It would be an unfair match. Imagine every single chess player, engine, and chess theory book ever all together playing a 3000 rated player with no opening knowledge. The first would get an advantage from the opening and convert it with high consistency.
Super GM versus Computer without Opening Book
Sort:
Are there any examples of this?
Hypocrism/Mephisto- So you are saying that the Super GM would completely crush the chess engine? I agree with you that the engine has no opening knowledge but how much of a advantage can be intiated from the opening? The engine has no opening knowledge, but it would (I think) still play to develop pieces and I don't think it would do anything that could hurt itself so badly. Of course the player would gain an opening advantage but how big would it be, and would it be enough?
ES_Lowe- I think it would be quite interesting to see results on this. It can really show the difference between general, just developing concept for the opening, compared to deep theory and understanding of the openings.
Mephisto wrote:
I am just saying that the super gm SHOULD acheive an advantageous position ,whether winning or not depends on the opening lines choosen and the level of participants .
Sorry, I wasn't really targetting you in the other post. I actually was replying to Hypocrism, and added you because you had the same idea of the GM having an advantage. It's just that he implied "would" and you "should". Anyways back to the topic, I guess the opening lines could also make a big difference. Maybe something with a lot of theory, such as the Grunfeld could go in big favor for the GM. I think this would also be a quite interesting study. Like which openings would have better results. Also could you play dubious gambits like the Lativian or Halloween, with good results?. How bout the Danish gambit?
FirebrandX wrote:
Mephisto wrote:
Super gm should win. Advantage in the opening cannot be obtained by intuition alone, it requires study . The engine will get an inferior position without the opening book and the super gm should be able to convert his advantage.
This used to be the case, but after the advent of multi-core cpu's and deep engines like Rybka, disabling opening book will merely cause the engine to come up with a borish choice of play that is both solid and drawish. For example as black, the computer will never pick the sicilian against e4. Most of the time it will evaluate either e5 or e6 as best, both leading to solid games for black. Even if the GM were to get a slighty better position, the engine's unrelenting and untiring defense will slowly frustrate the GM into making minor concessions that equalize the game at best, and cause an outright loss at worse. I've seen both happen.
So it has been tested. What happens if you sort of force the first 1-4 moves for the computer and make him play from there? (Not saying giving bad moves, but doing introductory moves of for example, the Sicilian and others). Would it go back to as Mephisto said an opening advantage that the GM could possibly/likely convert to a win?
adriaan1 wrote:
Hypocrism wrote:
It would be an unfair match. Imagine every single chess player, engine, and chess theory book ever all together playing a 3000 rated player with no opening knowledge. The first would get an advantage from the opening and convert it with high consistency.
but the computer can calculate 1 million possible outcomes of a move in +-3secs
adriaan1 wrote:
adriaan1 wrote:
Hypocrism wrote:
It would be an unfair match. Imagine every single chess player, engine, and chess theory book ever all together playing a 3000 rated player with no opening knowledge. The first would get an advantage from the opening and convert it with high consistency.
but the computer can calculate 1 million possible outcomes of a move in +-3secs
I stand by the idea that so many years of chess and so many genius improvements to opening theory will overcome the chess engine in the opening, even if the GM cannot solidify that advantage. Computer engines are good at overestimating the value of material and underestimating the value of the initiative.
From what I've read, the strongest chess engine, Rybka, can beat the best players in the world. But I've also heard that they rely on opening books, and without it, the engines don't really play the best starting moves. So how much weaker would the computer be without it. And would GMs have much better results against the strongest engine? Or would there really not be so big of a difference.