Saturday, June 27, 2009

How Good is Good?

So I drew a game against a master (with black), first time I've ever done so, and when I Fritzed it I was surprised. Not that I'd overlooked things, but rather that he'd overlooked so much, playing one or two pretty dubious moves. It's not that I assumed he'd play perfectly, since even masters clearly make mistakes, but even so the number of times he missed the best move seemed high. This guy has been a master for a long time too, with a peak rating around 2400, so he's no slouch. It got me thinking about what it means to be good at chess. By almost any standard, master is pretty strong. If you make it to master, you have a good grasp of the fundamental technical aspects of chess, and probably some areas where you possess truly exceptional understanding of the game. And yet...there's always somebody better (or something, in this computer era) to compare yourself to and come up lacking. So therefore we can't use purely comparative criterion. We also can't define chess mastery as not making mistakes because by that criteria nobody is a master. Beyond 'mistakes' vs. 'no mistakes', there are no absolutes. Mastery is an arbitrary definition we impose at an arbitrary rating of 2200.

Which brings us back to the comparative criterion. When I say we can't use it to define master, I mean that there is no formal way we can do so. In truth, mastery is whatever we say it is. John Nunn has done some comparative work that suggests that some participants in major tournaments at the turn of the last century were probably no more than 2000-2100 strength by today's standards, but in their day there were considered the leading chess masters. Considered is the wrong word...they were in fact the leading chess masters, based upon the knowledge developed at that time and the caliber of their opposition. In a way the development of knowledge in chess and the ease of disseminating that knowledge has made it harder to become a master, because it has raised the general level of chess play worldwide. The comparative criteria are stricter now for mastery than they've probably ever been.

In the end, how good you think you are is purely a function of who you compare yourself too. The better I get, the more I am bothered by small mistakes that wouldn't have worried me in the past. I find it hard to feel good about being 1800 (one point shy of 1900 now) because once you're there you recognize how much further you have to go. I think when you read GM annotations you get the sense that they feel the same way, because their self deprecation is almost always connected to errors they've made. The fact that those errors are so slight that a normal master wouldn't even rate them doesn't matter, no more than it matters to a master that I don't think his errors are gross inaccuracies.

This is a bit of a rambling post, but I really find this subject interesting. I remember one specific instance when I was talking to my buddy Garrett, who's about 2150. I was roughly 1700 at the time, and the conversation essentially consisted of him showing one of his games (which I believe he won) and repeating over and over that this move or that move was 'a blunder'. I recall thinking at the time 'wow, if that's a blunder then I wish I could blunder more often. These moves are at most small inaccuracies, 3rd or 4th best moves instead of 1st or 2nd'. Looking back, I completely understand that Garrett wasn't just being hard on himself, but was judging himself by the criteria appropriate to his level. As I keep getting higher, so do my standards. It's actually kind of a bitch, because it makes it hard to ever be satisfied with your play.

For today at least, I'm going to be satisified even though I took a draw when I could have pressed for more. It's still my first draw against a master. Here it is:


I won't be playing chess for another two weeks, so this may well be my last post for a minute. One final note: if you don't get New In Chess magazine and you can spare ~$100, then you should get it. The stories are great, and the annotations are by the top GMs themselves. I think the last issue I got had annotations by Aronian, Shirov, Leko, Carlsen, and Bacrot just to name a few. Congrats to Shirov for kicking some ass again. I would love to see him back in the elite mix. Peace out.

Monday, June 22, 2009

These Experts Ain't Shit...

I seem to have more trouble with other 'A' players than those over 2000 recently. In the last Boca Raton club tournament that just finished, I beat two experts, drew one, and the only game I lost was to another guy rated under 1900. My last round game was on the black side of a Colle against a guy I'd lost to in the same opening in the last tournament.

