openings books or computer moves ?

I'm having a hard time believing this question is coming from a 2000+ rated player. 

... i did that many times and the computer keeps getting to a wining positions !! , i will choose the move i know if i were playing a human but am just intersting in the computer choise ...

... yes they will after 22 to 28 moves the computer without the opening database keeps winnig :\ ...

The computer has no absolute chess knowledge, it just calculates to a certain depth and the result is depending on the used evaluation function. The computer is not always right. I am assuming you are genuinly interested in the difference between the evaluation in opening books (which is probably based on a mix of human experience, statistics, engine supported analysis, ...) and the evaluation in a chess engine (math function).

Perhaps it is a young player really interested in this question :-)

A proof that computers aren't always right:

Please let me be specific. I think on endgames engine purely just adds the material like on theoretical drawish positions when white has 2 pawns up and says +2. My friend interpreted that that engine says it's win for white but I don't think it is even saying that, but only that there is 2 pawn advantage. On middlegame with equal material +2 would be winning position. But on endgame at least the clear sign of winning position is more plus on those kind of positions because something like +5 means there will be new white queen on the board. At least I interpret +2 as something that "Engine doesn't know."

Even if I would be wrong I think it's clear that +1 means different things from engine evaluations on different positions. The human have to interpret what engine means.

On another position there was 1 B on both sides and white had 1 pawn. On that case anything from +3 or less is not win position, because that +3 would mean K+B vs K which is draw. Only evalatiaion that says win for white is +6 (or more)because it means Q vs B.