The problem with the crafty program when it comes to analasys is that it doesn't analyze openings that are in its book. Therefore I had to use another program, der bringer 1.9 (available on this site) to do it for me. After about 7 and a half hours and analyzing 8,360,684,392 positions after d4, it concluded that the best counter move is d4-d5, Nf3-e6, e3-Nf6, c4.... ( I couldn't see after this). Either way, it was strange since most people respond with nf6.

Well, computers say a lot of things. You'll find computers aren't always right. They have their strengths and their weaknesses. The problem is, a human goes into a game with the knowledge of what they want to achieve. A good chess player knows what position they want to put their opponent in...and chooses an opening that is best suitable.

 

A computer doesn't have this concept of desire, or even goal. They only calculate out the potential likely moves for a position and make a bases off 'whichever result gives them a higher point total.' That point total might not be a better position, it might not even be useful. And when you're calculating something so complex as an opening, odds are the computer will miss a lot of fundemental moves. Even a computer can't calculate everything.

 

So, that responce is probably one way to do it. But that doesn't mean it's the 'best way.' It's a possible option. But instead of worrying about what a computer says, find an opening you like. Because unless you can play like a computer (and know exactly why it prefers a certain opening), knowing a computer's favorite book isn't going to help you out at all. Using it would probably hinder you.

1...d5 is as good as 1...Nf6 so don't be surprised. Anyway, don't use computers to evaluate an opening, try using game databases instead. Computers just don't understand the opening principles and ideas.

I know, but it seems to have some value since they do play better than I do by a considerable amount :)

But that's because every single move they make is being analysed by the same database. If you take the opening 5 moves of a grandmaster, you're not going to use them wisely unless you know WHY the 5 moves are good :P

 

Basically, it would do you better to play 5 sub-par moves that you understand and can fully defend, than jump into a position you can't defend. And plus, half of learning is learning why. Using a database to come up with an opening because it says it's the best opening...won't help you.

Engines get stronger as the game progresses (as a rule)... In the opening they can be easily outplayed by even an average player provided they aren't using a book.

 

Anti-Engine chess rules tend to state that the first move out of the book is the best place to beat an engine... the reason being that they haven't formulated a 'plan' at that point. 

A couple of points. First, to reinforce some of the earlier points, computers suck at openings. This is generally acknowledged. Oh sure, they're great when they have opening books programmed into them; but remember that positions in the opening phase of the game tend to have the most possible moves and therefore require the most processing time. The engine, as you saw, will get bogged down until things clear up a bit.

     Second, consider what's going on here. You've gotten the computer to look just three and a half moves (seven half moves, or "ply" in computer chess jargon) into the future. Opening theory for pretty much any established line can go ten, fifteen, even twenty moves or more deep. This is the fruit of hundreds of years of labor by mankind, covering many millions of games and probably billions of analysis-hours. In short, the reason 1...Nf6 is played so much is that history has proven its value (though not, of course, its definite superiority to 1...d5).

     Third, according to the databases, the statistics for this opening pattern (1. d4 d5 2. Nf3 e6 3. e3 Nf6 4. c4) are not encouraging. The ChessLab database (Two million+ games at http://www.chesslab.com/PositionSearch.html) shows that 29% of the games from this position were wins for white, 30% were wins for black, and 41% were draws.

     In those seven and a half hours you spent waiting for "the answer" from your computerized oracle, you could have strolled down to your local bookseller, picked up almost any beginner's opening manual (assuming your bookstores have a chess section!), and read about the different kinds of games that can arise from 1...d5 versus 1...Nf6 (or even 1...f5, for that matter). This is the knowledge that will help you succeed: learn the meaning, not the moves.

DimKnight wrote:

A couple of points. First, to reinforce some of the earlier points, computers suck at openings. This is generally acknowledged. Oh sure, they're great when they have opening books programmed into them; but remember that positions in the opening phase of the game tend to have the most possible moves and therefore require the most processing time. The engine, as you saw, will get bogged down until things clear up a bit.

     Second, consider what's going on here. You've gotten the computer to look just three and a half moves (seven half moves, or "ply" in computer chess jargon) into the future. Opening theory for pretty much any established line can go ten, fifteen, even twenty moves or more deep. This is the fruit of hundreds of years of labor by mankind, covering many millions of games and probably billions of analysis-hours. In short, the reason 1...Nf6 is played so much is that history has proven its value (though not, of course, its definite superiority to 1...d5).

     Third, according to the databases, the statistics for this opening pattern (1. d4 d5 2. Nf3 e6 3. e3 Nf6 4. c4) are not encouraging. The ChessLab database (Two million+ games at http://www.chesslab.com/PositionSearch.html) shows that 29% of the games from this position were wins for white, 30% were wins for black, and 41% were draws.

     In those seven and a half hours you spent waiting for "the answer" from your computerized oracle, you could have strolled down to your local bookseller, picked up almost any beginner's opening manual (assuming your bookstores have a chess section!), and read about the different kinds of games that can arise from 1...d5 versus 1...Nf6 (or even 1...f5, for that matter). This is the knowledge that will help you succeed: learn the meaning, not the moves.

---

I see your point