What Is Overlearning?

I came up with a concept I call “overlearning” to help myself improve at chess. So far this fall I haven’t made much progress, partly due to upheavals in my daily routine, but I think if I stick with it, eventually I’ll start to get more traction. And I will stick with it.

What is overlearning? Overlearning consists of repetitive study of the same material, gradually expanding the core amount of material that you have learned and therefore understand.

A chessplayer who wants to get better at chess needs to improve his (or her) skills in many specific areas, including:

1)      Spotting tactics and creating tactical opportunities;

2)      The technical endgame phase;

3)      Positional play, a.k.a. “what to do when there’s nothing to do”;

4)      The opening;

5)      Practical decisionmaking (including time management) in games; and

6)      Game analysis after the game is over.

After identifying the skills you need to succeed, you will want to practice those skills again and again, even after you think you have mastered them. Why does overlearning have to be repetitive? Well, doing things over and over is how you get good at them and stay good at them, whether you are a football player running plays on the practice field, a concert pianist practicing scales, or a chessplayer solving diagrams of tactical positions. Every skilled activity has its drills that you have to practice repeatedly if you want to perform at a high level consistently. In a physical sport, you want to develop your kinesthetic sense (“muscle memory”) and a host of specific sport-dependent physical skills. In chess, you want to develop pattern recognition and a multitude of specific chess skills. Even non-competitive skills, such as singing, require that you practice, practice, practice. The octogenarian pop singer Tony Bennett is still on top of the show-business world. A recent article says: “At 86, Bennett, who is performing tonight at the Seneca Niagara Casino, just keeps on singing. He is healthy and fit. Pop stars, led by Lady Gaga, idolize him” (Mary Kunz Goldman, The Buffalo [New York] News, October 5, 2012). How does Bennett do it? In the article he says, “It’s important to keep up vocal training. As they say, if you don’t do your scales on the first day, you know it, on the second day the band knows it, and on the third day the audience knows it.”

Chess overlearning, to be useful, requires that you do actual chess work. This means making concrete decisions to solve practical problems, ideally of the type you are likely to encounter in your games. Playing good chess is essentially making a series of good decisions. You can’t read passively and hope to gain anything. You get better at chess not so much by learning facts as by improving skills. As GM Nigel Davies put it with great insight, “Focus on the how, not the what.”

Overlearning, according to my concept, consists of these steps:

1)      Identifying the essential skills that must be mastered;

2)      Listing items (e.g., books, DVDs, et cetera) that will help you practice these skills, starting with the most basic or easily mastered material and progressing to the most difficult;

3)      Studying and attempting to master the first item on the list (i.e., the easiest item);

4)      Repeating the study of the first item on the list, to increase mastery;

5)      Building on the confidence now gained and the skills now learned by adding the study of the second item on the list;

6)      Repeating the study of the first two items on the list;

7)      Building on the confidence gained and skills learned from the first two items by adding the study of the third item on the list; and

8)      Continuing in like manner, continually increasing mastery of items previously studied, and adding new items which in turn are gradually mastered. There is no end to this inexorable process of mastery until you decide to end it, or nature decides to end you.

How did I come up with my overlearning concept? Hard to say exactly, but I trace it to two things. First, I have always found it difficult to master the material in my chess books with only one reading. I like to believe I have a reasonable level of general intelligence, but experience has shown I have no remarkable chess gifts. I concluded I needed to read chess books more than once to learn the material well. Second, I remembered how I had tackled a somewhat similar problem in my college Spanish class back in the late 1970s. Every chapter in our first-year Spanish textbook included a longish diálogo, or dialogue, which we students had to memorize for the chapter test. We were given no guidance as to how to go about this task. I remember sitting in the college language lab and feeling baffled the first time I was confronted with the perplexing chore of memorizing twenty or thirty lines of Spanish dialogue. The task seemed impossible. Then I got an idea—where it came from, I don’t know; perhaps the Universal Mind whispered to my inner ear. I read aloud the first sentence. Then I looked away from the book, and repeated the first sentence flawlessly. That wasn’t too hard. Then I looked at the book again, and read aloud the first two sentences. Turning away from the book, I could repeat those lines, too. I looked back at the book, and read aloud the first three sentences. I can’t remember where I started to break down, but wherever that was in the diálogo, I repeated the material up to that point until I could do it flawlessly. Then I moved on. Eventually I was done, and you know what? It didn’t take very long to memorize the whole thing. Maybe half an hour. I got better at memorizing as I went along, and after all, it was a conversation, so the various sentences hung together as a coherent whole, with the earlier material pointing the way toward the later material, and the later material referring back to what came before. The idea had worked. I had developed an effective tool for memorizing longish chunks of Spanish text. The ancient Romans had a catch-phrase describing my method: Repetitio mater memoriae (Repetition is the mother of memory). Years later, I realized I could apply approximately the same method to learning complex chess material, too.

