Gilchrist is a legend wrote on 05/28/12 at 20:21:01:


Yes, that is the sitiuation we are discussing, and of course players will prepare for each position, but...

Gilchrist is a legend wrote on 05/28/12 at 20:21:01:


...good preparation requires good background knowledge, as I tried to illustrate with my KID example. And you will never have as much background knowledge of one single 960 position as you have about the "normal" starting position. Even if there would some day be as much theory about each position as there is about the "normal" starting position today, no strong player will ever study every single starting position in chess960 as thoroughly as he would study the only starting position we have today.

That is my stand, at least. But since there is no empirical evidence, and there won't be any until this game has been standard for a century, we probably just have to agree to disagree.

If I remember correctly, there was a post referring to having a Chess960-styled tournament, except only a limited number of starting positions with all players knowing which positions will be chosen for the games. I was referring to this hypothetical situation, and in this specific occurrence, of course players will prepare for each chosen position that the tournament arbiters or organisers have chosen. Eventually if more tournaments were created in this style, then eventually each position will have their own theory. Each tournament where the organisers have a limit, for example 10, positions and notify the players which positions will be chosen and when, then eventually, for example if there are 1000 tournaments like this or more, eventually all 960 will have their own theory.

The analogy with mathematics was not to link memorising formulæ with playing positions after theory, I was linking it to opening theory when preparation involves only memorisation of moves. Of course GMs forget memorisation of their moves, but many memorise their moves for their games. If they can memorise moves like a robot enough to play them on the board, it is the same as memorising like a robot mathematic formulæ. In Chess960, eventually I would think, perhaps in decades, that GMs will eventually expend extra effort to memorise all 10 positions if 10 positions were given to them prior to playing. Preparation is 10 times more than regular chess, but they will do what they must to succeed in the tournament, regardless if it signifies 10 times more effort and 10 times more time allocated to memorising theory for all 10 positions. Playing out theory without thinking because one memorised it an hour before the game is quite similar to writing down a formula without thinking because one memorised it an hour ago.

I find that highly unlikely. Even with today's one and only starting position, strong GMs sometimes forget. With many different starting positions, it will be much harder, and you cannot devote as much time to every starting position as you can to the one starting position we use today.

I am not buying your analogy to mathematics, since in mathematics, there is a definite answer to every question, while in chess, there is not. Well, theoretically I guess there is, since every position is a win, draw or loss, but luckily, humans are not strong enough to evaluate every position that way. For the same reason, your analogy in one of your previous posts,

Gilchrist is a legend wrote on 05/26/12 at 01:08:54:


is only relevant to me if the questions are so difficult that a lifetime of study will not be enough to answer them properly.

Tullius wrote on 05/27/12 at 12:42:25:


True. But if I have understood correctly, the purpose of chess960 is to avoid the fact that people can prepare 20+ moves of concrete opening lines. If people then start preparing for patterns rather than concrete lines, it means that the purpose is fulfilled.

Tullius wrote on 05/27/12 at 12:42:25:


100 % agreed. I see that my point wasn't expressed very clearly; allow me to reformulate it:

If I am playing the KID (which I am not, but it was the first example I could think of) and I see that my opponent usually plays the 13.Rc1 classical, I will focus my preparations on the 13.Rc1 line. Of course, my opponent may choose to deviate earlier on, but since I have studied the KID for some time, I know something about the other lines as well. Not everything, but something. Therefore, I prepare for the most probable line and rely on my knowledge in the other lines. Sure, understanding the arising positions is vital to success, but that understanding is, as you say, not learnt in the last two hours before the game.

In chess960, I will never have that much background knowledge of a concrete opening, and neither will my opponent. Therefore, preparing for a concrete position arising after 13 moves seems much less to the point.

Gilchrist is a legend wrote on 05/26/12 at 20:35:54:


But your educated guess will be based on a very reduced number of games, since your opponent will only have played a few games with that exact starting position.

Besides, you would know a lot less opening theory about each starting position than you would know about the one starting position every chess game starts with now. When preparing for a game in today's chess, you can look over your opponent's games and immediately recognize: "Oh, he is playing the 13.Rc1 line in the classical KID." You won't have to re-study the first 12 moves, since you know some ways to deal with earlier deviations. There is just no way you can learn every possible 960 position in the same detail as the "normal" starting position, and that will make your preparation harder.

Smyslov_Fan wrote on 05/26/12 at 20:56:06:


Who am I to argue against Kramnik, but I disagree that the rest of the game will not be substantially changed. In a lot of games, it is a key point that in a structure like a2, b3 vs b4, the a2-pawn is weak, and White is not able to create a passed pawn without help from his king or his other pieces. Without the en passant capture, White can simply change the whole picture by playing a4.

Kramnik once proposed removing en passant as a legal move.  He said that would substantially change opening theory for some key lines without substantially changing the rest of the game.

maybe invent something else than castling?

also we could add a new piece like the canon from chinese chess, which must jump over another piece to to take it ...
just one single new piece could transport chess on a whole new sphere, maybe replace the rooks by chinese canons =)