I really hate losing to offbeat lines like the Colle, Torre Attack, Morra Gambit, etc. It's not that they're so bad, it's that I feel a moral obligation to at least draw. I think it's because I write most of these openings off and so if I get beat by them, it feels as if all my work learning and playing main lines is wasted. I mean hell, if I could get just as good a position playing the London system and the English defense to everything, then I've been studying for nothing. But I don't think that is the case. Playing something offbeat with black can get you beat in the opening, but with white you certainly have bit more latitude. Most of these guys don't have study time, and so they just want a playable position out of the opening. While I can respect that, it's not for me. I do like to press in the opening, as I feel it's one of the strongest parts of my game.

Returning to the main point of all this, I lost to him last time from a very drawable position, and this time I was determined to at least draw. As it happened, he made a few errors (mostly subtle) and I won. The tournament raised my rating to 1899 (I couldn't have gotten one more point?!?), which is basically my goal for the end of 2010. I think I need to reassess my goals. New goal: get to 200o before starting business school. That means before fall of 2010...basically, I have a year. That's a pretty tall order if I get a job anytime soon, but otherwise I think it's probably attainable. Expert level has always been my goal in chess. If I could really do it, wow. I'd feel like I'd accomplished something. In any case, here's the game:


My opponent seems like a really cool guy, which was nice because it meant that I could talk to him about the game afterward. Post mortems are almost always valuable, even if you're quite a bit higher rated than the person you're playing. If nothing else, you end up giving a mini-lesson which helps you articulate what you were thinking at the board, always a useful way to retain knowledge.

So beyond the game, I've recently gotten the book Excelling at Combinational Play by Jacob Aagaard. I normally like Aagaard's writing and ideas but he has a horrible habit of disrespecting other authors which I find totally unnecessary and offensive. In this book he stays away from that (and actually makes a reference to not doing it, showing that he's aware of the criticism) and manages to focus on combinational play. If you've ever read a book of combinations, you know that most of them start out with some general advice on how to solve problems, the utility of doing so, etc. Aagaard is no different, and most of what he says is pretty standard for this sort of work. There was one thing however that really stood out, showing that Aagaard is still an active and practical player. He advises the reader to not just get the main line or see the idea of the combination, but to look so deeply into it that you know all the lines, such that if you were playing a game you could respond to any of your opponent's moves instantly. This is a tall order, and more work than most of do when solving problems. But think about it...if you're playing a game, you don't just want to play some sacrifice that gets refuted four ply down the line because you didn't look that deeply. You want to see to the end.

There's a connection here to the Hertan book Forcing Chess Moves which I reviewed recently. The way to know when you've analyzed enough is when you've analyzed all the forcing moves for both sides. You can't stop until you have, or else you risk a nasty surprise at the end of your combination. Any non-forcing replies are by nature non-critical, and so you don't have to analyze those. Certainly it's not always easy to even know what the forcing moves are four moves down the line, but that's a matter of board vision not the effort you put into solving. I find that if I do problems as Aagaard recommends, even four can take me an hour. When you think about it though, 15 minutes is not unreasonable for a critical tactical move during a game. I will go so far as to say that if you don't solve using this method of consistently looking through all the lines, then you're training yourself to overlook possibly critical responses during your games, like I did when I lost to Yin two weeks ago (see the archive for the game). If I'd looked at all the forcing responses he had, I wouldn't have miscombined like I did and would probably have retained the great position I gave up by miscalculating.

Returning to Aagaard, the problems in here almost all come from fairly recent games, and interestingly enough they're all from Sicilians. If you play 1.e4 or the Sicilian as black, then this book is definitely worth getting. Otherwise, it's just a good book on combinations with a few gems of advice amongst the boilerplate.

