ChessBase and Correspondence chess – part 1

Before we get into the re-presentation of this old column, please be aware that various correspondence leagues and websites have different rules regarding the use of computer chess engines as an analysis aid. Most leagues/websites still prohibit their use, except for special events or separate areas of the site which allow the use of a chess engine; ergo the comments I make below about their use are still by and large in effect.

WHY PLAY CORRESPONDENCE?

Correspondence chess has been a hot topic and subject of debate on the Internet lately, so let me be the latest to stick my hand in the hornet's nest...

I'm of the opinion that all chess players should give correspondence chess a try at least once. It's a totally different discipline from over the board chess. Obviously, you have more time to ponder your moves, which leads to a more relaxed atmosphere. At the same time, it requires a great deal more patience than OTB chess. Losing a correspondence game is somewhat less embarassing than losing one in OTB. This leads, in turn, to a desire to play more adventurous chess. Many gambits that are rarely seen in OTB play (the Cochrane Gambit is a good example) regularly appear in postal events.

With correspondence chess, you can play chess whenever you want. If a card comes in the mail and you don't feel like looking at it, you can just put it aside for a day or two until you're ready to consider your reply. And you always have the option of studying positions from your postal games while waiting for replies to come in the mail.

Correspondence chess is easier for older players. It gets tougher to visualize as one gets older; correspondence chess allows you to move the pieces around as much as you want before deciding on a move. It's also ideal for physically-challenged players who find it difficult to get to tournaments. The same thing applies to players who don't have access to OTB events due to geographical considerations. I first became involved with correspondence play when my wife was pregnant with twins. It was late in the pregnancy and I didn't have the luxury of driving two hours to play at some school or rec center that didn't have a telephone. Correspondence chess allowed me to play competitive chess in the comfort of my home while waiting for the magic words "Honey, it's time!".

It's really a liberating feeling to be presented with a knotty tactical problem and have three days to consider it, rather than the few short minutes allowed in OTB play. It's even sweeter when you find the correct continuation and end up winning the game!

There are social aspects to correspondence play as well. You'll sometimes find a letter instead of a postcard waiting for you when you open the mailbox. I carried on a lively correspondence with some of my opponents a few years ago; in fact, I spent more time writing letters than I did considering my replies! One can, however, safely disregard this aspect of play if one wishes. I'm currently in an event in which we just scribble a line or two at the bottoms of our cards; none of us really have time for lengthy letterwriting.

Correspondence chess is also a great way to improve your OTB game. Correspondence chess teaches you how to deeply analyze a position. Playing this type of chess can greatly improve your understanding of the game.

The world of correspondence chess is incredibly rich. It's the quest for that Holy Grail of chess: the perfect game. You have plenty of time to research openings, examine tactics, set and avoid traps, explore endgames. It's a wonderful form of chess, completely different from OTB or on-line chess. A chessplaying acquaintance once told me that he considers correspondence chess a form of meditation. I've never been quite that relaxed by it, but I do find it less stressful than OTB chess and I find it interesting to develop my concentration and research skills through playing chess by mail.

ChessBase and Fritz have a number of tools to help you as a correspondence player. If you have a large database of master and grandmaster games with a means of accessing it (e.g. ChessBase and Fritz), you already possess some of the important tools essential to correspondence play.

Here comes the part where I stick my head into the lion's mouth, by telling you how ChessBase can help you in your correspondence play.

Let's start by listing what types of computer use are generally legal in correspondence chess:

1. Using a game database to look up information;
2. Using a statistical tree program to combine games into a tree and give you won/loss percentages for specific moves;
3. Using endgame databases (Thompson, Nalimov, etc.) to look up endgame positions;
4. Using a computer program as a postal recorder, to keep track of moves played and time used;
5. Using a computer program to store your personal analysis.

Now let's list the things you generally can not use a computer to do:

1. Consulting a playing program to have it suggest a move;
2. Using a playing program to double-check your intended move for errors.

Why are some of these uses legal and not others? We'll go over the points one-by-one to examine them in-depth.

The use of database programs like ChessBase has always been the subject of contention ever since the mid-1980's when ChessBase was introduced to the market. "Old guard" players are offended by the idea that someone could use a database program to look up a game or position for use in their postal games. (A very few of these players also find the use of paper books/magazines/etc. objectionable as well, but this type of usage has generally been considered an acceptible practice for many years.)

