How good do chess engines play chess?

1. you mean by your saying, " score 6.5 does not mean win, it is just a mere evaluation" - You were the first to claim that 6.5 does mean white is going to win, I've asked you to provide some backup for your claim, you've provided nothing.
2. in post 203, you said," material difference is +8", the black pawn is about to promote as queen and white can not stop it by any tactical way except exchanging Queen which leads to equal position.How can you that +8 material difference? - if you check PGN you can see that it's white's move, white can prevent promotion, material difference will remain +8
3. If engine understands that the position [post 203 by you] is draw, why does not it understand the position is draw [post 166 by me]? do not say that ply does matter - engine doesn't understand that #203 is a draw, it merely gives it evaluation of +0.15, the rest is your interpretaion. Same goes for #166 - evaluation is +6.5, you interpret is as "engine thinks he is going to win", your interpretaion is wrong. When engine knows he is going to win the evaluation looks like "mate in N". And by the way ply does matter because we can't talk about some abstract evaluation, it's always evaluation for given ply, it's always evaluation of some future position after N moves.
   
4. you ignore the practical experince of score 1.00 or 5.00 - it's just that - practical experience. Yes in majority of positions 1.00 can be interpreted as slight advantage, 5.00 as decisive advantage, but majority doesn't mean all.

you are creating dramatic scene - like that

Earth64: Sun rises in the east

Uri65: No, you are wrong. Sun rises in the west.

Earth64: How?

Uri65: The direction you know as east, it is actually west.

The scene may be like that drama ---

Base: suppose,uri65 is trainer of earth64

Earth64: Mr uri, should i make Qe5 move or Qa4. There are no other valid moves except these two.

Uri65: What is your choice ?

Earth64: Qe5

Uri65: why?

Earth64: cause it scores 6.5

Uri65: It does not mean that you are going to win.It is just an evaluation.

Earth64: So lets make Qa4.

Uri64: How does it score?

Earth64: It scores 0.00. Lets make that move.

Uri65: no no, stop. Lets make Qe5.

Earth64: why?  score 6.5 does not mean it leads to win.

Uri65: It does not lead to win but it gives advantage.

Earth64: What will we do with advantage, if it does not lead to win.

Uri65: Advantage leads to win.

Earh64: ohho i get it. score 6.5 means it is an advantage. Advantage leads to win. That implies scoring 6.5  leads to win.

But why did you say at earlier time 'scoring 6.5 does not mean win'?

Uri65: Because i cut down  my several branches of common sense thought like stockfish while pruning for deep thinking.

Earth64: Can i attach common sense book with you?

Let me explain you some basics. Having advantage is not equal to having a forced win. Having advantage means having more chances for a win. Take an initial position in chess game. Does white have an advantage? Yes, just look at statistics - white wins in 37% of games, black in 27%. Stockfish agrees and evaluates it around +0.2 (depends on ply though). Does white have a forced win? NO.

DarkKnight, what's the engine score for that position? After all, White can sacrifice a Rook.

It's obviously drawn. Black can simply force draw by repetition.

And black can refuse it.

uri65,

Lets repair your basic. If it is not mate in N, it does not mean that it is not forced win. If one side has one rook surplus or two pawn surplus from the beginning, engine will never show that it is mate in N.

But we can say it on the basis of theoritical common sense that it is a forced win position. Moreover i will say it is one kind of mate in N position and engine can never be able to show it. After move by move the advantage will lead to Mate in N position which we and engine can not see for reasonable capacity problem.

Conclusion: Pure Advantage leads to win. Engine show the advantage with numerical value. So If engine shows score 6.5, it does refer to decisive [wining] position.

Dark knight,

u r amazingly funny.

In worst-case scenatio that's a draw by 50-move rule. Engine needs 100 ply to see it, I am not sure this is possible with actual software/hardware. May be some change in algorithm could make it detect that it's a draw?

Please define "pure advantage". How is it different from just advantage? Otherwise I have no idea what you are talking about.

How does your example of rook and 2 pawns advantage relates to your first example?

Your argument here and in post #199 is based on 2 assumtions:

1) that engine evaluation is equivalent to material advantage

2) that some material advantage (6.5 in this case) means forced win

Both assumptions are wrong - position in my post #203 is a counter-examples for both of them.

And I would like also to have your comment on initial position in chess -  white has advantage but no forced win.

If you do not want to understand and do not want to be convinced, then sleep.

I do not want to  waste your time as well as mine.

Goodbye then.

How can white sac a rook?

@Earth64: I have once escaped being a queen down (I had one, he had two) with a perpetual. Recently I have seen a game between two players in the Scotch tournament I am involved. White won, but he was for a very long time lost (from move 38 onwards):

In every game in which I might lose am I searching for imbalances that give me at least drawing chances. The most well known imbalance are unequal bishops, but effectively blocked pawns are another way. And I once was a rook down. There were still 8 pawns on my side of the board. You can make one guess how I got the first rook back. (In the end I won with a rook advantage. My opponent did not play very well to put it mildly as he was irritated that I did not give up after losing a rook.)

With perfect play on both sides will the side with the +6.5 evaluation win in the vast majority of positions, but not in all. Imagine all positions with a +6.5 or -6.5 evaluation. What percentage of stalemates exist? What percentage of positions is actually a win for the side with the worst position? They do exist.

Furthermore can you not expect perfect play from two players if the evaluation is +6.5 or -6.5. At least one of them has made a couple of bad moves.

Loek,

white missed a wining line.

If i am wrong, tell me.

R from b1 to b2 for example.

@Earth64: lol, I did not follow that game from the start. You are absolutely right that d6 is a directly losing move! Wow, how could both of them miss that?

Be7 is already a very bad move. Nf6 or Bc5 are the two main moves to challenge white. And if you play Be7 then should you play Nf6 or Bf6 as followup.

That only works if black takes