Final topic: I've changed the way I study openings. I used to spend a lot of time playing through one or two games trying to guess the moves of one side, thus forcing me to analyze the tabiyas of various openings. Not a bad method, but very slow. What I've been doing lately, and it seems to be working, is collecting a lot of annotated games (thanks !) in specific openings and play through them rather quickly, stopping to analyze only moves I don't understand. I find this gives me a good general feel for the opening, especially the differences between various branches of the opening tree. For example, I've been studying the Krasenkow (also called the Makaganov) variation of the KID, featuring an early h3. In this line sometimes you attack on the kingside, sometimes the queenside, sometimes you throw your g-pawn forward, or not, sometimes the bishop belongs on g5, other times on e4, etc. It's really hard to memorize all these variations, and it would take me a long time to learn the opening by playing through game after game in that slow, guess-the-move style. By playing through many games quickly (with pauses for deep analysis if I don't understand the point of a move), I'm starting to understand the differences between the systems and the underlying reasons for the moves the GMs play. The reason I can do this is that by looking at many games in a row I have the advantage of comparison, which is lacking when you just analyze one game very deeply. Give it a try. It's a lot easier mentally too, and would be much easier to make yourself do after a long day at work than trying to guess move after move.

Until next week. Peace out.

Tuesday, June 16, 2009

The Reign of Inconsistency

The story of every class players life, right? You play a great game, then you blunder a piece on move 12. You concentrate hard and get an overwhelming position, then one careless move and you're drawing or worse, losing. I've been told I play at 21oo strength. If you looked at some games I've lost recently, you'd conclude I played at 1400 strength. Not surprisingly, my rating splits the difference.

I don't really know how to become more consistent. Practice, I guess. I notice that mood makes a big difference for me: if I don't feel much like playing, I may as well not because I'm probably going to blunder. If I could wave a magic wand and improve anything about my game, it would be consistency. Not least because it's the only part of my game I don't know how to fix through hard work.

As you're probably guessed, I have some up and down games to show. The first is a very nice effort against an expert (the first one I've beaten) played in Boca Raton. The game is an f3 Nimzo-Indian. It's funny that with since I moved up to the premiere division at Boca I've played 3 Nimzo-Indians out of 4 white games. I have always read that the higher your rating goes the more Nimzos you play, but it's funny to find it so true. It seems that after about 1800 there's a big move away from the QGD and the King's Indian to Nimzo/Queen's Indian and Slav systems. While that's not so surprising given the greater complexity involved in playing the latter systems, it's also enjoyable to start seeing a broader range of defenses. It certainly validates all the time I put in studying them when I was ~1600. In any case, here's the game:


Not a perfect game, but I saw a lot and played with energy. I was very happy after this one. Not so much the next game. In this one my opponent also played the opening passively, but I got way too eager to punish him and over-reached. I blame studying tactics. I've spent so much time lately on tactical puzzles that I've started seeing almost every position as a tactical puzzle. While this might help me find some resources that I'd otherwise miss, it also makes it hard to step back and look at the larger picture. Sometimes you have to play with restraint. In this game I saw what looked like a good sac (it wasn't) and once I realized it wasn't good the only option I saw was to sacrifice more material (also incorrectly). When I analyzed the game, I found that a quieter continuation leading to space gains would have given me an overwhelming advantage. I didn't even look at it.

To my opponent's credit, I didn't just lose because I played badly. I set him some hard problems, and he found excellent solutions to all of them. I thought he played extremely well after the opening. Take a look:


That one really hurt. I also blundered a game in Margate last week that I'm not going to show here, because I literally lost a piece to a one move tactic around move 20 (and to a guy I don't like too, which made it even worse).

Moving to study, I'm finding it harder to study openings. They simply aren't as interesting as they used to be. I suppose that's not such a big deal since my opening repertoire is already much better than most players at my rating, but it's odd since the opening has traditionally been my favorite thing to work on. I don't have an explanation, though the pleasure I'm getting studying advanced tactics and endings may have something to do with it. Or it might just be a case of diminishing returns. I mention this because I've been toying with the notion of playing some 1.e4 again, though to do so means a lot of opening work. I'd just rather keep working my way through Dvoretsky's Endgame Manual and Forcing Chess Moves. It'll be interesting to see what helps my game more. I certainly do get some significant advantages in the opening. I hope I don't lose that during the process of improving other parts of my game.

One last note: I will be at the US Open in August. 6 day schedule. I'm super pumped. If anyone reading this is going to be there, leave me a comment and we can play some blitz. Peace out.