Let's step back and look at how postal chess games were researched before the days of computer databases, particularly in the area of chess openings. A player would receive a card in the mail from his opponent. He would record the move in his notebook or log, set the position up on a board, and begin to ponder his reply. At some point, he would turn to his library and attempt to find a reference to that opening or position.

Typically (at least between the mid-1960's and the mid 1980's) this would involve consulting the Chess Informant series. The player would have to first identify his game by ECO code, then start digging through volume after volume to find any games that used the same opening sequence. He would also consult various opening manuals (like ECO or Modern Chess Openings) in an attempt to find the theoretically "best" continuation from his current board position.

Obviously, research was a very laborious, time-consuming process. Players could spend literally hours finding raw information before they even got to the stage of analyzing it!

Computers have the ability to speed this process enormously. As an example, one can use ChessBase and a large database to locate in seconds information that used to take hours to unearth.

Quite a few postal players (primarily those that didn't own computers) rebelled against this use of technology and protested to the major correspondence chess organizations. Their protests were universally shot down. The fact is that there is no difference qualitatively between consulting a chess library on paper and consulting one on disk. The same information is being accessed; one technique is just infinitely faster than the other. Thus every major postal chess organization has ruled that the use of computers for data storage and retrieval is within the rules of correspondence chess. The idea here is that consulting a book or database is not the same as consulting a stronger player; it's still up to the player to locate and evaluate the raw information for him/herself.

The same idea applies to the use of statistical tree programs. All that these programs do is merge large numbers of games into a single "tree" of variations and provide success percentages for individual moves in the tree. Again, the "old guard" protested strongly against the use of these programs, saying that they felt the computer was "suggesting moves". Again, the protests were overturned.

The rationale behind this decision goes like this: let's say that a person is given a stack of printed gamescores in which a certain position appears. That person could review those games by hand and compile a list of all of the moves that were played from the target position. By taking careful note of the end result of each of the games and assigning a numerical value to each result (White wins equalling 1.00, Black wins equalling 0.00, and draws equalling 0.50), he could average the results and assign a success rating to each of the candidate moves.

This does not necessarily guarantee that a particular move is good or bad -- the final results could depend entirely on moves played later in the games. It's merely a form of "bean counting": how many games did White win after 9.Bd3 was played in this position?

Now replace the human from our example with a computer. The computer will tabulate the results much more quickly and accurately than a human, but in theory the results will be the same (assuming no mistakes are made by the human). Thus a statistical tree program merely provides a function that the human player could perform himself (given the time and inclination).

The Ken Thompson and Eugene Nalimov Endgame database disks work under the same principle as a game database; in fact, the endgame disks are merely databases of all possible positions with a certain combination of pieces (King and Queen vs. King and Rook, for example). You just set up a board position and the computer displays all the legal moves with the eventual result of the game, assuming perfect play for both sides. These disks fall into a bit of a grey area in the minds of some players. Isn't this cheating?

It's actually not cheating, by definition. The computer is not generating a move for the player; it's merely accessing a database of stored board positions and assembling them in logical order. In theory, a human player could do the same thing "by hand", though it would be prohibitively complex and time-consuming to do so.

By now, you can see a pattern emerging here. Using a computer to generate (think up) a move is illegal. But using a computer to ease a research task that a human player could (at least in theory) perform for himself is not illegal.

In other words, you can't use Fritz to tell you what to play or to double-check the move you've decided on for tactical blunders, but you can use Fritz or ChessBase to look for databased games containing the same board position or similar positions.

There has recently [as of 1998 -- SL] been extreme backlash and harsh criticism against using computers at all in correspondence chess. I believe two factors have contributed to this attitude: more than a few players have been caught cheating by using a computer to aid them at correspondence and on-line chess (making honest players more than a bit wary of computers), and Kasparov's loss to Deep Blue last year has generated the vision (in some quarters) of the computer as the "enemy of chess".

I recently saw an interesting argument on the Internet concerning the use of computer programs as postal recorders. It's common practice among computer users to keep track of their postal games in progress as part of a computer database instead of keeping a record of a postal game in a notebook or a traditional postal log. A gentleman on the 'Net argued vehemently against the use of a chess database program as a postal recorder, opining that such usage takes much of the error factor out of the correspondence game (moves written down incorrectly, boards set up with the wrong position upon receiving an opponent's move, etc). He felt that these errors are "part of the game" and that storing a game in progress on a computer takes away the chance of these errors occurring.

This gentleman obviously has very little experience in game input on a computer. I can't tell you the number of times I was entering a game by hand into a database and made the move ...Nf6 instead of the actual move ...Nc6 (or made similar errors). These errors are hard to catch until you reach a later point in the gamescore in which some other piece tries to move to f6. Only then does the error become apparent. It would be amazingly easy to incorrectly enter a (legal) move from your opponent into your postal database or even write down a move other than what you intended to play when you fill out your postcard. Using a computer doesn't eliminate the possibility of these errors; in fact such usage creates different opportunities for similar errors.

The same considerations apply to storing your correspondence analysis in a database. I've heard complaints that such analysis can't be lost or destroyed the way analysis in a paper notebook can be. People who try to make these complaints have never heard of hard drive crashes or power outages. I once played a postal game in which I lost a day's reflection time simply because of the fact that all of my analysis was on my computer and we had a major thunderstorm the evening I received my opponent's card in the mail. I couldn't turn on my computer because of the storm, so I didn't have access to my notes and analysis.

Earlier, I listed five legal uses of computers in correspondence chess. Now let's talk about the illegalities.

It is a basic tenet of correspondence chess that players are to receive no help or advice from other players. The definition of "other players" definitely includes computer chess software. You can't open the mailbox, look at your opponent's move, feed Fritz the position, and have the program tell you what to play. It is absolutely illegal to have a computer aid you in this manner, just as it's illegal to ask the Expert or Master down at the chess club how to play the same position.

This brings up an interesting question I was once asked. If you search for the position in your database and come up with a single game, played between two computers or a computer and a human, is it all right to play the same move as the computer? In other words, if your game's position appears in your database as part of a Kasparov-Deep Blue game or a game from the 1995 World Microcomputer Championships, is it legal to play it?

Yes, absolutely. The gamescore is part of the public record of the chess event and is no different than referring to a game between two masters. To draw an analogy, it would be illegal to ask Anatoly Karpov how to play a particular position from the Caro-Kann Defense, but it's certainly legal to consult a book he wrote on the Caro-Kann and use any of his notes that you find there (predicated on the fact that such printed information is [at least theoretically] available to your opponent as well as to yourself).

In other words, it's a question of intent. You can't go to a computer program and ask for a move, but it's OK to use a pre-existing gamescore in which one of the participants was a computer. In the latter case, you're not seeking active assistance from a computer program.

The other popular method of "computer cheating" in correspondence play is "move checking". You decide what move to play in your postal game, make that move against a chess computer, and see how the computer replies in order to find out if your intended move is a mistake (in other words, "blunderchecking" the move to see if there's a tactical refutation to it).

This is extremely illegal and is looked upon in the same light as having a chess program generate a move for you. But you'd be surprised to learn that some postal players think that move-checking is an acceptible practice. While working at ChessBase full-time years ago, I personally spoke to six or eight correspondence players who told me they regularly had Fritz "blundercheck" their moves. I explained to each of them that this was illegal and my explanation was met with near-universal apathy, if not outright distain. Only one player (bless him) expressed any remorse over this practice; he'd honestly not known that he was breaking the rules against getting advice. He called his correspondence organization and asked that his rating be reset to "0" and explained what he'd done. No sanctions were placed against him, but they did wipe his rating and started him over as an unrated player.

As a correspondence player, how do you know that you're opponent isn't cheating? It's simple: you don't. So don't worry about it. Just play your normal game and have fun! The players who use playing programs to help them in postal play are surprisingly few and far between. Don't let the last paragraph dissuade you from enjoying correspondence chess. Yes, I spoke to a few players who used computers, but you have to remember that I spoke to computer-owning chessplayers all day long every day as part of my job, and the overwhelming majority of them who played correspondence were honest players.

One thing to keep in mind about computers is that they are great at short-term tactics, but terrible at long-range planning. The difference between computers and humans is even more pronounced in correspondence chess than in OTB chess -- but in favor of the human player. It's even possible for average players to outthink a computer at correspondence time controls. Correspondence Master Jon Edwards (one of the strongest postal players ever produced by the United States) once told me that he actually hopes that his opponents are using computers, as he feels that he can outplay any computer at correspondence time controls, due to the computer's lack of positional understanding..

We've set the parameters for computer usage in correspondence chess. We know what's legal and what's not. But we have yet to talk about the nuts and bolts of how to legally use computers to aid you in correspondence play. We'll look at plenty of those tips and techniques in the coming weeks here in ChessBase Workshop. Until then, have fun!

You can e-mail me with your comments on ChessBase Workshop. All responses will be read, and sending an e-mail to this address grants us permission to use it in a future column. No tech support questions, please.

© 2007, Steven A. Lopez. All rights reserved.