25 September 2011

Openings Selection: Initial Considerations

Having recently started studying a new opening, I think it's worth spending some time looking at what factors go into selecting and then learning openings.  I discuss the evolution of an openings repertoire for experienced players in a companion post, which I think is a rather different animal than the process faced in choosing one's initial opening weapons, so it deserves its own discussion.  Here I'd like to focus on the first steps faced by all of us as players, partly to provide a background on my own choices and also to help set the scene for how this affects opening system choices later on in our chess careers.

Although it's somewhat arbitrary, I'll define the initial process of openings selection as lasting until a player settles on the major system(s) as White for the first move, along with Black defenses to 1. e4 and 1. d4, that they will have in their permanent career repertoire.  The above definition I think is generally a practical one, since it covers the essential decisions and challenges that every tournament-level player must face on a regular basis in the opening phase of the game.  For me, this initial process lasted around a year.

When first learning chess in a systematic way, I tried out a number of different openings in informal games.  In general, this is a great way to gain practical experience with the different systems' ideas and key positions, while also developing a gut feeling about which openings best suit you as a player.  Now that good chess software is easily found, it is easy to test drive different openings with a willing silicon opponent, although I think the feedback you can get from a human opponent is very worthwhile, as long as you take their own biases into account.  With the internet, it's also easy to find basic surveys of the various openings, to know what to try out.  However, I think there's still something to be said for owning a comprehensive reference such as Modern Chess Openings (MCO) in order to be able to know what is out there and see the basic descriptions for all of the different openings, along with suggested variations.

As White, I initially opened with 1. e4 and after some experimentation chose suitable variations against the major Black defenses, playing the Exchange Ruy Lopez and Closed Sicilian in tournament play.  Before that I had informally tried out the main Ruy Lopez lines and the Open Sicilian, but the complexity was at that point beyond my capabilities and inclinations.  I recall being somewhat puzzled by the 1. d4 suite of openings and was never really attracted to them as White.  The Reti (1. Nf3) was neat to try, but at the time I didn't really understand what White was supposed to do in it.  (So that you have a frame of reference for my chess ability at the time, my first rating obtained in tournament play was in the low Class C range.)

As Black, I looked at a range of possibilities, especially for defenses to 1. e4.  Informally, I played the Sicilian Najdorf, Alekhine, Closed Ruy Lopez, and French a great deal.  After some time, however, I gravitated towards the Caro-Kann, after having a serious go at the French.  It seemed to be similarly solid but without the drawbacks of the shut-in light-squared bishop and occasionally bare kingside.  As far as defenses to 1. d4 went, although I dabbled some with the King's Indian, I settled on the Slav as another solid choice.  Although the b7-c6-d5 pawn chain is a shared feature with the Caro-Kann, the positions arising from the two openings really aren't very similar, so the pairing doesn't save you much study time, unlike what you can achieve by playing the Modern (1..g6) or other very flexible openings that can be used against both 1. e4 and 1. d4.  As a small bonus, however, the openings selection did mean that I never had to fear the Blackmar-Diemer Gambit (1. d4 d5 2. e4) since I could just transpose to a Caro-Kann with 2..c6.

While my openings selection as Black held up to the test of tournament play during my first year, I wasn't nearly as satisfied with my results with 1. e4.  Essentially, I had discovered as Black that I played best and was most comfortable in semi-open positions.  While the variations I had chosen as White also largely reflected this, I felt that I did not grasp their ideas at a fundamental level, plus my opponents were all obviously well-prepared (or at least better prepared) to meet 1. e4 as the most common opening choice.  At this point I discovered IM Nigel Povah's How to Play the English Opening - a 119-page book published in 1983 - and decided to try it out.  The opening concepts were explained in detail (or as much as a short book could contain), complete games were provided and the positions were new and interesting to me.  I chose variations that I liked and after a few months of preparation permanently switched to 1. c4 as White.

There's a good deal of free advice out there on what openings are "best" to play, but I think it really boils down to whatever works for you as a player.  (I have a similar view of approaches to chess training in general.)  There is both a practical and an aesthetic dimension to this.  The first one is easy to understand - if you regularly end up in losing or inferior positions out of the opening against players of comparable strength, then your openings are not doing the job of keeping you alive on the chessboard.  Regarding aesthetics, if you don't have a genuine liking of the opening positions, or find your openings dull or incomprehensible, regardless of your results with them, they probably won't be a good permanent match for you. 

Taking that a step further, I think it is ultimately preferable to buckle down and master an opening that you are genuinely attracted to, although you may be having sub-par results with it, rather than try to convince yourself that a boring opening with good practical results is in reality fun to play.  (People's definitions of boring will vary widely; I think my opening repertoire is fascinating, I'm sure others would fall asleep looking at it.) 

Finally, let's not forget that last part about fun, because unless you're a chess professional, fun has to be a component of the chess experience, otherwise it's just work.  Chess improvement by definition requires work, but it should be the kind that leaves you with a smile at the end of it.

24 September 2011

Annotated Game #11: Upset (or half of one)

This game was the next tournament game I played after Annotated Game #10: Upset, taking place during the first round of a weekend tournament the following month.  In this case, the rating difference (over 500 points) was even higher, with me being a Class C and my opponent an Expert, the first I had ever played.  I had some momentum still going after my previous win, even a month later, and I was able to achieve a superior position, although settled for a draw.  One of the reasons I chose to examine this game, as was the case with Annotated Game #3, was to take another look at the final position and see whether taking a draw was a mistake.  At the time, I thought I objectively should have been able to win the game, but doubted my ability to bring the full point home.

