No not prohibited. It's specific, but so is a book on the opening. It's still not adaptive. Your opponent may not follow the lines in the book.

This is why correspondence is retarded, don't play it

A - ha! I now see Batgirl's citation in #62 - so books FOR OPENING MOVES are ok. 

I would strongly suppose that works ON THE ENDGAME would be NOT OK during the game - only after it, to aid in post-mortem analysis. The reason being, that it can easily get SPECIFIC. 

The FAQ's are poor, inspecific and out of date. The 'for opening moves' line is misleading and should be removed. It also doesn't say for opening moves only.

There is no specific definition of opening, middlegame or endgame moves. All static sources are OK other than engine analysed opening databases, this has been covered in the cheating forum in the past.

There has been extensive discussion in the past here:

http://www.chess.com/groups/forumview/evolution-of-the-site-rules EDIT: Useless link as staff never gave a proper answer.

Being a student of computers, I am not sure that computers are adaptive.

All they can do is follow pre-programmed sequences.

Adaptive would mean the computer, or its program, can change without an external input.

Surely there must be a better word here!

barefoot_player

A computer running chess engine software can definitely adapt to your moves.  They're really good at it actually.

That's the problem here. Computers are incapable of adaptation by themselves.

They don't change or improve with experience or age since they have no experience and they are not wine either. ;)

They can't reproduce so they can't change through evolution.

The can only follow pre-programmed instructions.

Very unadaptive.

Which is why I am looking for another word here.

barefoot_player

You are right batgirl and long. It is an automation that uses a power source.

But I still wonder if using this device, which prototypes have actually been built, can be considered cheating if used.

Surely it is not an engine, we all agree with that. Are mechanical tools illegal then?

A student of computers would know that this:

All they can do is follow pre-programmed sequences.

...is not really correct.  A chess engine creates a set of parameters of play and valuations for what constitutes "good" play.  Using that completely open framework, it instantiates a game and then plays against you better than any human being in history.  The only "pre-programmed sequences" a chess engine uses during play itself is the opening book it has ;). 

Your statement applies only to the earliest days of computing when computers were just glorified adding machines.  Do you imagine that the video games of today work by specifying out in detail the infinite moves a player can make?  They work by creating an adaptive construct with rules about how to react to anything the player does within the constructed environment.  That is adaptive.

Chess engines also "change themselves" on the fly.  You can see a perfect example in the current TCEC engine championship between Stockfish and Komodo.  Stockfish was winning +4 -0 =18 after 22 games before Komodo's accumulated knowledge of Stockfish's previous games kicked in and allowed to Komodo to start winning games.  

Komodo is "learning" to play Stockfish better every game as it goes along.  Stockfish is also changing, of course, and has maintained it's lead overall, but is not longer crushing Komodo as it was initially.

Whether a chess engine evolves or not, does not matter.  My copy of Fritz from ten years ago can still crush me.  But machines can learn -- check this out:

http://en.wikipedia.org/wiki/Machine_learning

Machine learning, a branch of artificial intelligence, concerns the construction and study of systems that can learn from data.

------------------

http://en.wikipedia.org/wiki/Genetic_algorithm

Genetic algorithm

In the computer science field of artificial intelligence, genetic algorithm (GA) is a search heuristic that mimics the process of natural selection. This heuristic (also sometimes called a metaheuristic) is routinely used to generate useful solutions to optimization and search problems.^[1] Genetic algorithms belong to the larger class of evolutionary algorithms (EA), which generate solutions to optimization problems using techniques inspired by natural evolution, such as inheritance, mutation, selection, and crossover.

Genetic algorithms find application in bioinformatics, phylogenetics, computational science, engineering, economics, chemistry, manufacturing, mathematics, physics,

This is interesting.

Do these computer things now change their programming as they go on playing? Does this new knowledge keep with them after the machines are turned off?

Can a copy of this program be transplanted into another computer, say one with a larger memory?

I know they were doing  experimental research into adaptive programming, but now you are saying these type of computers are used in applications?

Wow!

Whether the computer keeps the games played or discards them is up to the developer(s), but in the case of the top engines in their "public" downloadable states, I would say it discards them and goes back to it's "default" state between sessions.  After all, the engine can already kick the behind of any player using it, it would be like rubbing salt in the wounds to also keep your games and play even better specifically against you ;).

But in theory, any engine could store all games it plays in a separate opening book/database and refer to them going forward, pretty much like a human's memory, except a lot more exact.

I would be more impressed if the computer was to actually change its own programming and keep the new knowledge (like how to deal with an isolated, yet advanced pawn), in future encounters, even after it's power is turned off.

Nevver the less, I am always surprised with computer technologies!

barefoor_player

That is coming.  It won't be many years before the best human players can no longer explain how the best engines choose their moves anymore.  Today we have GMs that can see after the fact why an engine move is best, once the engine has found it, but someday, we'll even be past that point and teams of GMs will be floundering to figure out why a certain move was chosen over another in engine vs. engine games.

The initial set of valuations in a given position (like the isolated pawn example) has evolved over time, just like an egine does not use the pawn = 1, knight/bishop = 3, rook = 5 valuations anymore.  The engine may value a rook at 4.98, for example.

The more games that accumulate and can be analyzed, the better those numbers become.  Humans are still tweaking them for now, but if someone put a server up running Stockfish on some significant hardware the same way chess.com runs Computer-4 Impossible, it could play millions of games and refine those numbers even further, at which point instead of a human saying "let's try tweaking the valuation of a rook from 4.987 to 4.983 and see if the engine's play improves", the engine will tweak itself, and the developer will come back to discover "oh, the engine has decided that 4.9863125 is the best valuation".

P.S.  Those are also only "base" valuations in the starting position.  Valuations are already dynamic based on specific game positions.  An engine will value knights more in a closed position, for example, and value bishops more in an open one.

So all video viewing is illegal at all times in chess? If I watch a video then later play a game over the board I'm cheating? Gee Chess.com better take down all of their videos. They're making cheaters of us all.

LOL!

Let rephrase that!

A computer can not change its program by itself. It needs someone or something else to change its programming.

Spirit,

That was in reply to a question that has to with using a video or book while the correspondence game was in progress.

In fact, I encourage using a book, a video, a lecture, a friendly GM, as much as you like before a game starts.

It is using such resources during a game that his thread is about, what is allowed and what is not allowed.

barefoot_player  

I don't suppose that an opening database which incorporates engine evals of the positions should be forbidden. It's like a book and isn't adaptive. In addition, it is available to one and all. 

I saw no rule specifically prohibiting its use, in addition. 

If I have a K + P vs. K advantage, and I don't know anything about opposition nor trianglation, can I use an endgame book that just happen to have the K + P ending I am playing?

Ok - this has got to be one of the weirdest questions I've asked anywhere.

If so, how is using my marble machine different from consulting an endgame book with the same position?

Assuming that my opponent has the same access to this marble machine (it's available on an Internet site). They would both produce the same errorless result with absolutely with me not doing anything at all. Why would using my marble machine be considered cheating, and not blindly using a book (which is hard thing to do if you are blind, but you get my point! ;)

Maybe using both are illegal?

barefoot_player

Wouldn't it be worthwhile if chess.com gave explicit and detailed rules, in an obvious place, on what this site considers cheating in turn-based games,  especially since cheating is punishable by termination of accounts?