Learning Opening Lines

Lots of things to say about this...  Here's half-a-dozen or so little nuggets to ponder, and a bit more practical advice.

"Of my fifty-seven years I have applied at least thirty to forgetting most of what I have learned or read.  Since then, I have acquired a certain ease and cheer which I should never again like to be without.  (...)  I have stored little in my memory, but I can apply that little, and it is of use in many and varied emergencies.  I keep it in order, but resist every attempt to increase its dead weight." -- Emanuel Lasker

1. Is opening preparation mostly wasted in practical play?  A four-board match played away at Exmouth one Saturday...

The top boards each made a mistake on move 5: 1.c4 e5 2.g3 Nc6 3.Bg2 f5 4.Nc3 Nf6 now 5.e3(?) was possibly inaccurate, allowing 5...d5!? (highly recommended in Kosten#8217;s book, although NCO suggests White can still extract an advantage here) but Black didn't play it, preferring 5...Be7.  Did the players know better?  I had the idea they didn't know it at all, and were making it up as they went along.

Board 2, the players scampered along in the French Tarrasch: 1.e4 e6 2.d4 d5 3.Nd2 Nf6 4.e5 Nfd7 5.Bd3 c5 6.c3 Nc6 7.Ne2 cxd4 8.cxd4 Qb6 9.Nf3 f6 10.exf6 Nxf6 11.0-0 Bd6  This is the tabiya, where the real first choice lies for White, who followed the old main line: 12.Nc3 0-0 13.Be3 Bd7 14.Re1 Be8.  Now here all the books give this position as equal; although it is unbalanced I don't know why White wants to learn and play this line, Black has so much counterplay...  [GMs at the time were playing 12.Bf4.] We than had 15.Ng5 (a very important move) and then 15...Bh5? which was the first deviation from theory... White can now get an instant plus with 16.Bxh7+! e.g. 16...Nxh7 17.Qxh5 Nxg5 18.Qxg5 Nxd4? 19.Nxd5! but it was not refuted by White's chosen 16.Qb1?!  So White played into Black's hands, Black didn't know it and  made a mistake, and White couldn't take advantage of it...

Board 3 was a bit less theoretical, although it has been played before: 1.Nf3 g6 2.e4 c5 3.Bc4 Bg7 4.0-0 e6 5.c3 Ne7 6.d4 cxd4 7.cxd4 d5 8.exd5 after which Black surely should not have played 8...exd5(?) but 8...Nxd5, when there is some sort of plan against the IQP instead of a 'normal' White small plus with no prospects for Black.

And Board 4 had the unlikely start 1.e4 c5 2.Nf3 d6 3.h3... Not a GM move, but an attempt to play the opponent, and not the GM author of the opponent's latest book purchase.

So, opening preparation can be a complete waste of time!

2. We are often told: "Learn ideas not variations".  This is a good starting point: there's nothing sadder than watching a player trot out a move from a book in the wrong position, or even a move that leads to the right position, but the player doesn't understand what to do in that position when the moves run out. [Common examples include: playing c2-c3 in the Colle [1.d4 2.Nf3 3.e3] when you aren't forced to by Black's ...c7-c5; playing the Giuoco Pianissimo in every game but not having a clue what to do after 1.e4 e5 2.Nf3 Nc6 3.Bc4 Bc5 4.Nc3 Nf6 5.d3 d6 6.O-O O-O 7.Be3 Bb6, and then playing a3 or h3 or even b3; playing the Fried Liver sacrifice 1.e4 e5 2.Nf3 Nc6 3.Bc4 Nf6 4.Ng5 d5 6.exd5 Nxd5?! 6.Nxf7!? rather than the superior 6.d4!]

