Which Openings Should You Learn?

Dec 12, 2007, 3:10 PM |

Every beginning chess player who aspires to become better has asked this question.  The author of one introductory openings book that I read suggests that you find an opening that suits your style, and learn that.  But my immediate reaction to that advice was that I was a beginner.  My style involved lots of losing and I wanted to stay away from those kinds of openings.  So I kept looking for other answers to the question.

As I pursued those answers, I became intrigued by the question itself.  I started wondering about the criteria by which one would judge the validity of any answer.  Why should I learn the Sicilian, as opposed to the Ruy Lopez, for example?  Is there some metric by which I can evaluate the merits of one opening vs. another?  Better yet, is there an algorithm that I can learn which would allow me to play through virtually any opening and reach a middle game position that is at least as good as my opponent’s?

If you have had similar questions in your mind, then relax.  I’ve already done all the work for you and I am willing to share my findings.  You will, no doubt, be comforted by the results of this research.

I will begin and end the same way by presenting you with two simple algorithms.  The first such procedure may be regarded as slightly unorthodox, but I personally guarantee it will turn you into an openings expert.

Method 1: Monkeys at a Typewriter

Perhaps you’ve heard of the following thought experiment, which is sometimes presented in mathematics classes to explain the wonders of infinite processes.  Imagine a room that contains an infinite number of monkeys sitting at typewriters, banging away on them at random.  

(Note to readers under the age of 30:  If you are unfamiliar with the word ‘typewriter’ think of it as an analog computer that has a keyboard, but where actual physical paper represents the LCD screen, and a ribbon (I’m not making this up) that is drenched with liquid ink represents the printer, and these three components are connected by a set of mechanical levers.  When the user strikes a key with significant force, it propels a lever that whacks a small letter-shaped club against the wet ribbon, imprinting the image of that letter against the paper.  Your parents actually used such devices, which explains, at least in part, their current inability to use contemporary computers.)

 So if these monkey tap-tap-tap away for an infinitude of time, eventually at least one of them will produce “The Compleat Works of William Shakespeare” without the benefit of plagiarism.  With that conceptual introduction, here is the first suggested method of playing the opening.

1. Look at all of the legal moves on the board.
2. Select one at random.
3. If you lose the game, reject this move in the same position in future games and play a different move at random.

You will lose a few games at the start of this process, but if you follow the procedure as stated and if you play well in the middle and end game, over time it will converge to an optimal solution with the added benefit that you will never have to memorize an openings book.

The disadvantage of this method, however, is that it requires patience.  If you are pressed for time, then I have other very good solutions for you so keep reading.

Method 2: Statistics

We all want to win the most games possible, so why not choose those openings that have a history of success?  Point your browser to www.chessgames.com and then click on the link to “Openings” near the top of the page.  You’ll need a premium membership to explore fully, but anyone can explore enough to satisfy our current goal.  You will see a page of opening moves for White, ordered by their frequency of occurrence in a large database.  You will also see the winning percentages for those games in the database where that move was played.  As I write this it appears that 1. Na4 might guarantee you a win, since 100% of the games where that was the first move resulted in a win for White.  However, there were only 4 such games, so you may want to reconsider.  The most common first move is 1. e4 which results in a win for White 37.1% of the time, and a win for Black 34.1% of the time.  If you like to play e4, then click it and you will see a new page with the same kind of information associated with Black’s first move.  At the bottom of the page (after a few more clicks) you’ll begin to see the names of the openings that correspond to the moves up to that point.

You play chess.  You are therefore smart, and I’m sure you can see where I’m headed with this.  Click away and you’ll be able to discover the winning chances for both White and Black for virtually any opening, based on the results of nearly a half million games in their database.  Find an opening that appeals to you and which the statistics suggest may give you an edge, and learn that opening.  Some openings show better results for White and some for Black.

This is not an unreasonable way to decide what openings to learn, but there are other methods too, so let us continue.

Method 3:  Goofy Names

In response to a survey question on chess.com, I noticed that the vast majority of members identify themselves most closely with the knight.  Knights are odd and quirky -- traits which chess players also possess.   Thus, you might be attracted to an opening just because it has an odd and quirky name, and you’d like to be able to say that you play it.  If this describes you, then the openings to consider include the Grob, the Halloween, the Accelerated Dragon Sicilian (which may appeal to those who live in north Jersey), the Göring Gambit (for neo-Nazis), the Kalishnikov (for modern revolutionaries), the Pelikan, the Hedgehog Formation, and the Orangutang.   And if you just don’t give a damn, there’s always the Boring Opening.

Method 4:  Brute Force

In an earlier blog post I noted that Modern Chess Openings, 14th edition (MCO-14) contains about 15,000 moves for White and an equal number for Black for a total of approximately 30,000 individual tempos.  

Memorize them.

There is ample evidence that this is not as impossible as you may think.  First of all, any well-educated person has a working vocabulary of around 50,000 words, and a vocabulary of 30,000 is rather common.  So many of you have already learned this many items in another context.  

The Nobel prize winner Alexander Solzhenitsyn, who was a guest of the Soviet gulag for many years, wrote that one prisoner he knew mentally composed and retained in memory a poem of about 30,000 lines.  And any psychologist worth his or her hourly fee should recall from graduate school that it takes about 50,000 “chunks” of knowledge to become an expert at anything, be it woodworking or mathematics.  

So if you are young enough to begin the task, consider learning MCO-14.  This won’t help you to play well if your opponent strays from the approved lines, but if you can advertise your knowledge appropriately, you just might get invited to be a guest on David Letterman’s show.

Method 5:  Tree Pruning

The most popular opening these days in tournament play, and also confirmed by the statistics at chessgames.com, is the Sicilian, which begins  1. e4  c5.  Why not learn the opening that you are quite likely to end up playing in a tournament?  When I looked up the Sicilian in MCO-14 I found no less than 39 tables of moves, the most complex opening in the book.  Nick de Firmian, the author of MCO-14, suggests in his introduction to that book that the beginner find an opening that involves “a minimum of effort”.  

That sounds reasonable to me, so why not learn an opening with only a single table of moves, e.g. the Center Game?  1. e4 e5 2. d4.

The problem with this is that if I play 1. e4, my opponent will often play 1…c5 and we’re in the Sicilian!  So I wondered how I might be able to find a way to play an opening that I want to play, rather than one my opponent wants.

With this in mind, some time ago I took MCO-14 and sketched out the “tree” of moves from the opening position as cited in the table of contents.  That is, the root of the tree is written as a dot, and each first move is written as a branch from that root to another dot.  Black’s various replies are drawn as branches emanating from those dots, and so forth.  This is the same kind of tree discussed by Kotov in his classic book on analysis for those of you familiar with that.  

I did not draw the complete openings tree of MCO-14, but I did do the table of contents and a few more moves on some of the openings.

Then I considered the structure of this tree, and how I might select an opening that essentially forces my opponent to play the opening that I have learned.  If I play e4 as White, my opponent can always force the Sicilian on me.  But if I play 1. f4 as White, we are pretty much playing Bird’s Opening with only one MCO-14 table to learn.  Granted, Bird’s Opening is not commonly played, as it weakens the protection around the King, but your opponent is not likely to be familiar with it, so by playing it you can force him into unfamiliar territory.  And top players have had and still do have BO in their repertoire, including Larsen, Blackburne, Staunton, Henry Bird himself, and even Bobby Fischer.  (Sorry for the juvenile humor.)

If you are Black, you can reply to 1. e4 by playing 1…d5, and you have forced your opponent into the Scandinavian.

I said that I would begin and end this post the same way, with a simple procedure to learn that would result in optimal play.  I now present you with that final method, which was perfected by a former World Champion regarded by many as the best chess player who ever lived.

Method 6:  The Garry Kasparov Algorithm

1. Consider all legal moves on the board.
2. Select the best one.

This is not only simple to remember, but it has the added benefit of applying to any stage of the game, not just the openings.  If you follow it correctly, you will win many games.

I hope you found the above entertaining, if not a little instructive.  Regardless of what openings you decide to learn, keep this one rule in mind.  Choose your move carefully, in chess and in life.