Wednesday, May 07, 2008

Advice for chess beginners

This is a summary of a simple improvement plan for chess novices. It is meant for those trying to get from beginner level to the fledgling club player in about a year. It is based on my experience trying many ways to improve, when I went from 950 to about 1450 in slow games at the Internet Chess Club (ICC). As points of reference, an absolute beginner who just learned the rules would be rated about 500 at ICC, while a grandmaster would be rated over 2600.

The following suggestions are roughly in order of importance.

1. Play and do postmortems
Much like learning to sing or swim well, chess skills are built up largely from extensive practice. Hence, as a beginner, it is most crucial to simply play chess. Some people attempt to improve at chess by focusing on everything but playing: they do puzzles, read lots of books, or (worst of all) spend lots of time studying openings. As I discuss below, some of these things are important, but if you are truly a novice you need to develop some intuitions about the mechanics of the game by playing as much as possible.

Slow games (e.g., 30 minutes or more per side in a game) are the best bet if you want to develop your chess, but don't worry if you enjoy the occasional fast game too.

Unfortunately, it isn't enough to merely play. Just as there are horrible singers that have vast experience singing along with the radio, there are plenty of players rated under 1000 that have played thousands of games of chess. It is crucial, in addition to playing, that you get feedback on your performance.

How do you get feedback? By doing postmortems of your games. A postmortem is when you critically examine a game soon after you finish it. Look for good moves you made as well as blunders that you want to avoid in the future. Explain to yourself why a move was good or bad, and take a mental note of the important features of the position: that way, in similar contexts you will be more likely to remember the lesson learned (e.g., his Bishop was far from my Queen so I didn't notice it could take her).

Do postmortems for all of your games, but especially your slower games. Some folks may tell you to only go over your losses: I disagree strongly with this sentiment for reasons I spell out here and a follow-up post here.

After going over the game on your own, find someone or something better than you to go over it one more time. The chess program Fritz is a wonderful tool--it plays as well as a Grandmaster, never gets tired, and loves to show you places in your game where you could have done better (and importantly, where you played well!). Even better than Fritz is a human coach or advisor. While Fritz just shows you the moves you should have made (and then you need to figure out why), a person will be able to explain it to you, answer your questions, and generally give more usable and practical feedback.

If you are playing and analyzing two or more slow games a week, in addition to the suggestions below, that is great. It can also be helpful to perform postmortems on your faster games, as such analysis will reveal practical tactics, the best types of tactics with which to be familiar.

Summing up this first recommendation: it is absolutely crucial that you play/postmortem in addition to working on the other things I mention below.

2. Read a general overview of chess
Before getting focused on specific areas of the game, start by reading a general overview of chess. This will give you a useful bird's-eye view of the game and provide a foundation for the sometimes strange terminology used by most chess players.

My favorite beginner-level book is Wolff's Idiot's Guide to Chess. If you have pride about buying one of the books in the 'Idiot's Guide' series, swallow it. This is a great book. He describes how to play in the opening, middle, and end of the game, basic tactics (see below), strategy, and even how the pieces move. Each chapter includes many problems that are explained in the back of the book. Work on those problems, as chess is a game best learned by doing, not reading. This is a book I wish I had when I first started playing.

3. Study tactics
I define tactics as sequences of moves that lead to a gain of material or mate (there is some debate about this definition, but as a beginner you don't need to worry about it). For instance, you have doubtless overlooked a mate-in-one or left your queen open for the plucking, losing a game in ignoble fashion. This happens to everyone starting out, and such blunders are the most decisive events in all games of the novice.

When you are playing, you should be focusing almost all of your energy on not committing tactical blunders, and exploiting those of your opponent. Really, that is the key to reaching 1400 at ICC. You will read a lot about pawn structure and other chess 'strategy', but until you stop leaving pieces open for the taking and missing simple mates, this stuff will just not make a big difference.

Aside from playing and doing postmortems, how can you improve at tactics? There are lots of resources for learning the basic mating patterns, such as the wonderful little book Simple Checkmates.

There are also many resources out there for learning tactics beyond simple mates. I'd suggest developing your tactical muscle in two steps. First, read an explanation-heavy book that will familiarize you with the basic themes, a book such as Littlewood's wonderful Chess Tactics. Another nice book with less explanation that contains extremely simple tactics is Pandolfini's Beginning Chess.

Once you have studied an explanation-heavy book on basic mates and tactics, I recommend using software with lots of tactical problems to build experience and intuition. Chess Tactics for Beginners and Personal Chess Trainer (aka Chessimo) are both great programs. Do ten-or-more problems every day--you will learn faster if you spend two hours spread out over a week than if you spend two hours one day a week on tactics.

