When playing correspondence chess (traditional CC, no chess engine assistance), where is the line between what is and is not acceptable when using databases? For example, let’s say a game is five to seven moves deep on a fairly standard opening line. By consulting a chess database, or even printed chess books, I can find examples of games that have had the exact opening or a similar opening that results in the same position. What is done next with that information is where things can get grey.
So, let me ask some hypothetical questions, of increasing dubiousness:
Is it legal and ethical to consult printed chess books?
Is it legal and ethical to consult a purchased chess database with millions of games to find similar positions?
Is it legal and ethical to consult a purchased chess database that contains an ANNOTATED game with a similar or identical position?
Is it legal and ethical to play out more of the opening line found in the exemplar games mentioned above?
Is it legal and ethical to play out more of the opening lines found in the chess databases mentioned above to maximize the percent chance of success?
Is it legal and ethical to use a high-end chess database system to run a deep tactical analysis on a recorded chess game and use that analysis to influence your strategy in your own similar game?
I’m very interested in your opinions. Please note I’m asking what you feel is legal and ethical, not what is tactically wise. For example, someone who relies on #5 above probably isn’t going to do too well once the game progresses.
Using a database is not that much different from using books that compile games that you can read and explore. Using an analytical engine from a program running on a fast computer or a dedicated device rather than your own brain is a different story. Reading and doing your own thinking is okay. Getting help from a machine or others is not. Of course, different correspondence organizations have their own rules on these things.
My basic principle was always that if it’s in a book it’s fair game, and thus anything analogous to that on the computer is also fair game. If the machine does one iota of calculation, it’s not.
But then, I gave up correspondence chess before tablebases. Is a tablebase analogous to Basic Chess Endings?
Positions of 7 pieces or less have been solved by computers and are thus readily available on endgame databases-databases of regular games include computer recommended moves and computer evaluations. GM’s writing opening books today consult heavily with the engines-like it or not computer engines are a part of correspondence chess either directly or indirectly.
I don’t know if this still hold’s true or even if it did, the future is unclear: A correspondence GM once said, a few years ago, that even with chess engine and all the books and databases you have at your finger tips, there is still room for some human element.
I don’t know if that still hold’s true or not. With Intel making a 2 qbit chip, it’s only a matter of time before quantum computers are in your house.
Far as I can tell, the 2 qbit quantum chip is the only one made, so as far as desktop quantum computers, it’s a few years down the road. I don’t know how many qbits one would need in a desktop to make it better than a regular cpu.
My question boiled down to how one uses tablebases. If I have a 7 piece endgame, for instance, if I would have to punch in my position and have the computer show best play, that wouldn’t be right. If I just punch in the specific 7 pieces on that exist and it shows me how to win from a position with those 7 pieces, that’s more like looking it up in BCE and ought to be permissible.
I do not understand the distinction. On both the 6 and 7 piece endgame databases, all possible positions have been analysed.So you set up the position on the databases chess board and hit the “evaluate” ( 6 piece database) or “probe” (7 piece database) button.
Bingo if there is a mate in say 90 moves-it will list all 90 moves-exactly like BCE would show you how to win say the Lucena position. If there is no mate ,it will simply give 1 move and say its a drawn position.( unlike BCE which would show a bunch of moves analyzing why it is drawn) ICCF recognizes the 6 piece database as a final determinate to a game but interestingly not the 7 piece.
Not sure why that is.Hopefully this answers your question but to me it is essentially semantics.
i don’t see a difference between looking up an opening or using endgame tablebases. At some point, at least some correspondence organizations decided it was impossible to enforce not using engine help. In the same line of thinking, opening book and endgame tablebases are just part of what it is to use and engine.
In any event, even a few years ago, it was all but proven that a chess engine and no tablebase is against a chess engine using an endgame tablebase is statistically even as far as performance of getting a win or draw. Of course, there are positions in which the use of a 7 game tablebase would make the difference between a win or draw or losing, but as a percentage of possible endgame positions, it’s negligable.
Of course it’s not hard for a talented human to come up with a position that would give a chess engine fits comparing to an engine using a tablebase, but how often that comes up in normal games would be minimal.
I like to think of it like this: My favorite number, never calculated is Graham’s Number. A massive number no mortal or quantum computer could ever calculate. Yet, as big as that number is, it’s effectively zero compared to Tree(3).
So there are probably a massive number of those positions that would be played better with the help of an endgame tablebase, but compared to the total number of actual possible positions, it’s effectively zero for normal play, at least comparing it to the performance of a high end chess engine.
I don’t want to get into Graham’s Number but to compare it to Tree(3), Graham’s Number would be 3^^^^3G64. ^ is not an exponent, but Knuth’s up arrow notation.
You have to use a different system to describe Tree(3), but in that system, Graham’s Number would be k64 and Tree(3) k(n) with (n) to the power of 65,000 rounding it out.
Some correspondence chess organizations (ICCF, for example) have an actual rule that allows a game to be claimed as a win, loss, or draw based on the 6-piece endgame tablebase.
As the years pass by, I wonder just how much human element is left in correspondence chess. I guess with Intel figuring out how to put a quantum computer on a chip, it’s only time before it eventually gets into people’s living rooms. At that point, what’s left for a human to ponder, even in correspondence chess, that a quantum computer would already see in it’s horizon?
I think it’s safe to say some of the highest quality games of chess is from the world of correspondence chess, but perhaps someday in the future, over 50% of the games in Chessbase’s Corr Database will have come from machines and no human interaction.
Maybe at that point, the future of correspondence chess is with prisoners and the person on the outside will sans all help other than good old fashion paper & ink books, to give the prisoner a chance to actually win games.
Chessbase is already making opening books primarily from games in the engine-to-engine room. Although they use some human games, it’s probably more of a marketing ploy, because they’d sell less of those opening books if people felt they only used engine-to-engine games to populate the opening book.
In fact, if you’re not doing 1, 2, and 3, you’re probably losing. And you should be. Correspondence chess is a research exercise.
In 4 and 5, I don’t know what you mean by “play out”, so I can’t judge. In 6, it probably depends on how close your live position is to the position where you start analysis.
All this presumes a rule set in which engine analysis is prohibited. In ICCF, engine analysis is permitted, and all that you describe is certainly fair game as long as “play out” doesn’t involve analysis with carbon-based intelligence. That doesn’t make it any less of a contest; blindly accepting engine analysis or failing to push an engine beyond its limits is a sure way to lose.
I found ICCF’s policy out halfway through my first ICCF tournament. It was disappointing at the time, but there’s still life and enjoyment in the game if you approach the task with an open mind.