My opponent chose the Exchange variation of the Caro-Kann, which all the opening books say is harmless for Black.  This assessment jibes with my own limited experience; I've played two tournament games in the variation, winning my first one against an opponent rated almost 200 points higher and then drawing this game.  Basically, it causes Black no real problems and there is nothing very complex about it, so while it gives White a decent game, opportunities to pressure Black are much reduced compared to other variations.  The variation was touted for a short while in the Fischer era and then essentially disappeared from Grandmaster-level play.  I think there's still a bit of mystique at the club level as a result of Fischer's recommendation.  It's certainly a safe choice against the Caro-Kann, especially if someone doesn't expect to face it much and doesn't want to spend a lot of time studying a more complex variation that has better prospects of yielding an advantage.

In the game, my opponent was even less ambitious than is normally the case in the variation, playing 7. Ne2 instead of the critical 7. Qb3 line.  By move 10, Black has fully equalized and then pursues a strategy of trading down, successfully removing from the board two pairs of minor pieces in short order.  White suddenly decides not to accept his drawish fate and launches a kingside attack with 16. f4, seeking to create something out of nothing.  He misses that Black's queen can sally forth to grab the f-pawn, the result of a neat in-between move, and then safely move out of danger.  Black then concentrates on shutting down any possible White attack on the kingside, playing remarkably accurately; I chalk that up to having a good day and also having a simple enough position where I was confident in its soundness.

On move 35, just as I was activating my kingside pawn majority, my opponent offered a draw.  After some thought, I accepted it, although I partially regretted it at the time.  Looking at it now, however, it wasn't a bad choice.  The double rook and knight endgame would have been very difficult to win and Houdini gives an evaluation in the final position of Black having less than a 4/10 pawn advantage, despite being a full pawn ahead.

This and the previous game taught me not to fear Class A and Expert level players, which for a Class C player is a very valuable lesson.  I've subsequently always played much better when focusing on honing my own performance in a game, rather than worrying about my opponent's skill level, although I admit I haven't always succeeded at that in the past.  In any case, this was another watershed game for me.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Expert"] [Black "ChessAdmin"] [Result "1/2-1/2"] [ECO "B13"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "70"] [EventDate "1987.??.??"] {B13: Caro-Kann: Exchange Variation and Panov-Botvinnik Attack} 1. e4 c6 2. d4 d5 3. exd5 cxd5 4. Bd3 Nc6 5. c3 Nf6 6. Bf4 Bg4 7. Ne2 {Most common by far is Qb3, with Nf3 a distant second; Ne2 is rare.} e6 8. Qb3 Qd7 9. O-O Be7 10. Ng3 O-O {First move out of the database. The two DB games continued ..Nh5, a clever idea which shows up next move; Houdini agrees with me and prefers castling first.} 11. Nd2 Nh5 12. Nxh5 Bxh5 {In addition to having exchanged off the pair of minor pieces, Black now is ready to play Bg6 when necessary.} 13. Rae1 Bd6 ({Interesting to see how Fritz preferred} 13... Rac8 14. Re3 { but Houdini agrees with the human move made.}) 14. Bxd6 Qxd6 15. Qc2 Bg6 { this is a common maneuver in the Exchange variation, taking away threats on the diagonal.} 16. f4 {Hoping to generate an attack from an equal position. The engines recommend that White instead exchange on g6; among other things, Black's bishop currently ties down the Qc2.} Ne7 {Both Fritz and Houdini recommend exchanging bishops on d3 first, which would be a cleaner defensive strategy. I recall thinking that White looked rather menacing in the resulting position, however.} 17. Nf3 {probably the idea behind f4, but White must have overlooked move 18 for Black.} Qxf4 $15 18. Ne5 Bxd3 19. Qxd3 (19. Rxf4 $4 { Fritz says greedy!} Bxc2 20. Rf2 Bg6 $19) 19... Qg5 20. Rf3 Nf5 {the best defensive idea} 21. Ref1 (21. g4 {doesn't work, even without Black taking advantage of the pin. For example} Nd6 22. Rh3 Ne4 23. Rh5 Qe7 24. Qh3 Ng5) 21... Rad8 {taking d7 away from the Ne5} 22. Qe2 a6 23. a4 Qe7 24. Kh1 { too slow, Black begins to consolidate his position} Nd6 25. Rh3 h6 26. Ng4 Ne4 27. Rf4 f5 (27... Rfe8 28. Qe3 Ng5 29. Rg3 $17 {is preferred by Fritz, but again Houdini likes the original human move, after a bit of calculating.}) 28. Ne5 Qg5 {here Houdini much prefers Qe8, which would allow Black to start generating queenside counterplay by pushing the b-pawn and moving the Rd8 over to the b-file. I of course had all my attention focused on the kingside.} 29. Rf1 Qd2 {forcing the queen exchange kills any remaining hope by White of counterplay} 30. Qxd2 Nxd2 31. Rd1 Ne4 32. Kg1 Ng5 33. Rhd3 Ne4 34. Ng6 Rf6 35. Ne5 g5 {At this point my opponent offered a draw.} 1/2-1/2

18 September 2011

Chess Tactics Server

I've used the Chess Tactics Server enough now to make a few observations.