You will learn tactics fastest if, in addition to trying to figure out the problem, once you determine the solution you construct an explanation of each move in the problem. E.g., 'I moved the queen to the back rank because the enemy King is hemmed in by his pawns, so he can't escape mate.' There is a study that demonstrates that such narratives improve people's performance when a similar position is seen again.

4. Study the endgame
As you improve at tactics, more of your battles will reach the endgame. You will be playing better players and the games will no longer be decided by move 20. So it is time to learn something about how to finish the game.

Of course, if you already know the basic mates, you already have a bunch of endgame knowledge without knowing it. In addition to basic mates, it is important to learn about pawn play in the endgame (the fact that pawns can be promoted to Queens make their importance skyrocket in the endgames), and many other topics that will help you tremendously.

Luckily, Silman wrote a wonderful endgame book that will tell you all you need to know about the endgame, Silman's Complete Endgame Course. Once you master the first few chapters of this book, and have stopped missing elementary tactics in your games, you will easily reach 1400 at ICC.

5. Don't worry much about the opening
You might be tempted to memorize a bunch of specific openings. This is relatively easy to do, and is tempting because the opening occurs in every game. Unfortunately opening study just won't help you very much. You are losing your games because of tactics, not subtle opening errors. If you are losing in the opening, it is because of opening tactics that you should be able to target and remedy in your postmortems.

A few simple principles are all you need in the opening, and those can be found in Wolff's book (Step 2 above). Sure, there are interesting openings that violate the principles, but you are guaranteed to make it safely to the middlegame if you just follow the principles while keeping your eyes open for tactically-justified exceptions.

6. Develop good thinking habits
When playing, you will be tempted to make the first move that pops into your head. Don't. Relax, sit on your hands, and take your time to think through the moves, consider your opponent's replies to your moves, and how you will respond. This will be hard at first, but it is crucial. Dan Heisman has written a lot about this, such as his classic article Real Chess. Once you have read Wolff's book, you might also take a look at my extensive discussion of thought process in chess, found in PDF form here. As you might expect, it advises that on every move you should think about tactical threats before you do anything else.

7. Have fun, relax, don't try to force it
Being smarter than the average Joe is only marginally important for chess improvement. For even the smartest person, getting better at the game takes a great deal of time, energy, and commitment. You will lose a lot of games. This is very important to do, as losses are wonderful learning opportunities.

So unless you are the second coming of Bobby Fischer, you will not get better as fast as you want to. Learning chess is not like learning World History, getting better is not a matter of memorizing facts and reading books. It is a much slower process, in which we control the amount and quality of training, but the improvement tends to come at its own pace, just as a well-tended sprout will follow its own time-course as it grows into a flourishing rose bush.

So, relax, don't let yourself get too frustrated in your chess improvement efforts. You can't force yourself to improve, but you can study, and I guarantee the improvement will come. Chess is a beautiful, complex, often exhilarating game, and if you follow the simple advice given here you will come out with a greater appreciation of the game in addition to being able to crush your Uncle at Thanksgiving.

Once you approach 1500 (this should take between six months and three years, depending on how much time you put into chess study and other factors such as your age), aspects of the game such as strategy and the openings will merit a bit more attention. That is, simple tactics and endgames will decide fewer of your games. Your course of study will need to become more individualized and expanded to include things such as the study of annotated master games. At any rate, at this level I recommend finding a good coach or advisor, and generally talking to people much better than me about how to take your game to the next level. Good luck!


Thanks everyone for reading my blog: I'll be taking a break from blogging for a while. With the help of writing here, and the excellent comments, I easily surpassed my goal when I started of reaching 1200 at ICC, and I'm no longer focused on chess in a serious way. Please look at my Blog Highlights for what I consider the most helpful or interesting material on the blog from the last few years, most of it geared toward novices struggling to improve at this complicated and fun game.

Tuesday, May 06, 2008

Another de la Maza hater surfaces

Jon Jacobs, a noob who probably thinks he is being original posts some Circles hate, and I respond here. (See Blogotype #4). Ah, the memories. Every six months or so one of these turds folks floats to the surface.

Of course I love smart discussion of chess improvement, especially criticisms of the Circles method of tactical training, but blustering simplistic bombast posing as sage instruction is annoying.

Hey, Liquid Egg Product, I bet you'll be surprised to find out you are a member of the Knights Errant.

Friday, May 02, 2008

Who should do the Circles, and how?