I got more out of reading Fine's " Ideas behind the chess openings" than reading any number of more detailed books; it's terribly out of date now, but to anyone taking up the Queen's Gambit or French Defence, I'd certainly recommend starting with Fine's discussions there.  Tracing the history of ideas-led opening books, I can remember enjoying How to Play the Caro-Kann (Keene et al., RHM), where literally every move is discussed at some length, and How to Play the Sicilian Defence (Levy O'Connell, Batsford), which emphasised systems over variations.  Many modern opening books for club players use ways of explaining I associate with Bellin and Ponzetto in their 'Mastering the...' books, showing themes and piece placements related to structures as well as to variations.

3. "It may sound like a platitude, but openings have to be studied in accord with your own tastes. Another point is a little less obvious: in constructing an opening repertoire, you need to take your own powers of memory into account(...) I know from my own experience what an excruciating labour it is to memorise 'theory' before a game.  You have it all written down in notebooks, you have gone through it ten times before starting play, and you still can't remember it.  If this is so, it may well be better to concentrate on what I call 'opening schemes' -- logical systems with a smaller amount of theory, in which it is more important to understand the position and know about typical ideas and resources than to memorise specific details and precise move-orders." -- Mark DVORETSKY

At a minimum, you should have some key games in your memory: the most simple wins for each side, the central ideas, the most common tactical patterns, and, vitally, any traps that are around in your opening.  Dvoretsky says, keep a card index, but these days I guess you would use a computer.  Although modern computers have a huge capacity, treat them like Lasker treats his memory: keep any databases small and well-ordered.  Just having the games saved on your computer won't transfer them magically to your brain; you need to go through them and see what is typical and untypical of each game, and what reason if any might explain the differences.

4. "Some players learn a lot of variations by heart and repeat them in play whenever possible.  But their opponents may by so  ignorant or unkind that they step 'out of the book' long before that wonderful advantage-for-me sign has been reached.  I remember a young player who said he lost three years of his life studying the Najdorf!  He realised that he had learned variations, not chess."

"On the other hand, it is perhaps overly simple to say that you must learn the ideas behind the openings ... But the trouble with chess is the opponent: if you know only the 'ideas behind the variations' and he knows the ideas and a lot of variations, he is likely to beat you." -- Bent LARSEN

5. "As Victor Korchnoi points out in his book of best games, when you study openings you don't remember the variations. What happens is that the strategic motifs become familiar to you at a deeper level - which means getting the kind of positions in which you know what to do." (...) "...It's really better not to try to memorize anything - the only stuff that will stick are things that can be hung on hooks of deep understanding. If you have good abilities in calculation and vision, much can be worked out during the game; what is 'theory' anyway other than a collection of imperfect games which have, for the most part, been analyzed rather badly." ... " You're best forearmed by understanding the middlegame structures that come from your openings." -- Nigel DAVIES

I think the other significant 'hooks' to hang things on are your own games.  Make a card index or a database of all your own games, find out where you or your opponent diverged from what your book says, decide if it was a mistake or an improvement or just as good.  You should find that each time you play a line, the 'tree' of moves that you can recall grows.

6. "What do most readers look for in an opening book? Unfortunately, something that they won't get. TWIC readers, for example, apparently want pretty much what my own students keep asking for: a book which explains all the 'ideas' of an opening, but isn't cluttered up with all sorts of nasty variations which one will never run into over the board anyway. I'm sorry to report that this is just a fantasy. Learning an opening by accumulating abstract ideas is a little like learning a language by reading a grammar book. Worse, actually, because generalizations in chess don't apply with nearly the consistency or predictability of grammatical rules. If a chess opening could be learned by absorbing the opening's 'ideas' (whatever those might be), the opening phase of the game would be universally mastered and of little interest. The fact is, no verbal description of what squares are important or where the pieces 'usually' go can describe the dynamic interplay of tactics and positional factors in any major opening. Whereas, by contrast, the straightforward study of enough examples will lead to a nuanced and practical knowledge of how to play that opening. In addition, by studying in context, you will automatically get a much better grasp on how those important squares and typical manoeuvres work than you would have from reading a general description." (...)

"To summarize this lengthy review, then:

(a) openings can only be learned by study of numerous examples and variations, not by learning abstract 'ideas';

(b) books can use either the tree structure or the illustrative-game approach successfully, but the second choice often leads to the omission of key material;

(c) watch out for books which fail to attribute analysis--there's a good chance that the material is copied from elsewhere;

(d) before you purchase a book, if you get the chance, compare a few variations in it (preferably ones you play) with what you have in your database. If a simple collection of database games gives you more information than what's in the book, try to assess whether other qualities of the book (line selection, instruction, guidance) outweigh its lack of original information. Barring that approach, your best bet is to go by the company's and author's reputation, and hope for the best!"

-- John WATSON

Well, Watson is an IM, and is perhaps telling us what we might need if we want to be an IM who plays 'major' openings.  Club players who don't know much about any openings, major or minor, can't acquire a lot of opening knowledge at once.  I would expect that we have to lay in the broad, rough pencil outlines of understanding before we can paint in each finesse.  I think all of the above quotes are probably true for different levels of player, with different amounts of study time available; what are your ambitions, what can you manage?

7. "Reading and nodding is not learning", says Jonathan Rowson (after Nigel Davies), and he's absolutely right.  You have to do something more deliberate to get the information in, and you have to practise getting it out, too.  Read and nod, yes, but put a line in the margin beside everything that is important, then take some notes, and see if you can get all the main points of a chapter down on one side of a piece of paper.  Shut the book and read your notes and see if they still make sense.  Then put aside your notes, get out a set and a board (or an empty game in a database) and see what you can recall.  Wait a day, then try writing it all down from memory, or, even better, explaining it to someone else and see when you dry up and which questions you can't answer; go back to the book to patch up what you know.  Wait a week, and do the same.  Play a bunch of 15-move games, write the moves down, see where you or your opponent went wrong, see if there was a better way to take advantage of any mistakes. 

8. Whatever the green lobby tell me, I always think that the natural resource in shortest supply is time.  There are ways to cut down the amount of theory you need to know:

Choosing non-confrontational openings: Purdy once said (re-published in Action Chess) you can learn all the opening theory you need "to get by" in 10 hours.  He recommended the Colle System (1.d4 2.Nf3 3.e3) as White, and the Rubinstein French ( 1.e4 e6 2.d4 d5 3.span style="font-style: italic;"Nd2/Nc3 dxe4) as Black against 1.e4.  Against 1.d4 and other closed openings, he recommended you aim for a standard formation of ...d5/...e6/...Nf6/...Be7/...b6/...Bb7/...c5, which, against the Queen's Gambit, is called the Tartakower variation (first seen at the London 1922 tournament ).  He didn't say these were great openings, just systems where there was a decent trade-off between strength and ease of use.  If you want to do more than "get by", maybe you need to study more, but Purdy was very clear that spending 10 hours studying the middlegame or endgame would do more for your chess results than any 10 hours on the opening.

Choosing early deviations: "Is it possible to play the Ruy Lopez with just strategical ideas? Probably not. Is it possible to memorize all the lines? Probably not." -- Alexei SHIROV.  I liked Andrew Greet's championing of the Worrall Attack in the Ruy Lopez, a deviation from the main line by 5.Qe2.  [He then rather spoiled things by writing a book large enough to act  as a shelter during inclement weather (or even nuclear attack), but  the  approach is sound.]  Retracing his steps, we might recommend the Ruy Lopez Exchange Variation (4.Bxc6), or the Scotch Game (3.d4), or Bishop's Opening (2.Bc4), or even Bird's Opening (1.f4) as ways of making it more likely that the battle takes place on ground of your choosing.  Then you have to learn less.

White and Black: You can adopt similar systems as White and Black; obvious examples include the King's Indian Defence and the King's Indian Attack, Bird's Opening and the Dutch Defence, the English Opening and the Sicilian Defence, and so on.

Sister openings: You might think about playing the Semi-Slav (...d5/...e6/...c6) as Black against 1.d4, and perhaps the Caro-Kann (1...c6) or Scandinavian Defence (1...d5) as Black against 1.e4.  These openings share some features in common, and your pieces will often end up on similar squares.  These would form a decent set of openings to use if you play the Colle System as White.

Moving in the same street: rather than taking up an entirely new opening, you are better off using as much of your existing knowledge as possible, and moving to a different variation of the same opening.  So, if you already play the Ruy Lopez, you might prefer on move 5 to play the Centre Attack (5.d4), the Four Knights' (5.Nc3), the Andersson-Steinitz line (5.d3), or some other variation where a lot of your homework will still prove useful.

Traps: I have no problem if people include lines in their repertoire where traps arise.  But you shouldn't choose an opening just because it has a trap, and you may find that none of your rehearsed traps ever arise in practice.  Ask around... I know Ray has had some success with 1.e4 d5 2.exd5 Nf6 3.c4 e6!?, getting in a few standard tricks, while although I like very much to play the MacCutcheon French (span style="font-weight: bold;"1.e4 e6 2.d4 d5 3.Nc3 Nf6 4.Bg5 Bb4), I never get to play it, because White nearly always plays 4.e5 or 3.e5...  So everything I know about this line is just 'dead weight' in practice.

Wacky openings?: Grob's Attack (1.g4 e.g. 1...d5 2.Bg2 Bxg4 3.c4) will always have a fan base, but it's only a surprise once and antidotes can be found.  The trouble is, while any opening can be played with shallow ideas or deep ones, it's easier to find a layer of deeper ideas in a mainstream opening.  You are then faced with learning the whole of a new opening, with your hours of study of the obscure lines of the Grob being wasted.  "...kids need to move on to real openings as quickly as possible..." opines Nigel Davies.

10. Learning from what?

Dvoretsky recommends building up your own card index; most of us I think would prefer to take a pencil and a stack of sticky notes to a book.

Books: There are different levels of book with different styles which might suit different levels and styles of player.  There are excellent books about things I couldn't have imagined when I starting reading chess books... I guess there's a ladder of difficulty which can be hinted at in the title: "Starting Out..." "Explained..." "Winning with..." "The Complete..." and even "The Ultimate..." represent some sort of hierarchy, and depending on your grade, memory and study habits, I expect you could decide what's appropriate for you.  Some books have key ideas, some use a tree structure, others are based on complete games, some use a mix of approaches. Complete games undoubtedly help you remember themes, but are also undoubtedly used as padding.  It's not enough that a book contains detail, it needs to be useful, organised and well-explained detail with a point.  So, even if you see detail, don't assume it will do you any good.  As a rule, more elementary books use illustrative games, and more advanced books use 'trees' of variations, but Lev Psakhis arranged all his material in his monumental tetralogy on the French around illustrative games.

Show and tell: I've always thought that videos, DVDs and so on, however memorable, have a terribly small amount of content.  I've had to transcribe every video I've ever owned because (a) such a transcript wasn't packaged with the product, and (b) it's essential to compare the lines given with other sources, otherwise you have to keep zipping back and forth to find the position you are interested in.  And when you do this, you might find a video costing pound;15 has about a page of variations in it.  They really have to have a substantial accompanying booklet to be worth it.

Software: Most databases have all the features you need to organise a repertoire: trees, sort functions, training modes and so on.  There is even software ( BookUp) which is designed specifically to help you organise and test yourself on the openings you play, and boasts a transpositional feature which I haven't heard about elsewhere.

Hybrid DVDs: The latest development seems to be a DVD bearing a database linked to videos of talking heads.  I haven't seen many of these but if the talking heads are telling you important things, then it's the same pain as with a video finding the part where they talk about the position you are interested in, and if they're not telling you important things, well, it does beg the question, why the talking heads...

Class: