Monday, June 25, 2007

Prelude: Preparing an Opening Repertoire

Part of me is loathe to discuss openings yet again, but I see people tie themselves up in knots over this. I have had a monster tutorial on how to go about opening preparation rolling around my head for some time now, so I'm going to try and spit it out over the next few postings. I wanted to emphasize a few things right off the bat.

The idea is to decide for yourself what you like to play in a position, and keep track of those decisions. Unless you have a photographic memory, this means you need to maintain a record of some sort. If you're "old school", this could even be notebooks or index cards, but if you're reading this it's likely you own a computer, so I'll focus on using one to manage your repertoire.

Programs such as Bookup or the free download Chess Position Trainer allow you to move back and forth through the variations, and will spot transpositions. For example, let's say after you enter 1.e4 e6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 you write a note about the Rubinstein variation of the French. If you then enter 3.Nd2 (the Tarrasch variation of the french) dxe4 4.Nxe4 you'll see your previous note and realize that you just transposed out of the Tarrasch and into the Rubinstein. There are other fancy things you can do with such opening databases, like automatically import your internet games for analysis if you want. Bookup has recently been revamped as the Chess Opening Wizard. I'm not ready to plop down money on it when I have the free CPT plus ChessBase 9 at my disposal (as well as the old Bookup Express).

Alternatively you can use software like Fritz, Chessbase (which has a new, free CB Light version) or whatever. I'm sure there's even free software out there that will do the job as well. With these, you would have to save your repertoire as separate lines and variations, so you have to spot transpositions between openings yourself. There's a tutorial here on one method of constructing a repertoire database in this fashion.

The ideal repertoire, in my opinion, consists largely of moves that follow good opening principles, that you consider the best in the position (or that you have come to understand are the best). Theory can help steer you towards more reliable lines, but what's most important is that you like and understand the moves you make. If you do this, memorization is less of an issue. Also, if you're consistently playing the moves you've decided yourself that you like, then there is no need to waffle and hop from opening to opening. If you really think 1.e4 makes more sense than 1.Nf3, then why play the latter? If you think 1...d5 in response to 1.d4 makes more sense than 1...Nf6, then why not play it?.

As I've mentioned in my previous opening guidelines, I would strongly encourage you to choose systems that tend to follow classic opening principles: play just enough pawn moves to get your pieces out, develop your minor pieces (knights before bishops, unless you know what you're doing), castle and connect your rooks. Do so in a manner that gives you a say in what happens in the center. If the "book" opening move violates any of these principles, understand why or play something else. (btw: if you still don't have a good feel for what good opening play is, Emms' Discovering Chess Openings is a terrific book).

Where opening theory is helpful is in guiding you towards good choices and away from bad ones. This way, you don't paint yourself into a corner. If you look at the ground and follow the most attractive path, you may find yourself looking over a cliff. For example, perhaps you may like the Ponziani opening, but theory pooh-poohs it. If you can understand, and agree with, that assessment, you may find that you agree more with the Scotch, Italian or Spanish than the Ponz. On the other hand, if the Ponz move sequence makes sense to you, and you can't understand why they're dissing your opening, then play it. Play your favorite move until you realize and understand that it's inferior. The goal here is to consistently play "your" moves instead of hopping from opening to opening.

Here's a better example of what I mean. I am in the early stages of figuring out what I want to play against the Sicilian. I currently lean towards systems with Bc4 (although I also like the feel of systems based on Be2, Be3 and f4). Here is a standard Sicilian position that arose in an ICC game:

If you crack open a theory book, 7.Bb3 is the overwhelming favorite here, with barely a mention of 7.0-0 except to say that it usually transposes after 7...b5 8.Bb3. So, if I were a lemming theory monkey (hmm... maybe I should avoid metaphors with multiple species) I would move the bishop twice in the opening, play Bb3, and perhaps not even try to find out why the move is played. I would be relying on my memory at this point. In contrast, 7. 0-0 is completely playable, and is the natural developing move. It's not quite that simple...there's actually a lot of interesting material surrounding these two move options that I'm sweeping under the rug, and certain black responses can be met with a rude shock. The point is: 7.0-0 is good, it's natural, the move order subtlety is overlooked in a lot of books and by playing "my move" I have a possibility of surprising my opponent if they play bookish moves robotically. I may have to play Bb3 next move anyway, but I've looked the lines over and decided I'll move it when I have to and no sooner.

I very much like the approach to opening preparation mentioned in Johnsen and Johannessen's book on a Spanish repertoire for black. Study the main line of the opening system you've chosen and try to get a mastery of that line. "Your attitude should be something like this: I may not be a grandmaster in chess, but on this particular branch of the game, I want to be an authority!" For the club player, I'd temper that as "I understand this entire branch up to the end of the opening (pieces out, castled, rooks connected...move 8-12ish) and have an idea about how the middlegame goes from there". As you encounter deviations in your games, determine what you'd play the next time you encountered it and follow the main line from there. One nice thing about Chessbase is that the database statistics can help you find out what the "main" line is.

In the short term, playing simple systems allows you to concentrate on other things such as tactics. However, if you approach opening study the way I recommend, you won't spend much more time studying a main line opening than your pet system. Basically, after you play a game (even internet blitz), look at the opening, see where it leaves theory, and determine what you would like to play the next time you encounter it. Also, make sure you're familiar with the main line itself. Make a record of this line. There, you're done! This automatically keeps your opening study down to the appropriate level. If you play the Spanish as White, for example, you may find that the "minor" 3rd-move deviations such as the Steinitz, Classical and Cozio occur more frequently than the main-line closed Spanish. If you routinely do an opening checkup after each game, you'll consequently be studying the lines that occur in your games the most.

In the long run playing main lines (in my opinion) will be the best for developing your chess. I have read several masters that have said "I would be a better player if I had started playing main lines (the Spanish is a common one that comes up in this context) at a younger age." If your intent is to progress in chess as far as you can, I say choose a repertoire that will grow with you. It's often said that chess masters can play any position well. I think choosing a repertoire that exposes you to many different kinds of pawn structures, while initially intimidating, ultimately helps you to become a better player. I think the more pawn structures out of Kmoch that you see in your own games, the better.


Blue Devil Knight said...

This seems quite sane, and I need to remind myself of this stuff when I get my dickie in a knot over subtle opening stuff.

Grandpatzer said...

One thing that's really becoming obvious to me is how limiting books are in coming up with a repertoire. There's a lot of interesting lines out there that aren't encountered much because they're not fashionable, and because opening book authors aren't dealing with them.

Reading through Marin's 1...e5 repertoire book, I'm coming across a lot of lines for white that Black has difficulty meeting. It's fascinating that with a well-written and objective book on an opening, you can reverse-engineer it and come up with your own nasties.

I'm also interested in old books and games where they state quite authoritatively that some oddball treatment of an opening is best. I'm relying on memory here, but I think Nimzo said it was best to decline the Evans Gambit, and Alekhine insisted that after 1.e4 d5 2.Nf3 e6 the best move was 3.Be2. Today at Dortmund there occured a Nimzovich Sicilian and it was good enough for a draw.

Opening principles and main lines are useful guides. But, if you're willing to follow your nose and do the research you could create your own killer repertoire. I'll touch on this in my repertoire construction series.

takchess said...

in reference to a past post: Are you aware there is a pgn file of Polgars book that can be downloaded ?

Grandpatzer said...

takchess: ooh, where?

wang said...

Great post! I have something similar to this I wanted to post on my blog, but its the opposite not to develop your opening repertoire. I have come to the same conclusions you have, it just took me 10 years of beating my head against the wall to get here. I probably won't be quite as eloquent as you are.

I do advocate using books a bit more than you, but buyer beware! You can't just slavishly follow a book or books, thinking that it is going to be the gospel for you.

Thanks for the insight.

Grandpatzer said...

Wang: I'm a book junkie myself (you can probably tell from reading the blog) but I try and imagine the plight of players that can't afford all the bells and whistles. I try to assume the minimum of a computer and a chess program, since they're using a computer to read my blog.

I'm finding that Chessbase is a godsend here, however, particularly for looking at the lines you encounter less frequently. Between its statistics, Fritz, and your own brain, you can figure out reasonable lines yourself. Books help provide insight and context, but for "bang for your dollar" Chessbase easily beats 6 books on openings you don't frequently encounter.

Grandpatzer said...

Whad'ya know, google polgar and .pgn and hey presto.

takchess left this as an exercise to be solved by the reader.

frank said...

For all those interested in the Polgar PGN book, I have placed a copy of it on my website. Just click "Books" when you get to the Menu page. The address is: