Should Amateurs Try to Specialize?

GM Nigel Davies recently posted the following remarks:

A few weeks ago my son won a couple of games with the £$%^&! (sorry, I can’t say what it is) opening, something that has become a firm favorite of his. It’s a kind of secret weapon.

There’s a lot to be said for having such weapons, but with certain reservations. On the one hand they can help summon up enthusiasm for games and studying its wrinkles. On the other hand a fixation on one particular line of play can take attention away from general improvement and a broader sweep of ideas.

I do not disagree with anything GM Davies says here.

Chess amateurs almost by definition have severely limited time available to improve their chess skills. Should we focus our efforts on trying to acquire a basic competency in all essential areas? Or should we accept that we will never reach a high level in our general play, and indulge ourselves by studying a favorite gambit that we may rarely get to use, or by focusing on (say) pawn endings to the neglect of other endings, if pawn endings are what we enjoy?

Asking such questions raises another more fundamental question. Why do we play chess? If it is to win as often as possible, perhaps we should take a severely practical view of our chess studies (to the extent that chess studies can be described as practical). If it is to have fun, perhaps we should spend our limited chess time on bullet chess or whatever activities bring us immediate gratification. Most of us, I presume, fall somewhere in the middle: we want to have fun right now, but we also hope to win games in the future. So we compromise, and allocate our time accordingly.

An argument could be made that amateurs who specialize, by spending most of their time on a few chess topics, may reap certain benefits not available to their more sober peers, who strive for solid mediocrity in all areas, without indulging themselves in any flights of impractical specialization or suffering from any outrageous gaps in their skill set.

Consider, for example, Player A, who knows a moderate amount about most openings he plays, including the basic plans. He is fair at tactics and has read Jeremy Silman on positional play. He is familiar with basic positions in many types of endgame.

By contrast, Player B spends most of his study time on three topics: the Evans Gambit, tactical problems, and pawn endings. His knowledge of other openings is relatively shallow, his positional understanding remains undeveloped, and his knowledge of rook endings is limited to a shaky grasp of the Lucena position.

Let us assume that Player A and Player B are equally talented and put in the same amount of chess work. However, due to his more practical allocation of resources, Player A is rated 2100 Elo, while Player B is rated 2000.

Ratings don’t lie. The ratings tell us that Player A is likely to be more successful than Player B in any given tournament. But one could construct an argument that Player B may, from time to time, exceed statistical expectations to a greater degree than would be predicted by a normal distribution of tournament results.

The reasons for this anomaly are twofold: Player B’s idiosyncratic approach to chess, combined with the very real existence of luck in chess.

In theory there is no luck in chess, but we all know better. In some tournaments, our opponents play into our pet openings. Or they go in for middlegame tactics, and the types of positions that arise happen to be among those we have studied. Or we get three pawn endgames in six rounds, and we love pawn endgames. These things really happen from time to time. When they happen to Player B, he will thrive, because when he is bad he flounders, but when he is good he excels.

Consider what could happen in a five-round weekend tournament, called the Random Walk Open. Probably won’t, but could.

In round 1 of the Random Walk Open, Player B is paired up to a master, to whom he would normally lose badly. But this time he gets the opportunity to play his pet Evans Gambit and wins with a book trap he knows. (I have succeeded this way against better players a few times, though not with the Evans.)

In round 2, Player B survives an opening he doesn’t know. Out of ignorance, he plays original moves that lead to an unusual position. Then he outplays his generally stronger opponent because the tactical patterns that randomly appear on the board remind him of positions he knows well from a book of combinations. (Been there and done that.)

In round 3, Player B is crushed by a strong master. (This has certainly happened to me many times!)

In round 4, Player B gets a mildly inferior pawn ending against a tough opponent, but makes the most of his chances and wins. (I have done this exact thing a couple of times, including once in Philadelphia against future U.S. Women’s Champion Jennifer Shahade, before she “got good.”)

In round 5, Player B gets paired against a much stronger player who really ought to beat him soundly—a “heffalump,” as Simon Webb calls a big, scary opponent in his charming book Chess for Tigers (with thanks to A.A. Milne). It’s the last round and the stronger player wants to win—money is on the line—so the stronger player goes in for aggressive tactics. In an unclear melee, though Player B has been outplayed to this point in the game and is down material, the stronger player makes an oversight. Player B lands a lucky punch and wins. (See my last blog post!)

Player B does not finish first in the Random Walk Open, but he scores 4-1 for clear second place. He wins a trophy and large money prize. He achieves a certain notoriety in local chess circles, and has memories for a lifetime. The excitement of his success on this memorable occasion will sustain him through many years of lesser results. (I have never strung together quite as many lucky results in one tournament as Player B … but it could happen!)

Meanwhile, Player A has performed roughly to expectations, with a 3-2 result. He maintains his 2100 rating but gets no prize. His result is like most of his other chess results, and therefore not memorable.

(We have described two idealized chess amateurs. What if they were real people? How would their different results in the Random Walk Open affect their future lives? Perhaps as follows. Player B, flush with cash and endorphins, goes out to a pub after the last round and parties with his friends. At the pub he meets a nice girl; his good cheer impresses her favorably; in time they marry. He becomes a successful salesman, continues to play his optimistic brand of chess with varying results, and lives happily ever after. Player A, whose creditable but completely expected result nets him no prize and no joy, glumly returns to his dark and chilly apartment, where no birds sing. He becomes a chartered accountant and gives up chess as an impractical time-wasting pursuit. He never marries and dies alone, surrounded by women’s shoes.)

Ahem—back to the main line of our discussion. In the next few tournaments after the Random Walk Open, neither Player A nor Player B enjoys any unusual degree of luck. Player A continues to play solidly and maintains his 2100 rating. Player B’s many weaknesses are on display; he loses back his “lucky” rating points and regresses to his former 2000 rating.

Perhaps our example was too extreme. Perhaps Player B, due to the gaps in his play, is very unlikely to score 4-1 against a strong field, even in the Random Walk Open. But let us grant that he is dangerous even to better players in certain types of position, and may well knock off a stronger opponent from time to time, who happens to wander into his minefield. Over time he is almost guaranteed to accrue some memorable games, even if they do not all occur on the same weekend. Famous historical players of this type arguably include Frank J. Marshall and David Janowsky, both of whom could play spectacular attacking chess, but whose reckless style often did not hold up against more solid, well-rounded players. A player’s skill in the endgame is often cited as an indicator of his over-all level, but Janowsky is reported to have said, “I despise endgames.”

Player A, on the other hand, will perform at a higher average level than Player B. He may score fewer upsets, but his performance will be more reliable and he will win more games overall.

If both players are ambitious and willing to do some work, Player A, because he has no serious weaknesses, may continue to increase his rating advantage over Player B as time passes. Over time, Player B’s strengths and weaknesses will become common knowledge. The higher he manages to rise—if he is able to rise at all—the more skillful his opponents will be in avoiding his strengths and exploiting his weaknesses.

According to an old saw, the second-best swordsman will always lose to the best swordsman. But an unschooled country bumpkin swinging his blade wildly may accidentally beat down the guard of the best swordsman and win. Think of Player A as the second-best swordsman. He will play sound, predictable chess, and will lose predictably to better players. Player B, of course, is the bumpkin who might get lucky!

The question I asked at the top of this post was, should amateurs try to specialize?

Put another way: would you rather be Player A or Player B?


Author: Tim Hanke

Tim Hanke is a U.S. amateur who still believes, despite much evidence to the contrary, that he can become a decent chessplayer.