...e6 9.0%
1...d5 5.9%
1...c6 4.6%
1...d6 3.3%
1...g6 1.7%
1...Nf6 1.6%
1...b6 1.4%
1...Nc6 1.1%

The data is somewhat skewed. My database of 7000 or so games goes back 12 years or so, and I only counted positions that I would face in my current repertoire. I've been quite faithful to my repertoire over those years, but there are notable exceptions (I used to play the exchange Spanish, and I avoided the open Sicilian with c3 or Bg5 systems for much of that period). In my current repertoire, no Sicilian line has been seen past move 5 more than 1% of the time! I need to see if I can do some sophisticated filtering to separate out the games that agree with my current repertoire.

The data also shows that at lower levels of play, "minor" variations actually become major. For example, in the Ruy Lopez the Steinitz, Cozio and Bird variations are now "major" lines to contend with.

My repertoire database has been getting a bit weedy, so I may start a leaner, meaner repertoire database using just these 23 lines, and making sure that I map each main line out several moves further. I can repeat the exercise for my Black games, although there I expect an even stubbier tree of variations because White gets to vary with the first move. The goal is to declare a chunk of theoretical turf where I know anything outside its boundaries is encountered less than 1% of the time.

Thursday, June 18, 2009

DKos Chess Tournament, Round 1

I'm taking part in an informal online tournament organized by the Daily Kos community. Apparently Mig Greengard caught wind of this, so there are actually prizes to be had, including a book autographed by Kasparov. I doubt I'll win anything, but you never know.

Here is my first round game, where I had White. The time control was 30 minutes...no incremental, no "insufficient losing chances" appeals to an arbiter. Rybka found a lot of interesting variations throughout the game, but I'm going to try and focus on key moments. Also, I may get some traffic from players that are new to the game, so I'll try to elaborate more on certain concepts that are unfamiliar to newbies. I'll also have a summary list of take-home lessons at the end of the game.

For those unfamiliar with chess annotation symbols such as "?!" or "+/=", I direct you here.

FreeRadical - FightingRegistrar [B35]

DKos Tournament (1), 16.06.2009

[Rybka 3 32-bit (120s)]

Rybka 3 is the name of the analysis software I used to help me analyze the game.

1.e4 c5 2.Nf3 Nc6 3.d4 cxd4 4.Nxd4 g6

I'll have some commentary on the opening because I'm a bit of an opening wonk. I want to re-emphasize for the Kossacks that, for most of us, studying openings in depth isn't the best use of our chess time....but I enjoy it.

The "Accelerated Dragon". A standard Dragon inserts 4...Nf6 5.Nc3. When Black omits 4...Nf6 in the Sicilian, it can allow White to play 5.c4, leading towards a "Maroczy Bind" position. The pawns on e4 and c4 tend to cramp Black's game. In Beating the Sicilian 3 by Nunn and Gallagher, this is the approach they choose for White. I find the weakness of a1 and b2 early in the opening uncomfortable, but I should look into these positions more.

In Winning with the Fischer-Sozin Attack by Gary Lane, the author advocates Bc4 and kingside castling against dragon systems. This is the approach taken in this game.

5.Bc4 (5.c4 Bg7 6.Be3 Nf6 7.Nc3 0–0 8.Be2 d6 9.0–0 Bd7 10.Qd2 is the Maroczy Bind main line) 5...Bg7 6.Be3 Nf6 7.Nc3

7...0–0 So far the main line of this variation. I have 5572 games with this position in my reference database.

7...Nxe4 8.Nxe4 d5 is the "center fork trick" that White has to keep in mind. Here it fails for Black: 9.Bb5 dxe4 10.Nxc6 bxc6 11.Bxc6+ Bd7 12.Bxa8.

8.Qd2 A stereotyped move in these Dragon positions, but not the choice of Grandmasters. 8.Bb3 (avoiding the center fork trick) is the main line, e.g.: 8...d6 9.h3 transposing into a regular (i.e. not accelerated) Dragon.

8...a6 The center fork trick 8...Nxe4! works now, e.g. 9.Bxf7+ (9.Nxe4 d5) 9...Rxf7 10.Nxe4. Black isn't "winning" but does have the bishop pair and a central (albeit hanging) pawn pair. White has a developmental advantage and fewer pawn islands. 9.0–0N ("N" stands for "novelty", meaning we finally have reached a position that's not in my database of about 3.6 million games) 9...b5 10.Bd3 Bb7

11.Rae1 11.f3 was considered during the game. Rybka likes it, but the weakening of the a7-h2 diagonal is unattractive. Also, a long-term consequence of pushing the f-pawn in these positions is that, if Black succeeds in penetrating the queenside (typically down the c-file), he can then sweep towards the kingside down the second rank. In such cases a pawn on f2 often provides important shelter. As an aside (because it was referred to in the DKos open diary), Bc4 and f3 are most commonly associated with queenside castling, in the Yugoslav Attack main line of the Dragon.

11...Ng4 Usually, if White has played this opening properly, ...Ng4 is either prevented with f3 or h3, or is allowed when Bf4/g5 is possible. White does not want Black to capture the e3 bishop if it can be helped. 12.Nxc6 Bxc6 13.Bf4 Qa5

This was the "oh, crap" moment for me, because now it appears that I can't avoid a shattering of my queenside pawn structure. Rybka, however, found a solution.


Rybka found 14.Nd5! Qxd2 (14...Qxa2 15.Nxe7+ Kh8 16.b3 Qa3 17.Nd5+/=) 15.Nxe7+ Kh8 16.Bxd2 Bxb2= Both sides now have isolated pawns on half-open files.

14...Bxc3 15.Qxc3 Qxc3 16.bxc3=/-

The isolated doubled pawns on the c-file are ghastly. However, White does have the bishop pair. For a while, Black's N is offside and the bishop pair helps White stay alive. For those new to chess: a pair of bishops is usually considered an advantage, but a small one (about half a pawn). For more information on tallying the relative values of the pieces, see this article.

16...Nf6 (16...f6 17.c4= Rybka) 17.f3 Rac8

I now wanted to repair my queenside pawns, but I chose the wrong way. 18.a4?! Instead, 18.c4! looks best to me. It gets rid of one pawn off the c-file, and the remaining pawn can be backed by a rook and advanced. Rybka shows that complications can be found by Black, however: 18...bxc4 19.Bxc4 Bxe4!? 20.Bxa6 Ra8 21.fxe4 Rxa6 22.e5 Nd5 23.Rd1 with an unclear position to me. 18...bxa4 19.Bxa6

My plan was to offset pressure down the c-file with my own pressure down the a-file. However, Black isn't so much defending his a-pawn as attacking with it. The passed pawn is very dangerous and much more easily advanced than White's doubled, but not passed, c-pawns. The constellation of pawns on the queenside actually gives Black a space advantage. White will be cramped trying to attack the a pawn and defend the c pawns at the same time.

19...Ra8 20.Bc4 Actually at this point having three loose pieces on the c-file bothered me a lot. I was worried about ...Rfc8, and Rybka agrees. Rybka prefers 20.Bd3 d6, but I felt at the time I needed my bishop on the a2-g8 diagonal to help manage Black's passer. 20...Ra5 (20...Rfc8!?-/+ Rybka] 21.Be3 Thinking of using the two bishops to cause problems for Black's rook(s) on the a-file (21.Ba2 Rybka) 21...a3 (21...Rc8! Rybka) 22.Ra1 Rybka has more faith in White's defence at this point than I do, and considers this position roughly equal. 22...Rfa8

23.Rfb1 Rather than doubling rooks on a1 and a2, which seems futile, I wanted to keep one rook active. I felt that a rook on the b file could alternate between attacking the a pawn, defending a c pawn, and possibly creating other threats such as back-rank mates. 23...Ne8 (23...d5!? is given by Rybka), but activating the knight makes sense) 24.Bb6 Ra4 25.Rb4 trying to tempt Black into undoubling my c-pawns (25.Bb3 R4a6 26.Bc4 might lead to a draw by repetition as well) 25...Nd6 (25...Rc8!? Rybka: 26.Rxa4 Bxa4 27.Bd3 Rxc3 28.Ra2=)


Over the board I couldn't see through the complications after 26.Rxa4, which Rybka prefers: 26...Bxa4 (‹26...Rxa4 27.Bb3 Ra8 28.Bc5±) 27.Bd5 Rc8=

26...R4a6 27.Bd4 simultaneously protecting the c-pawn and indirectly preventing a2 27...Nb5 (27...a2? 28.Rxa2 Rxa2 29.Bxa2 Rxa2?? 30.Rb8+ Nc8 31.Rxc8#) 28.Kf2 I was willing to give up the bishop pair to get rid of the active N and fix my pawns 28...d6 29.Bc4 Pressing the issue 29...Rb8

In several variations in the next few moves (omitted for brevity), I saw Rybka taking advantage of the fact that the knight is pinned to the b8 rook (e.g. Be3 is possible without dropping the c3 pawn). I was aware of some of the possibilities but did not appreciate at the time how valuable an asset this was. 30.Rb3 e5 31.Bxb5? I was worried that at some point the knight would be unpinned and become dangerous, but this is a premature release of the tension. White has given up the bishop pair and simplified the position. I think this should be a black win as long as an opposite-coloured bishop endgame is avoided.

For the novices: endgames with just pawns and opposite-coloured bishops (one on light squares, the other on dark) are the easiest for the defender to draw. In a nutshell, you can always defend with more force than the attacker, so for example your bishop and king can blockade passed pawns on, say, dark-coloured squares and be impervious to a light-squared enemy bishop.

31...Rxb5!-/+ 31...Bxb5 seems OK for White after either 32.Be3 or 32.c4. 32.Be3 Rba5? Rybka is adamant over the next few moves that "passed pawns must be pushed":...a2! 33.c4?! Rybka blockades the pawn with 33.Ra2! In the game I didn't like putting both rooks on the light diagonal, but Rybka is ok with it: 33...d5 and White just avoids …34.exd5?? Bxd5. 33...Ra4 (33...a2!)

34.c5? Finally getting rid of the doubled pawns, but Rybka considers this a clear error (34.Ra2!) 34...dxc5! 35.Bxc5 a2! 36.Rb2 Rc4 I think better choices include centralizing the King, advancing the f-pawn, and (my favorite) cycling the B to e6 via d7 to overprotect the pawn and commence working on another weakness. 37.Be3 Rca4

38.Bh6 At the time, I thought encouraging ...f5 and trading some kingside pawns would be advantageous. Actually, I think that gives Black further play options, e.g. creating another passer. At this point, however, all of White's options seem poor. Rybka initially liked 38.Bc5, but on further analysis seems no better: 38...Bd7 39.g4 Be6 40.Rb8+ Kg7 41.Bf8+ Kf6 42.h4 g5 43.hxg5+ (43.h5? Ra8 44.Rxa8 Rxa8 45.Bb4 Rb8 46.c3 Bc4 47.Ke3 Rd8 48.Kf2? Rd2+ 49.Kg3 Rb2) 43...Kxg5 38...f5 39.exf5 gxf5 40.Kg3?

I was thinking of penetrating with the K to grab loose pawns...if the 4th-rank barricade could be lowered. 40...Kf7?

Rybka, however, found a very interesting set of lines that gave Black a clear advantage, based on threatening to trap White's bishop with 40...f4+! I started to analyze these lines, but it gets rather complicated. However, given that White was under time pressure this complication would have been very uncomfortable.

41.c3? 41.c4! Without going into too much analysis, this provides a wrinkle in the line 41...f4+ 42.Kf2 The c4 pawn prevents ...Bd5, so Black's bishop gets in the way: 42...Bd7 (42...Ba8 43.Bg5 Rc6 44.Rbxa2 Rxa2+ 45.Rxa2 Bb7 46.Ra7) 43.Bg5 and the B can escape.

41...Kg6 (41...f4+!) 42.Bf8 e4 43.fxe4 fxe4

44.Rb4? Rybka considers this an outright error leading to a losing position. I think simplification helps Black, but at the time I wanted to get closer to an opposite-coloured bishop endgame. [44...Rxb4–+ 45.cxb4 Ra4 The move that bothered me was 45...e3! , since the king can't get in front of the pawn]. 46.Bc5 (covers both b4 and e3) 46...Kf5!–+ 47.Kf2 Ke5

48.g4 trying to get some counterplay as my clock is getting low on time. After losing several endgames because my opponent's pawns on one wing were further ahead than mine, I've come to appreciate such advances. In this game it turns out that my advanced kingside pawns save me. 48...Kd5 49.h4 Kc4 50.Ke3?! Kb3?! (50...Ra3+! Rybka. The rook on the third rank is powerful) 51.Bd4? The time crunch is taking a toll. It wasn't neccessary to give up the pawn. White could have continued with 51.g5, e.g. 51...Kb2 52.Bd4+ 51...Rxb4 g5

52...Ka3?? 53.Bc5!= White doesn't win the exchange because of his own weak rook. However, the resulting endgame should be a draw, even though White must give up the bishop to do so. 53...Kb2 54.Bxb4 Kxa1 55.Bc3+ Kb1

56.h5! Rybka had a hard time seeing that this leads to a draw. When you've used computers enough you realize that there are positions where they have difficulty "looking over the horizon" and need to be coaxed along. I had to drag Rybka through the variations kicking and screaming before it would agree. a1Q 56...Be8!? was Rybka's move, but the computer took a while to find that 57.g6! draws, e.g. 57...Bxg6 (57...hxg6?? 58.h6+-) 58.hxg6 hxg6 59.Kxe4= 57.Bxa1 Kxa1

58.g6! I was desperately short on time here. For a while I thought that maybe I should have pushed h6 first and then g6, but that doesn't work because of 58...Be8-+. 58...hxg6?? My opponent felt that 58...h6 was the move he should have made. Rybka agrees but still doesnt' see a win for Black. 59.Kd4! actually threatens to win for White by taking d5 away from the defending bishop: 59...Bd5! (59...Kb2 60.g7+-) 60.Kxd5 e3! 61.g7 e2 62.g8Q e1Q and if anything White should have the advantage in this queen endgame, but it's complicated and White would probably lose on time. 59.h6+- now White technically has a win, but only seconds on the clock. 59...g5 60.h7 g4 61.h8Q+ Ka2 62.Qc3 Ba4 63.Qd4 63...Bc2 64.Qd5+ 64...Kb2 65.Qg5 g3 66.Qxg3 Kb3 67.Qb8+ 67...Ka2 68.Qa7+ There are many forced mates in this endgame. I'll just show one mate in 4, because that's at least possible to see with about 15 seconds on the clock remaining: 68.Kd2 Bb3 69.Kc3 Ka1 70.Qxb3 e3 71.Qb2# 68...Kb1 69.Qc5 Kb2

70.Qxc2+ [70.Kd2 e3+ 71.Qxe3 Bb3 72.Qb6 Kb1 73.Qxb3+ Ka1 74.Qb4 Ka2 75.Kc1 Ka1 76.Qa3#] 70...Kxc2 71.Kxe4 ½–½ Draw forced with 3.6 seconds left on the clock.

Themes this game:

  • the center fork trick (moves 7-8)
  • using piece exchanges to weaken an opponent's pawns (moves 14-16)
  • the bishop pair (annotation to move 16)
  • "passed pawns must be pushed" (moves 21-35)
  • the pin (moves 29-31)
  • opposite-coloured bishops (annotation to move 31)
  • entrapment (moves 40-41)
  • advancing pawns as endgame insurance (move 48)
  • the skewer (move 53)