Of course, learning chess skills that you hope to apply over the decades of your chess career is a much more complex intellectual activity than memorizing lines of Spanish that you will allow yourself to forget as soon as the chapter test is over. For one thing, you must approach chess with a critical mind, always asking questions about the material as you examine it. Do not accept any chess information uncritically—even from a top grandmaster! Let me illustrate this very important point. Just yesterday, I inserted a new chess DVD into my laptop: Sergei Tiviakov’s “Sicilian Defense with 2.c3, Alapin Variation.”

As an aside, it’s not so easy to overlearn opening theory efficiently, with a critical mind. At least, I haven’t figured out how. For me, it’s obvious that you can overlearn tactics by solving lots of diagrams, leavened with occasional injects of theory (for example, examples of the basic mates, or discussions of how to carry out a typical attack). Endings are a bit more of a challenge. My current approach is to read about the theory of a position first (say, R+P versus R), then ask myself a few questions about the position (“What if the defender did this instead?”), and finally to test myself by playing out the position against Fritz. Unfortunately Fritz is not an ideal opponent in theoretical endgames, because Fritz often will make a random, obviously losing move when it has decided all roads lead to a loss and there is no hope; whereas a human opponent would continue to make good moves that pose practical challenges. I haven’t yet figured out how to overlearn openings effectively, yet one must nevertheless try to learn openings.

Let us now return to Tiviakov’s DVD. GM Tiviakov is an expert on the Alapin Sicilian who has scored very well with it; he says that during 2005-2006 he won twelve games in a row with it! The DVD starts with one of his early Alapin games, played in 1989 when he was only 16, against the 14-year-old Vladimir Kramnik. Tiviakov achieved a good opening advantage, won a pawn in the middlegame, and finally prevailed in 54 moves. He says he is still proud of this early effort. On the DVD Tiviakov lavished much time and explanation on the game’s early moves. However, he did not explain one early move that would be good for an amateur to understand:

In the above position which could have occurred in the game, Black could capture a second pawn with 7…dxc3, but instead 7…d5 is recommended without comment. I guarantee you that an amateur with the Black pieces would consider 7…dxc3 apparently winning a pawn. Already at this early point in the opening, I had arrived at a position I didn’t understand. I stopped the DVD, opened Fritz, and looked at the position for a while. I could see that 7…dxc3 8 Nxc3 led to some White pressure, and White eventually won back the pawn, but after thirty moves or so, Fritz evaluated Black as having slightly the better of an equal position! Admittedly I hadn’t given Fritz much time to think, but clearly the problems in the position after 7…dxc3 are non-trivial for a player of my level, and I would need to spend more time looking at them.

After Tiviakov won a pawn in his game against Kramnik, he said essentially, “Now we have entered the technical phase, and I will play through the remaining moves of the game quickly.” Which he proceeded to do. I tried to follow the rapid-fire moves as best I could, but it was almost hopeless. Then I saw the position below flash past, and I said, “Whoa!”

White has just played 45 Ne3. Here, according to Tiviakov, Kramnik played 45…Bd7. But wouldn’t 45…Rd4++ be better? I doubt Kramnik, even at age 14, would have missed this move, which I, a lowly amateur, noticed as the moves zipped past me at light speed in the DVD presentation. I halted the DVD and inserted the winning move 45…Rd4++ as a variation (perhaps you can just make it out in the screen shot above).

What is the explanation for 45…Rd4++? Did Kramnik miss mate in one during severe time trouble? Not likely, at move 45; especially since Tiviakov did not mention that Kramnik missed mate in one. I’m guessing, and this is only a guess, that Tiviakov was showing the game from memory, and perhaps transposed some moves by accident. Haven’t we all tried to show off a game to our friends from memory, lost our way, and then said, while rearranging the pieces, “Anyway, we got to this position …” and gone on from there? Whatever the explanation, my point is the same: you must always study chess with a critical mind, taking nothing on faith. In fact, you should continually be trying to falsify whatever information you are given! Only thus can you cultivate the independent mind you need to succeed as a practical chessplayer.