Thursday, June 4, 2009

Two Games and a Lesson

So I went to the Miami Chess Academy for the first time. Had a lesson with the owner, IM Blas Lugo. It was very productive. It never ceases to amaze me how very strong players can just look at a position, any position it seems, and quickly come to a conclusion about the proper plans and ideas for both sides. If you're in Miami and you want some chess lessons, I'd highly advise the Academy. It is a little pricey, but chess lessons are expensive everywhere. One funny note: Blas told me $50/hour, which is pretty normal. I got there, and we sat down in a room without a clock. I had assumed we'd go an hour, even though Blas hadn't said anything of the sort. I think I assumed it because that was how the price was quoted. Anyway, we went through some games and then did some problems, and I was thinking 'this is more than an hour, is he giving me some extra time because it's my first lesson, or what?' only when I went to the bathroom and saw that three hours had passed did I realize that the rate was in fact, as Blas had said, an hourly rate and not a rate per lesson. For an unemployed guy (me), $150 is a lot to spend on a chess lesson, but it was extremely instructive. My wife insisted, by the way. She just laughed at me when she heard the price, because she knew it would bother me more than her. In the future I'll probably just stick to game analysis since problem solving involves lots of (expensive) time in which I just sit there and analyze. It's great to do it with a strong player there, but I can't afford that. $100 every 3-4 weeks wouldn't be so bad though. I plan on going back when I have enough questions to fill 2 hours. Not every game is worth going over with an IM, after all.

Speaking of games, here are two that I played recently. Both came down to endgames, which is a new thing for me. Think about it: when you are ~1600, your games are often effectively over by 30 moves or so. Someone drops a piece or otherwise blunders, and the endgame (if there is one) is usually just mopping up. Now that I'm getting good enough where I don't drop pieces or pawns that often, I'm starting to reach endings in almost every game. I really like the endgame, but I'm not so good at it yet because I've played so few.

But I have been (as I usually do) studying my ass off with Dvoretsky's endgame manual. I think one of the things I like most about it is how hard it is. The exercises are very challenging, and it's like having a teacher with high expectations. You get a sense for how hard you really have to work to get to a high level, and depending on your penchant for study that either excites you or scares you. I get excited because as I work and study I realize that I can in fact do it, I can see deep enough and calculate accurately enough if I really put in the effort. It's tough but rewarding (that could apply to almost any activity worth doing I suppose). I think it's already payed off, as you'll see from these two games. In the first, played in Boca, I made some mistakes in the middlegame and got into a clearly worse ending. I ended up sacrificing the exchange to maintain a passed pawn on the 6th rank which distracted my opponent enough that I was able to draw. He made some mistakes in the later stages, but they weren't obvious and it just goes to show how hard the endgame is to play well.


I really should have lost that game, but I feel I defended well and I'll always take a draw in such a position. I do feel that endgames tend to favor the defender. In the middlegame the attacker often has natural squares for his pieces and a variety of tactical threats, while the defender is forced to walk a razor's edge to avoid defeat. In the ending this situation is often reversed, and the side that's better has to find 'only' moves to avoid turning a win into a draw. If you have pressure in the middlegame, it's often sustained whereas in the endgame one inaccurate move can completely change the evaluation of a position. What makes things worse is that mistakes are often not obvious at all. I must say the subtelty of endings is one of the things that attract me to them the most. The next game is not so high quality in the opening or middlegame (especially on my opponent played the middlegame well despite a pretty bad opening), but the ending is a pretty example of how strong, subtle threats can get you a win without your opponent making any obvious mistakes. I am somewhat proud of my endgame play in this one. I didn't miss a beat, and that's something I can hardly ever say about my games.


I feel that Bruce played above his rating in this game, which makes it even sweeter to get the win. I believe it was Spielmann that said "play the opening like a book, the middlegame like a wizard, and the endgame like a machine". I've got the opening part down (at least for my level), and I'm rapidly improving on the third part. I wonder where I go to learn magic..?