My post on pattern recognition as seed planting was the first of two major topics I've ruminated on since finishing the Circles. The second topic is this--the one serious problem with the Circles isn't that they don't help (they do, to varying degrees), but that they tend to lead to burnout or boredom.

[For those who don't know what the Circles are, see the description here.]

Clearly the way Michael de la Maza (MDLM) did them, with 1200 very tough problems in CT-Art, doing all of the problems with no repetitions until that Circle was finished, is one of the best ways to burn out (for those who don't know, first Circle 10 minutes spent on each problem, second Circle 5 minutes, and eventually Circle seven when you do all 1200 in one day, 30 seconds each in a marathon session).

So we have two questions. Who should do the Circles, and what is the best way to modify them to avoid burnout? I don't claim to have the answers, but look at this as a brainstorming session to which others can contribute.

Who should do the Circles?
The idea behind the Circles is that gaining tactical acumen is largely a matter of learning patterns. Tactics are often hard to 'figure out' over the board. Such calculation is time and energy consuming, so we want to learn to recognize simple tactics quickly, effortlessly, as easily as we would recognize a friend's face. Most people agree that a key to gaining such skills (as opposed to book knowledge--in which case you would be a chess scholar not a great player) is repetition. It's like training for tennis, you practice the overhand over and over again, doing it with a coach and on your own until you just get it right without trying

So, who should do the Circles? If you are a tactical car wreck, overlooking simple tactics (where 'simple' depends on your skill level), and that is clearly the weakest part of your game, and if you already read and understood one or two of the text-heavy tactical books out there, then the Circles may help you get better at tactics.

That, anyway, is a necessary condition for doing the Circles. What else should you consider? First and most important, do the Circles sound fun to you, or like a hellish chore? To me, the Circles sounded quite fun, and I was just so horrible at tactics that something extreme seemed necessary, so I was quite excited to do the Circles. If they don't sound fun before you do them, then perhaps you should try to find a different approach to remedy your tactical flat tire.

Also, do you have the time to devote to them? They require at least a half-hour a day devoted to studying tactics. Do you have this time? Will you have enough time to actually play games?

Finally, the Circles are anything but a balanced approach to chess. Will it bother you if your tactical muscles grow at the expense of some of your other chess skills?

Frankly, I would hesitate to recommend the circles to a true beginner. A beginner should probably spend a year or so working on more general issues, of course spending a great deal of time on tactics, but also taking the time to study the endgame, and especially the majority of time actually playing and analyzing games. Perhaps once you hit 1000-1100 at ICC, then you might consider doing the Circles, as it will pretty handily take you to 1400 or higher levels at ICC (if my experience is any indicator).

How should the Circles be modified to prevent burnout?
This has been discussed by many Knights, as I catalogued a couple of years ago here.

MDLM, who was helped tremendously by the Circles, focused exclusively on chess for over a year. He wasn't working at the time. Doing the full-blown MDLM style circles is very very hard to do, so modifications are almost mandated if you want to maintain sanity and balance (though don't let me discourage you if you want to be like the Knights Errant of old and be a total bad ass).

So, how to modify them to prevent burnout? Again, I don't have the answer, but here are three suggestions:
1. Mini-circles.
Do circles on small sets of problems. E.g., one hundred at a time, learning them extremely well over a course of a few weeks rather than months. I did minicircles, and I did get a little burnt out but because I probably went a bit overboard near the end and I didn't take any breaks (see suggestion 3).

2. Use simple problems.
CT-Art has some very hard problems, many of which have mistakes in the solutions. For those reasons, I used Chess Tactics for Beginners. This was great for me, as it hammered home the basic mates, but by the end had 3-5 move combinations which helped me appreciate how basic tactics interact over the board. Personal Chess Trainer also seems to have many simple problems (for a comparison of CT-Art and PCT, see this post).

The whole point of the Circles is to help you improve at chess. So pick a set of problems that start out a bit too easy for you, but progress to problems that you would tend to miss in over the board situations. I believe it is a mistake to start out with problems more complex than you are missing in real games. Start simple, and build up to the more complex problems that are presently a bit out of your reach in real games.

3. Take regular breaks
MDLM was hard core. He says that you have to do the circles every day to improve. My guess is this is just wrong. Taking one day a week off is probably a good idea. Even better, between sets of mini-circles, take a break from the circles for a few days. Play some chess. Relax.

After the break, you can come back to previous problem sets to refresh your memory, to make sure you are still sharp.

Those are my main ideas. Again, I don't claim to have any final answers, and consider this a brainstorming session.

Thursday, May 01, 2008

100 Chess Book Reviews Portal