thanks for helping.
I thought it before posting and ended up with the same conclusion. But I'm not sure if both of us are correct.
(anyone else wanna join?)
thanks for helping.
I thought it before posting and ended up with the same conclusion. But I'm not sure if both of us are correct.
(anyone else wanna join?)
In my opinion, I think that engines have plans, but not THAT kind of plan. Their "plans" is just best moves stacked on each other, and our plans are steps to make an attack or make a defense.
But still I still have the same opinion, just another way to tell the difference.
Don't engines use a database of winning games to make their opening and middle game moves? That is kind of a plan.
So can they make a plan for playing?
No, hell no, no way, and not even close. I speak as a computer programmer.
We know that engines analysis positions. And they just analysis one position at a time. So can they make a plan for playing?
Not the way humans do. Engines look at moves, consider all possible responses, then choose the move/line that results in the highest evaluation score.
Some positional factors go into an engine's analysis but they're very basic compared to how a human analyzes a given position. Engines will give you raw output of variations 20+ moves long and - in a sense - that is how they "plan." They cannot yet assess a position the way we do and give you a paragraph on why one side is better or why it's correct to use an abstract idea like color complexes to force a winning position. I think that one day, they will be able to.. but not yet. However, reading between the lines of those variations that engines give you, you can often see a "planned" series of moves that compare with moves that top players consider and make themselves; the difference is the method by which the two arrive at their desicions. Humans rely on logic and a generally small tree of calculations, engines on raw calculations with primitive positional analysis programmed by a human.
I agree.
A engine calculates, but they don't plan. They just calculate for the best moves. We calculate for plans.
Don't engines use a database of winning games to make their opening and middle game moves? That is kind of a plan.
Esteemed chessmate;
Certainly they may use opening database and endgame tablebase, but, already, it is not a plan by itself, it is a programmed option that may help the engine with its work with already existent endgame and opening knowledge.
There's no strategical or positional thought on it. 😜
Sure it is. Isn't that what the pros do when they prepare for a match? They know the openings to be played for the most part and memorize the best lines before using strategy to plan their moves don't they?
Engines don't make a "plan" per se, instead they play the game by estimating positions on a move by move basis. More advanced engines retain the already analyzed lines, as I understand, so you could say they have countless plans in their silicon brains on what to do in response to opponent's possible moves, slightly changing the plans as their assessment of the positions deepens.
Don't engines use a database of winning games to make their opening and middle game moves? That is kind of a plan.
Esteemed chessmate;
Certainly they may use opening database and endgame tablebase, but, already, it is not a plan by itself, it is a programmed option that may help the engine with its work with already existent endgame and opening knowledge.
There's no strategical or positional thought on it. 😜
Sure it is. Isn't that what the pros do when they prepare for a match? They know the openings to be played for the most part and memorize the best lines before using strategy to plan their moves don't they?
It isn't real plans.
So can they make a plan for playing?
No, hell no, no way, and not even close.
Is the short and sweet answer.
No.
Is the even shorter answer.
lol.
Computers go through a permutation of move and decides the best based upon which it evaluation function and a mini-max algorithm. The computer can store what it thinks is the best line.
Now the problem is what do you consider a plan. If my plan is to play Nc7+ and take the rook next move, then the computer can plan that. If my plan is to pawn storm due to opposite side castling, there are functions which suggest this to the computer, and the storing of move is its plan.
The other side is most RTS. The basic plan is to gather resources, build an army, make defensive structures (of both buildings and men), and attacking. The computer doesn't really develop a plan, but follows a predetermined set of instructions. This implies no intelligence and, therefore, no plan.
For the final thought argument, consider the human brain. We follow the same guidelines the computers are given, and we often come to the same conclusion. Do we make a plan just like a computer would? We are able to change our plan during a game, but this may be just using different guidelines, which equals that we are worse at making a plan due to our limited RAM.
So can they make a plan for playing?
No, hell no, no way, and not even close.
Is the short and sweet answer.
No.
Is the even shorter answer.
It's often the commonality of two things that is important. I detect a certain commonality in these two responses. ![]()
Is the short and sweet answer.
No.
Is the even shorter answer.
It's often the commonality of two things that is important. I detect a certain commonality in these two responses.
[QUESTION DELETED]
So why do you think this answer?
Is the short and sweet answer.
No.
Is the even shorter answer.
It's often the commonality of two things that is important. I detect a certain commonality in these two responses.
[QUESTION DELETED]
So why do you think this answer?
For example, one of the first things an engine will do is list all legal moves. It then calculates a line for each legal move, no matter how obviously terrible the move is. As it calculates, it can't keep considering every legal move (well it could, but that would be too slow) so the programming tells it when to ignore certain lines.
By its nature a plan (at least in part) will start from the general and from there work out the specifics. For example my plan might be to attack on the kingside by using the dark squares, and the specifics would be "I'll play Nf6." An engine only does specifics, 1 move at a time, and anything resembling a logical sequence of moves (like an attack) is unnecessary and unintended. In fact often an engine doesn't play "logical" moves (at least what a human would call logical). As Carlsen said playing the engine is like playing an idiot (but the idiot always wins). It doesn't win due to superior planning, it wins because humans make too many short term errors. The engine makes small long term errors, but in a practical game that doesn't matter.
so I got your idea. do u mean that the couple of the engine's moves might be a plan or else?
Is the short and sweet answer.
No.
Is the even shorter answer.
It's often the commonality of two things that is important. I detect a certain commonality in these two responses.
[QUESTION DELETED]
So why do you think this answer?
For example, one of the first things an engine will do is list all legal moves. It then calculates a line for each legal move, no matter how obviously terrible the move is. As it calculates, it can't keep considering every legal move (well it could, but that would be too slow) so the programming tells it when to ignore certain lines.
By its nature a plan (at least in part) will start from the general and from there work out the specifics. For example my plan might be to attack on the kingside by using the dark squares, and the specifics would be "I'll play Nf6." An engine only does specifics, 1 move at a time, and anything resembling a logical sequence of moves (like an attack) is unnecessary and unintended. In fact often an engine doesn't play "logical" moves (at least what a human would call logical). As Carlsen said playing the engine is like playing an idiot (but the idiot always wins). It doesn't win due to superior planning, it wins because humans make too many short term errors. The engine makes small long term errors, but in a practical game that doesn't matter.
so I got your idea. do u mean that the couple of the engine's moves might be a plan or else?
Seems English may not be your first language.
Get the guy who made this topic to translate for you:
https://www.chess.com/forum/view/general/come-on-this-is-a-really-big-bug-talking-thread
ouch.
----------
1. Got it.
So you mean that engines only make specifies, but their plans have no logical. And we make logical moves.
But we both make plans, so your answer refer to both answers.
2. Please look at the first post again.
We know that engines analysis positions. And they just analysis one position at a time. So can they make a plan for playing?
(I mean like an level 3 computer in chess.com)