  • Problems are all "real life" in the sense of looking like normal game positions. The device of having your "opponent" (the server) make its move before you helps underline this impression.
  • The wide range of different problem types also assists in providing an over-the-board type challenge to the player. They range from simple recaptures to material gain to checkmate, so you need to actively figure out what the position requires, rather than programming your brain for a single type of solution.
  • Statistics and records are useful when revisiting failed problems.
  • The timed rating feature encourages quick moves without checking multiple lines, in a very Pavlovian fashion.
  • The rating change indicator for each problem appears to be incorrect, as a ratings penalty always seems to be assessed if the solution is not found in the first few seconds.
  • Alternate solutions are not recognized. This affects only a small minority of problems, but is still occasionally annoying.
Some common advice is to do at least 15 minutes of tactics problems like this a day, in order to keep the mind sharp and expose you repeatedly to key tactical concepts and patterns.  I think that's about right for me, any more than that in one session and it tends to become more like rote exercises, rather than my intention of training for serious play.  It's also important for me to keep a tournament game mentality to the extent possible, rather than simply blitzing moves out without checking for blunders or alternatives.  As long as I don't really care about the CTS rating, then the server actually works pretty well for this.

17 September 2011

Annotated Game #10: Upset

This is my greatest tournament game upset.  At the time, I was a Class C player and my opponent was a strong Class A player rated 400 points above me.  He was also the tournament organizer, which perhaps added some insult to injury for him.  In the tournament itself, a four-round weekend Swiss, I had dropped the first two games as Black but won my third as White, thereby drawing a last-round White pairing.  This tournament marked the beginning of a sustained rise in my over-the-board performance, which lasted until I went to university and stopped playing competitively.

No doubt in part due to the ratings gap, I recall that my strategy was to simplify down and keep a drawish position.  My playing style then (as it more or less is today) was to aim for a solid game and then let my opponent make the first mistake.  The English opening is well suited to this as White, since there are usually no targets for Black to attack early on.

My opponent surprisingly chose the Symmetrical variation, which is by nature more drawish.  Perhaps he assumed (rightly) that I had less experience with it, but the ideas are not hard to grasp and if White is not looking to push for an edge (for example as could have been done by playing 7. d4) then Black cannot easily complicate matters.

In the game, I succeeded in exchanging off a pair of minor pieces early on, although could have more profitably focused on additional development.  In my assessment Black was overly aggressive with his h5-h4 push, as White could just as easily switch his own forces to the h-file after the g-pawn is exchanged; I'm sure the ratings difference helped dictate my opponent's choice of strategy.  Black set a trap on move 14 by exchanging his g7 bishop for my c3 knight and leaving the e7 pawn unguarded, which however I avoided.  I was positionally better but should then have refrained from executing my own bishop-for-knight swap, which would have allowed Black an attacking advantage had he recaptured with the queen on the long diagonal.  Black then attempted to force the issue on the h-file, leading to a successful counterattack by White.

My attitudes have shifted to the point where I no longer wish to take rating differences (either way) into consideration during play.  I find that I enjoy the games more that way, feel much less pressure, and usually play better chess.  However, I will say that an upset win like this never loses its satisfaction.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "ChessAdmin"] [Black "Class A"] [Result "1-0"] [ECO "A38"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "53"] [EventDate "1987.??.??"] {A38: Symmetrical English vs ...g6:4 Bg2 Bg7 5 Nf3 Nf6} 1. c4 Nf6 2. Nc3 g6 3. Nf3 Bg7 4. g3 c5 5. Bg2 d6 6. O-O O-O 7. d3 Nc6 8. Bd2 Bg4 {Keeping the symmetry with ..Bd7 is more common and scores better for Black.} 9. Ng5 { Now out of the reference database. Standard developing moves are Rb1 and Qc1.} Qc8 10. Nge4 Nxe4 11. Bxe4 {The original idea of the knight maneuver, to exchange off a pair of minor pieces.} h5 {A bit overoptimistic on Black's part. } 12. Rc1 h4 13. Bg5 hxg3 14. hxg3 Bxc3 {Neither Fritz nor Houdini like this. Houdini prefers the attacking plan of f6 followed by Kf7 and Rh8.} 15. bxc3 { The bishop for knight exchange slightly favors white, who now has taken the d4 square away from Black and has the two bishops.} Kg7 $11 16. Bxc6 {I may not have seen the issue with taking the e7 pawn at this point, but was also looking to trade down further.} (16. Rb1 Rh8 $11 {would have been fine.}) 16... bxc6 {Missing the opportunity to seize the long diagonal and immediately increase the pressure on White's king position.} (16... Qxc6 17. f3 Bh3 18. Rf2 $15 f6 19. Be3 Rh8) 17. f3 $11 ({I didn't fall for} 17. Bxe7 Re8 18. f3 Rxe7 19. fxg4 Qxg4 $19) 17... Bh3 18. Rf2 {Houdini prefers Re1, leaving the f2 square open for other pieces in future variations.} f6 19. Be3 Qe6 20. Qd2 Rh8 21. g4 ({A more solid defense is} 21. Rh2 g5 $11 22. Kf2) 21... Rh4 (21... g5 { would keep Black more in control and lock away White's bishop. Houdini gives} 22. Bxg5 fxg5 23. Qxg5+ Kf7 24. Rh2 {as best for White, otherwise the piece is useless.}) 22. Rh2 $11 g5 {a move later, this causes Black problems due to the Rh4.} (22... Rah8 {is equal.} 23. Kf2 Kf7 $11) 23. Bxg5 $14 Rxg4+ $2 (23... fxg5 {would lead to a difficult but playable endgame featuring a small White advantage.} 24. Qxg5+ Kf7 25. Kf2 {guards e3 against the queen fork} Rah8 26. Kg3 R4h6 27. Qf4+ Kg7 28. Rch1 Bf1 29. Rxh6 Rxh6 30. Rxf1 Qxe2 31. Rf2) 24. fxg4 $18 Qxg4+ 25. Kh1 fxg5 26. Rg1 Qe6 27. Qxg5+ (27. Qxg5+ Kf8 28. Qg7+ Ke8 29. Qh8+ {wins the rook.}) 1-0

12 September 2011

Simple Openings Repertoire Database System

In response to a comment from laramonet, here is a description of what could in nice terms be described as a simple and straightforward system designed to store an openings repertoire in a database.  (In not so nice terms, I suppose it could be labeled crude and unsophisticated).  I started my personal openings database quite a while ago, so while I'm sure I could redo the whole thing now in a more sleek and powerful manner, it's not really worth the time investment, since it works just fine for my needs.

The desired objective is to keep a single database (I simply call it "ChessAdmin opening book") containing all of the variations that I consider significant and noteworthy in the openings I use.  My basic criteria for inclusion is that there should be at least a decent chance of facing the line in a tournament game someday, with the corollary that the variation itself is complex or important enough that I need to document and (theoretically) remember it.  For example, I don't have anything in the database for Black defenses against 1. b4 (Sokolsky), 1. g4 (Grob's Attack), etc.  However, the database is always a living document and I regularly update and refine it based on my ongoing practice and study.  For example, I've started to recently encounter a Queen's Gambit Declined setup more often against the English, so have added that to the openings book database.  If in the future I encountered a Grob's player on a regular basis, I'm sure I'd add something on that as well.

More specifically, I will have a "game" in the database for each major variation in an opening.  For example, I currently have nine Caro-Kann games in the database, named as follows:
  • Caro-Kann - 2. c4
  • Caro-Kann - 2. d3
  • Caro-Kann - Advance
  • Caro-Kann - Exchange
  • Caro-Kann - Fantasy
  • Caro-Kann - Main Line - Bronstein-Larsen
  • Caro-Kann - Main Line - Classical
  • Caro-Kann - Panov
  • Caro-Kann - Two Knights
In this case, in the game data "White" is "Caro-Kann" and then "Black" is however I choose to name the particular variation. This allows for easy sorting of the lines by title in the database.  For each of the games, I manually entered the main line of each variation and then selected sub-variations as I desired.  Naturally, some "games" are much more complex than others (like the Caro-Kann Classical), which others are simpler (like the 2. c4 variation) being shorter and with fewer sub-variations.  For my side of the opening in question, I treat it as a repertoire, only giving one preferred line against the opponent in the vast majority of cases.  The idea is not to have a comprehensive openings database, but rather my personal repertoire which I should study in order to prepare for a tournament or other serious play.

I prefer to have the lines named and organized in the above way in the database, as it parallels how openings book authors typically organize chapters.  I can then easily look up a particular major variation in the games list of the database.  Conceptually and practically for me it is cleaner than trying to put everything into just a few or even one massive database game.  Where one draws the line is probably a matter of personal taste, although using a "chapter" analogy is probably a good guide.  For example, I decided to have two Main Line "game" variations (also happening to parallel the framework in Starting Out: The Caro-Kann) largely because they are so different from each other and the large number of branching variations made putting the two in a single database game unwieldy.

One thing that I think is very helpful about a database structure is being able to easily include variations and sub-variations that aren't main line or are never used at the professional level, but can be dangerous or annoying if you aren't aware of them.  Your future opponent likely doesn't know or doesn't care if GMs play the line or not, so if something looks obvious they may just go ahead and play it.  There are a fair number of these types of moves that may even be refuted by a particular sequence, which makes it worth documenting rather than just trusting myself that I can work it out over-the-board.  I will also use game annotations to leave myself notes, for example when my side has a common plan in response to whatever the opponent does, rather than including all of the possible choices.

With my opening book organized per above, it's easy to immediately link from a particular position to the main database in order to look up full games and then load those separately, perform statistical operations based on the position, etc.  With the above setup, I'd hesitate to merge complete games into the base "game" variation, but I've considered setting up a different "examples" database to break out those key games that are worth referring to repeatedly.

11 September 2011

Book completed - Starting Out: The Dutch Defence

Starting Out: The Dutch Defence by [McDonald, Neil]

Today I finished Starting Out: The Dutch Defence by Neil McDonald (Everyman Chess, 2004).   I went through the book with a tournament-size chess set in front of me, playing through nearly all of the lines and illustrative games.  This type of approach is what I base my opening study methods on; I also happen to find it more entertaining when complete games are presented.  Particular lines that I am interested in are entered afterwards in my personal openings database.  As with the Caro-Kann book in the Starting Out series that I also own, this book presents a general survey of all major variations and approaches in the opening, with a balanced approach from the author.  My impression was that it was a somewhat less comprehensive treatment of the opening than can be found in the Caro-Kann book; in quantitative terms it certainly is, clocking in at 20 pages shorter.  This impression may also have been influenced, however, by the fact that I had considerable experience with the Caro-Kann before going through the other book, while I have never played the Dutch (yet!)

A more detailed description of the book can be found at the above link; here are my own observations:

  • Formatting: the single-column format used by the Starting Out series works well due to the large amount of text relative to variations presented, although occasionally the diagrams (always paired for space reasons) appear in slightly awkward places.  The diagrams are large and can be especially helpful for the reader when referring to the book without a chess set, although they do not appear frequently enough to allow a read-through of the book in this manner (unlike for example Logical Chess: Move by Move).
  • Typos: I found several move typos, concentrated toward the first half of the book.  Sometimes these were incorrectly numbered moves after explanatory text, although on occasion typographical errors appeared in the moves themselves.  While the reader can overcome this without much effort, it was annoying to see the editing lapses.
  • Stonewall Dutch: this was the most interesting section for me, as a number of new ideas (well, new to me) were introduced; it was also the first of the three main Dutch setups presented.  The Stonewall is not a very popular opening variation, but I find it fascinating and it is interesting to see how many world-class players have used it over the years (from Alekhine to Kramnik).  While all of the games contained in the book were well worth replaying, I found it a little strange that the theory narrative focused on the Modern Stonewall setup with ..Bd6, while a number of the games presented featured ..Be7 instead, with no real discussion of the differences.  That said, there are a large number of ways to reach a Stonewall-type setup and there is relatively less theory out there on it in general.
  • Classical Dutch: while I probably had the least amount of personal interest in this section beforehand, it proved to be nearly as fascinating as the other two major setups.  GM Kevin Spraggett's games in particular stand out (as they did in the Stonewall section).
  • Leningrad Dutch: I was the most familiar with this form of the opening, having a while back purchased Steffen Pedersen's The Dutch for the Attacking Player.  I found the Starting Out presentation of the Leningrad setup (characterized by g6/Bg7 and ..d6 in the center) to be quite useful, especially regarding the major choice faced by Black in the main line on the 7th move (..Qe8, ..c6 and ..Nc6 being the options).  This book uniquely looks at the ..Nc6 option seriously, rather than dismissing it like most other works.
  • Anti-Dutch: while not necessarily exhaustive, the ideas presented in the book treat the wide range of possibilities quite well and provide the Black player (my perspective) with good, easily understandable ways to meet the various White gambits and alternative setups.
When an author has experience with an opening and can present it in an interesting and engaging manner, as was done here, the reader can get a great deal out of it.  As someone new to the Dutch in general, I also found it quite useful when he made points about move-orders, transpositional possibilities between the various systems, and specific strategic and tactical concepts that show up repeatedly.  This book accomplished exactly what it was aimed at doing, which is to provide a broad survey of the opening that at the same time has enough depth to be useful as a guide to practical play as well as future, more detailed study.

10 September 2011

Annotated Game #9: Caro-Kann Classical with 5.Nc5

This tournament game features an intriguing sideline of the Caro-Kann Classical variation with 5. Nc5.  On the face of it, it looks aggressive, but if Black keeps his cool and White plays only standard-type moves, Black ends up in a favorable version of his usual setup, as occurred in this game.

White can test Black in the variation I used with 5..Qb6 if he plays 6. g4 as a follow up, similar to the idea that occurs in the Caro-Kann Advance after Black develops his bishop to f5.  In large part due to a prominent loss by Beliavsky to Bronstein in 1975, the Qb6 variation isn't a popular response to 5. Nc5.  However, Kasparov (yes, that one) and Shakarov in their 1984 book Classical Caro-Kann: 4...Bf5 recommended it; more recently, Houdini slightly prefers Black as well in the 6. g4 line.

In the below game, White plays the more normal-looking 6. Nf3 and Black then forces the c5 knight back as part of his usual developing sequence.  By move 12, Black has a comfortable game with opposite-side castling, which gives him the easily understood strategic goal of advancing his pawns and pressuring White on the queenside, which White cannot ignore with his king there.

The middlegame instructively points out how my positional play was weak as I exchanged down (and offered to once again) an effective attacking piece for a less effective equivalent.  After going a bit astray with the attack, Black spots a deflection tactic on move 26 and emerges with a pawn and a won endgame.  The endgame itself is also instructive, as White could most likely have held with a more active defense, piece activity being the key in these types of positions (in this case R+N vs. R+N).  White's king was also shut out of the action on the queenside once the theater of war shifted back to the kingside.

Since this game, due to my studies I believe my attacking play has improved (although it still has a long way to go) as well as my understanding of the role of piece exchanges in the middlegame.  Among other things, this game reinforces the lesson that simplification on the board can lead to a dissipation of an advantage or even a loss.  I was also pleased at being able to see the key tactical motif on move 26, along with the necessary in-between move.  This was relatively simple, but also illustrative of the role that experience with the opening and resulting middlegame played, since I was comfortable and familiar with Black's play in that position.

[Event "?"] [Site "?"] [Date "????.??.??"] [Round "?"] [White "Class B"] [Black "ChessAdmin"] [Result "0-1"] [ECO "B18"] [Annotator "ChessAdmin/Fritz/Houdini"] [PlyCount "120"] [EventDate "2006.??.??"] {B18: Classical Caro-Kann: 4...Bf5 sidelines} 1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Bf5 5. Nc5 {an aggressive-looking sideline} Qb6 6. Nf3 (6. g4 Bg6 7. f4 e6 8. Qe2 Be7 9. h4 h5 10. f5 exf5 11. g5 Nd7 12. Nb3 Qc7 13. Nh3 O-O-O {is the critical line from Bronstein-Beliavsky.}) 6... e6 7. Nb3 Nd7 8. Be3 ({Standard is} 8. Bd3 Bxd3 9. Qxd3 Ngf6 10. O-O Be7) 8... Qc7 9. Qd2 $146 {First move out of the DB. Only two other games are shown in it, one with Bd3 and the other with Nh4.} Ngf6 10. Bd3 {as you may have guessed by now, a standard idea in the Classical Caro-Kann. Black's light-square bishop is too strong to leave there and there is little else for White's bishop to do.} Bxd3 11. Qxd3 Be7 12. O-O-O O-O {opposite-side castling favors Black slightly here, as White does not have as many kingside threats as in the main line.} 13. h3 h6 {although this takes g5 away from White, better would be to get the queenside ball rolling with ..a5 and not give White a possible future target at h6.} 14. g3 Nd5 15. a3 {takes away the b4 square from the attack} Nxe3 {Example of a beginner's exchange, taking simply because you can. The Nd5 is much superior to the Be3 and no benefits accrue to Black from this.} 16. Qxe3 Nf6 {moves the knight away from the attacking zone. Better would be to develop a rook with Rfd8 or push the a- or b-pawns. Black's potential attack has lost momentum.} 17. c4 {Luckily White decides to loosen his kingside and offer a target for Black, giving him back his attack.} b5 18. Qd3 {this allows Black to open lines on the queenside.} ({Better defense would be} 18. c5 a5) 18... bxc4 19. Qxc4 Rab8 20. Rhe1 (20. Nc5 {would remove the knight from the line of fire and let it go to d3 to help shore up the b-file}) 20... Qb7 {bad attacking form, putting the queen in front} ({More effective is} 20... Rb5 {followed by Rfb8}) 21. Rd3 Rfd8 {with the idea of transferring it to b5 via d5. This is too slow, however.} 22. Re2 Rd5 23. Nc5 {disrupts Black's plan, I recall being surprised at this.} ({Houdini suggests} 23. Ne5 {which would be more active for White, for example} Rc8 24. Rc2) 23... Qb5 {again with the beginner's exchanges; Black's queen is better on the attack and is also helping defend the queenside weaknesses.} ({Houdini and Fritz both prefer} 23... Bxc5 24. dxc5 {which leaves Black with a positional plus, at least, if not an attack.} Rbd8) 24. Rc3 {however, White doesn't seize the opportunity to exchange queens and misses a tactical deflection theme after the captures occur on c5, allowing Black to win a pawn.} Bxc5 25. dxc5 Rxc5 $1 26. Qxb5 (26. Qxc5 Qxe2) 26... Rxc3+ { an important in-between move.} ({Black needed to be careful here and not do the obvious recapture} 26... Rcxb5 $6 27. Nd4 Rh5 28. Nxc6 $11) 27. bxc3 cxb5 { After the smoke clears, Black has a winning endgame, with 4 vs. 3 pawns on the kingside and connected vs. isolated pawns on the queenside.} 28. Rb2 Rc8 29. Rc2 Nd5 {another beginner's move, placing a piece where it is obviously less effective at controlling territory.} (29... Ne4 30. Kb2 $19) 30. Kb2 $17 Kf8 31. Nd4 a6 32. Kb3 Nb6 33. Kb4 Nc4 34. a4 e5 35. Nb3 (35. Nf5 bxa4 36. Kxa4 e4 $17) 35... Ke7 36. axb5 axb5 {the pawn is tactically defended, due to the threat of Na3+} 37. Ra2 Nd6 {I had thought this necessary to protect the pawn.} ({However, the engines show this:} 37... Ke6 38. Kxb5 Nd6+ 39. Kb4 Rb8+ 40. Ka4 Ra8+ 41. Na5 Nc4 42. Kb5 Kd5) 38. Ra7+ {looks useful on the surface, but all it does is help Black activate his king. The white rook can't do any damage to Black's kingside pawns.} (38. Nc5 $5 {would make it much more difficult for Black.}) 38... Ke6 $17 39. Nd2 g6 {preparing to get the pawns out of the way of the White rook and push in the center.} 40. f3 f5 41. Rg7 Kf6 42. Rd7 Ne8 { far too passive.} ({Much better is} 42... Rc6 $17 {which holds everything together and keeps White's king cut off.}) 43. Nb1 {returning the passive favor, instead of Rd3.} e4 44. Na3 Ke6 ({More forcing would have been} 44... exf3 45. Rd3 f2 46. Rf3 Nc7 47. Rxf2 Nd5+ 48. Kxb5 Rxc3) 45. Rd2 $17 Nc7 46. Nxb5 $2 {the endgame simplification error, to which I've been subject to myself.} Nxb5 $19 47. Kxb5 Rxc3 {Black now has a completely won game, with White having no counterplay with either rook or king.} 48. fxe4 fxe4 49. Rg2 Kd5 50. Rd2+ Rd3 51. Rg2 e3 52. g4 Ke4 ({An easier and quicker approach would have been} 52... Rd2 53. Rg1 e2 54. Re1 Ke4) 53. Kc4 Rd2 54. Rg1 Kf3 ({The engines bemoan the fact that I didn't play} 54... e2 55. Re1 Ke3) 55. Kc3 Kf2 56. Rb1 Rd6 57. Rb7 e2 {making it harder for myself} (57... Re6 $5 {and Fritz says Black can already relax} 58. Rb1 $19 e2) 58. Rf7+ Kg3 59. Re7 Kxh3 { rather than trying to queen the pawn, which seemed too hard, I just simplified down.} 60. Rxe2 Kxg4 {White's rook can't stop the king and two connected passed pawns, with his king out of the action.} 0-1

How to Publish with Aquarium 2011

As a follow-up to the post on chess software and by popular demand, here's how to use Aquarium 2011 in order to publish a game on a blog.  An example of the results can be seen here.  Warning: not a user-friendly process.

1.  Update Aquarium.  Get the latest Aquarium 2011 update from ChessOK.  The original forum post describing this is here.  Once installed, your web publishing options will now be correctly shown and enabled.

2.  Set up your blog style sheets.  Once you've loaded a game in Aquarium (any game will do), on the Publish tab select the Web Export - iBook HTML for Blog option.  (It doesn't matter if you actually want to publish in the iBook format or not.)  The program will bring up a new window with style sheet code at the top and the iBook code at the bottom.  Copy the style sheet code to your clipboard (there's a link on the new window to do that for you).  This then needs to be added to your blog's template html.  As an example, if you're using the latest Blogger interface, you can access this via the Design - Edit HTML option.  I suggest pasting the code near the top, just after the line where you see your title section, which should begin with < title >.  Then click Save Template.  This enables your blog for Aquarium publishing.

3.  Publish a game in a blog post (with playable board and game score with annotations).  On the Publish tab select the Web Export - HTML for Blog option.  This brings up a new window with the HTML code, copy this to your clipboard.  In a new blog post, when you are ready to paste the game in, make sure you are looking at the HTML rather than normal text (Blogger has "Compose" and "HTML" buttons, choose the latter).  Paste in the HTML code where you want the board and attached game score/annotations to appear.  To avoid having a lot of white space above your chessboard, under Post settings make sure the option is chosen to have line breaks use the < br > tag.

There's also a ChessCafe article that covers most (but not all) of the above.

Observations:  practically speaking, it's better to have any other text you want in the post above the published game.  Also, if you go back in to edit your post after publication, the game board may not function afterwards and you will then need to delete and re-paste the game HTML code.  Finally, I've noticed that no other blog posts below an Aquarium game will show up on my main blog page.  This isn't a big deal for me, since I prefer only having 1-2 posts on the main blog.  Not sure if other users have been affected by this issue, but I spent a lot of time Googling in order to track down what was necessary to get the publishing process to work properly and don't recall seeing another mention of it.

05 September 2011

September 2011 Chess Carnival

This month's edition of the Chess Carnival is now up at the Hebden Bridge Chess Club site.  Well worth checking out, among other things the host has an excellent post on his experience at the British Chess Championships.

04 September 2011

Comparison of Chess Software - features and uses

Since by now I have a fair amount of experience with the various chess software packages that I've been using as part of the training process, I thought I'd put down some observations on their strengths and weaknesses, along with some thoughts at the end on how they are being used together.  This is not intended to be a comprehensive review of each package's features, but rather some informed commentary by a player using them to improve his chess.

Fritz 12
  • Links to ChessBase reference database for best openings reference comments when doing game analysis
  • More flexible complete game analysis feature; select evaluation threshold for variations/comments, select game move from which to start analysis
  • Sometimes silly, pointless, or poorly translated "natural language" game commentary
  • Decent variation of playing options (sparring, "friend" mode) with handicaps
  • Don't fully trust engine evaluations with material imbalance
ChessBase 10
  • Allows UCI engine plugins (i.e. Houdini) for position analysis
  • Quick searches for game references based on board position
  • Sorts games for various tactical and strategic themes
  • Game statistics and more sophisticated DB analysis easily generated
  • Easy to use for annotation purposes
  • Expensive, not worth upgrading to ChessBase 11 at this point
Rybka Aquarium 2011
  • Excellent database included, but openings reference in game analysis is based on move-order rather than position
  • Complete game analysis produces better/more understandable results and includes numerical position evaluation in commentary
  • Publishing of fully playable annotated game to blog or webpage is possible (although not user-friendly)
  • Analysis and commentary features are strongest; focus is on this rather than playing
  • Rybka 4 engine included has legal issues; can substitute another one if desired
Chessmaster: Grandmaster Edition
  • Best playing options by far for simulated opponents and 3D boards
  • Comprehensive training package included
  • Haven't bothered with the analysis and database functions, which are not state-of-the-art
So a typical cycle is to play a training game, afterwards conducting a complete game analysis using Rybka 4 in Aquarium and/or Fritz 12.  The game (in PGN format) is then loaded into ChessBase for opening reference purposes and commented on in Aquarium, while I step through the game move-by-move for my own analysis.

I'm still working out the best practical way to do this for new training games.  One technique that seemed to work reasonably well for Annotated Game #8 was to have ChessBase loaded with the Fritz 12-analyzed game and then have Aquarium open on the left 3/4 of the screen so I can easily compare the different variations and evaluations.  Although I like the ChessBase + Houdini engine combination for conducting annotations using my tournament games database, which was all previously analyzed using Fritz, the Aquarium analysis and commentary features seem more powerful and cleaner overall, now that I've become more used to them.

Unfortunately, there doesn't seem to be one package that can do it all, so I expect I'll continue to mix-and-match to some degree.  If I had to pick just one to own, however, it actually (perhaps surprisingly) would be Chessmaster: Grandmaster Edition, because its playing and training features aren't replicated by the others and there are now a variety of good freeware programs available for database (SCID, ChessBase Light) and analysis (Houdini 1.5, Stockfish) purposes.

Annotated Game #8: Training game (English with early ..d6)

This training game was played against the "Max" Chessmaster personality (a high Class C on my system).  "Max" varied from the main lines immediately with 1..d6, which is however a good and flexible move.  I chose to head for a standard position featuring a fianchettoed light-squared bishop and queenside expansion plans.  The benefit of the move-order chosen by Black, however, was that it allowed him to exchange off the bishop early.  This was not particularly upsetting to White, although my openings database showed an improved line starting with 6. Rb1 that keeps the bishop on the board.

White had several major decision points in this game which had critical effects on the game.   The first was on move 10, where I chose to go for a somewhat more unbalanced central position with the idea of being able to then undermine Black on the e-file.  Black's decision to castle queenside on move 11 greatly simplified things strategically for me, as my planned pawn pushes would now threaten his king position and not just a gain in space.  Black struck back in the center, however, demonstrating that was the correct reaction to a flank attack.  Better defending on move 14 would then have given Black an equal game.

I again decided to pursue the more active, attacking path on move 15, passing up winning a pawn in favor of deploying the queen and increasing the pressure on Black's king position.  Black passed up shutting out White's dark-squared bishop with ..d4 and paid the price, as this time I calculated the correct follow-up on move 18 and attacked down the half-open b-file.  Despite some missed mating ideas for White, Black was nevertheless on the ropes and then allowed a nice tactical finish to any hopes of resistance.

The principal learning points of this game for me are:
  • Understanding the ..Be6 and ..Qd7 ideas stemming from an early ..d6 and ..e5 opening sequence in the English
  • The exchange of the fianchettoed Bishop can either be avoided or improved, per the game notes
  • The center should be watched more carefully when planning an early queenside expansion
  • I should perform additional calculation when the position is relatively unclear, for example moves 12-15
  • Calculating a forced variation to a winning position is fine, even if better moves are missed along the way
The analysis process itself was also rather interesting.  I again had both Rybka and Fritz perform a complete game analysis and then compared the results while I stepped through it myself.  There was general agreement on the key points, although some interesting differences in the move recommendations and timing of things such as Black's potential counterplay with an ..h5 push.  Some other observations on the strengths and weaknesses of the computer tools will be made in a later post.

8 8
7 7
6 6
5 5
4 4
3 3
2 2
1 1
ChessAdmin - Max (CM Class C)
1-0, 9/4/2011.
[#] 1.c4 d6 +0.22 2.Nc3 e5 3.g3 +0.11 Nc6 +0.18 4.Bg2 +0.15 Be6 the point behind an early ..d6 5.d3 Qd7 6.Nf3
[Main line is 6.Rb1 g6 7.b4 Bg7 8.b5 Nd8 at which point a4 and Nd5 are both popular and rated equally by Rybka; Nd5 has fewer games but scores much higher in practice]
6...Bh3 +0.48 7.O-O +0.26
[Another possibility is 7.Bxh3 Qxh3 8.Nd5 Qd7 9.O-O ]
7...Bxg2 8.Kxg2 Nge7 9.Rb1 +0.15 First move out of the database. Interestingly, Aquarium's hard-coded openings reference gives this move a +0.44 evaluation, but the Rybka engine gave a lesser result and preferred Nd5. 9...Nd4?! +0.56 This loses time for Black.
[9...g6!? 10.Qb3 Rb8 11.Qa4 Bg7 12.b4 +0.15]
10.Re1 +0.33 first major decision point. I had also considered the alternative e3, which Rybka prefers slightly and in hindsight is probably preferable.
[10.e3 Nxf3 11.Qxf3 c5 12.Bd2 ]
10...Nxf3 11.exf3 With the idea of undermining the e-pawn, the reason Re1 was played. 11...O-O-O +0.41 Strategically makes White's game much simpler, as queenside expansion now also leads to an attack on the king position. 12.Be3 +0.29 Kb8 13.b4 +0.15
[13.d4 is preferred by both Rybka and Fritz, although they disagree on Black's reaction. This would among other things rule out the threatened pawn fork on d4, which I did not see at this point.]
13...d5 +0.29
[13...h5 is suggested by Fritz, starting counterplay on the kingside. Rybka and Fritz both prefer this idea, either now or later in the game (for example move 16) as more active for Black.]
14.Nb5 -0.15 exchanging on d5 would simply bring Black's pieces to life, so the pawn fork needs to be avoided.
[Rybka's preferred method is 14.Bc5 d4 15.Qa4 Qxa4 16.Nxa4 Ng6 17.h4 +0.29]
14...Nc6?! +0.72
[14...d4!? 15.Bd2 f6 16.f4 a6 17.Na3 Ng6 -0.15]
15.Qa4 -0.07 The second major decision point for White. I chose this in order to maintain pressure, instead of the alternative, which wins a pawn.
[15.cxd5!? Nd4 16.Nxd4 exd4 17.Bxd4 ]
15...a6?! +0.56
[15...d4!? is again a superior defensive move.]
16.Na3 Be7? +2.38 17.b5 axb5 +3.75 18.Rxb5 Third major decision for White and the one requiring the most calculation. I saw the bishop sacrifice idea on b6 and considered the Rxb7 idea as well, although did not see that it would lead to mate. 18...Bxa3?
[White still has the attack after 18...d4 19.Reb1 b6 20.c5 Bxc5 21.Rxc5 ]
19.Qxa3 This misses a mating opportunity:
[19.Rxb7+ Kxb7 20.Qb5+ Kc8 21.Qa6+ Kb8 22.Rb1+ Bb4 23.Rxb4+ Nxb4 24.Qa7+ Kc8 25.Qa8# ]
19...dxc4 20.Reb1 the mate idea with Rxb7 was still present 20...b6 21.Bxb6 cxb6 22.Rxb6+ Kc8 23.Qa8+
[The engines find a remarkable mating sequence starting with 23.Qc5 Qc7 24.Rxc6 Rd7 25.Qa7 ]
23...Kc7 24.Rb7+ Kd6 25.Qa3+ Ke6 26.Rxd7 Rxd7 27.dxc4 I had calculated the attacking sequence (in multiple stages) as far as this and considered it winning, based on both the material balance and the two passed pawns. 27...Rhd8 28.c5 f5 29.Qa6 Rc7 30.Rb6 Kd5
[After 30...Rdc8 I had planned to simply run the a-pawn up the board, with some Queen activity thrown in, which would have led to further material losses for Black. The move played allows a nice tactical solution.]
31.Rxc6 Rxc6 32.Qd3+ Ke6 33.Qxd8 Resistance is now futile. 33...Rxc5 34.a4 h6 35.a5 Rd5 36.Qb6+ Rd6 37.Qb7 Rd7 38.Qc8 g5 39.a6 Ke7 40.Qxd7+ Kxd7 41.a7 Kd6 42.a8=Q Ke7 43.Qc6 h5 44.Qg6 g4 45.fxg4 fxg4 46.Qxh5 Kd6 47.Qxg4 Kc5 48.Qe4 Kd6 49.h4 Ke6 50.h5 Kd6 51.h6 Ke6 52.h7 Kd6 53.h8=Q Kc7 54.Qhxe5+ Kb6 55.Qg6+ Kb7 56.Qe7+ Kc8 57.Qg8# [1-0]