ELECTRONIC T-NOTES


CHESSBASE USA'S WEEKLY ON-LINE NEWSLETTER


FOR THE WEEK OF OCTOBER 21, 2001


THE USE OF DATABASES

by Steve Lopez

A few years ago, I was talking with a fellow friend in the software business. My friend asked me what kind of software I sold.

"Oh! You sell programs that play chess!" she replied. "Neat!"

"And we also sell chess database programs."

There was a moment of silence. "Ummm...what's that?"

"You know, it's like a regular database. You can store hundreds of thousands of chess games and then find the ones you want in just a few seconds, then play through them right on the screen."

Another moment of silence. "Uhhh, OK. But why?"

I'll freely admit that I was taken aback by the question. If somebody's a chess freak, they already know the answer to that question. But what about people who aren't gonzo over chess? Or casual or new chessplayers? Or even serious players who haven't been exposed to databases?

It wasn't the last time I've been asked that question, not by a longshot. And I'm sometimes still asked that question by people who consider themselves pretty serious and accomplished chessplayers.

There's no short answer to that question, though I wish there was one. Any answer I give is either incomplete by the questioner's standards or insufficient by my own standards. Chess databases aren't the major miracle of modern civilization but, for chessplayers serious about improving their play, they're dang close. But how do you explain them to someone who's never used one?

A good start is to repeat my answer to my fellow software vendor. She'd asked me why anyone would want to replay old chess games. After considering the question for a few moments, I found what I thought was a decent, albeit basic, answer.

"Chessplayers, moreso than fans of any other game or sport (with the possible exception of baseball), are totally obsessed with the past. Just like baseball players who argue endlessly over whether Cobb or Williams was the better hitter, chessplayers will squabble at the drop of a hat over whether Capablanca or Alekhine was a more talented player. Just as there are modern-day fans of baseball players who have been dead for decades, many chessplayers still worship the play of Paul Morphy. But unlike baseball fans who are still fans of Shoeless Joe and have no chance of ever seeing the man actually play the game, chessplayers have an advantage -- we can still replay the games of players who have been dead and gone for years or even decades."

So why do we do this? Some of us do it for enjoyment. I still get a huge charge out of pulling up some games of Morphy, Tchigorin, or Anderssen and watching the brilliant sacrifices and tactical fireworks. But the biggest reason for replaying old games is to learn about the game of chess, to increase our knowledge so that we come away with a little bit more than we brought to the table when we sat down to replay a game or two.

Any chess author worth a rip knows this. Just about every chess book you've ever read (or will ever read) provides illustrative examples from past games, showing the relevant ideas and concepts in action. The author will typically discuss an idea (the fianchettoed Bishop, for example) and then follow it up with a game or two (or bits of several games) that show the fianchettoed Bishop in action. Some authors will take a complete game and describe the action (like a color commentator on a sports broadcast), pointing out the important positions as he goes with the aim of making the reader a better chessplayer -- Irving Chernev did this wonderfully well in his book The Most Instructive Games of Chess Ever Played.

Few people learn solely by doing. Oh, there are a few hyperbrainy mugs who start off their first tournament by playing like a patzer and end up killing everyone by the fourth or fifth game, but such prodigies are rare. Most of us need to be shown a concept, idea, or procedure before we make it our own. In fact, many of us synthesize the two processes -- we're shown an idea, we try it a few times, and it becomes part of our store of knowledge. But seeing the concept in action first is critical for most of us.

And this is why we replay old chess games. A lot of chess mastery consists of successful pattern recognition -- the ability to see recurring patterns in slightly dissimilar positions. The most basic form of this is when we study checkmating patterns as beginners. We learn the basic mates and then look for them in our own games. Pattern recognition comes into play much more frequently than many players realize; I recently reviewed the basic mate with two Bishops in an endgame book I was flipping through -- and sure enough, I had opportunities to successfully put my opponents away with that mate (or a variant of it) twice in the last month.

Where do databases enter the equation? Prior to the advent of chess databases, we were at the mercy of chess authors and editors. An author might introduce a concept and give some examples, but sometimes two examples just aren't enough. Or worse yet, maybe he gives some bad examples (I can think of one book in particular in which this was the case, but I'll not embarass the author by revealing the title). Editorial choices also influenced what chessplayers were exposed to -- for example, the Chess Informant was the bible of tournament chessplayers for many years, but the Informant carried only the recent games that the editors thought were "theoretically significant". Thta's all well and good if you enjoy playing openings that are the "flavor of the moment". But for players like me who enjoy unorthodox openings, the Informant rarely contained much of specific interest. I enjoyed playing over some of the epic battles in top-level chess (I always loved the times when Kasparov and Kamsky duked it out over the board) but I didn't see much of practical value to myself in the Informant regarding my own opening choices.

While a little guidance by strong players is absolutely a good thing, why should we be constrained by seeing only the games they select for us? A chess database allows you to access anything you want to replay, whether it's a noted theoretical duel in the King's Indian Defense played by top grandmasters last week or a Halasz Gambit played by "mere" masters or experts in an obscure European correspondence event five years ago.

The key idea here is that a chess database allows you to find and replay exactly what you want to see -- whether it's a specific middlegame motif or some old 1800's games that you get a kick out of replaying. And you can find this stuff in seconds or minutes, without plowing through volume after volume in printed form, searching for that one elusive Doug Root game in which he played 4.Qg4 in the French Advance, published in a back issue of the Informant that's lying around buried somewhere on your bookshelf.

We're already seeing the answer to my friend's question taking shape here. There are a lot of reasons for using a chess database to find and replay past games (whether long past or a game played just last week): education, information, enjoyment. You can find answers to specific chess questions, look up many examples of a particular motif in action, or simply pull up a batch of your favorite grandmaster's games for your viewing entertainment. And, if you take your time and pay a little attention when you replay these games, you'll often come away with a new bit of chess knowledge that you didn't have before.

The "new knowledge" angle is critical. We as chessplayers can learn an awful lot from the past, and we'll look at some concrete examples the next time around. Until then, have fun!

IMPORTANT NOTICE: As of Nov. 1st, 2001, I will have a new e-mail address. The old e-mail address given in previous ETN issues will no longer be active. From now on, you can e-mail me here with your comments, suggestions, and analysis for Electronic T-Notes.