The problem with learning tactics: they're unrealistic!

depends on the time control. daily chess you have loads of time to think about moves. rapid you have a decent amount of time. bullet and blitz there's no time I agree there. that's why discovered attacks work really well in bullet.

I strongly encourage all my future opponents to " just sacrifice this piece and see what happens." I get to go home early, and with the full point!

Here's the thing: It's true that in most positions there is no immediate tactic. But tactical opportunities occur far more often than you would think, and they are game changers! so they are more than worth looking for. Mating your opponent, or winning substantial material gets you the full point, so why not spend a little extra time looking for those opportunities? 

IMO that's one benefit of working with books instead of online... there's no reward or punishment for guessing, so you're not tempted to just play a move to see what happens.

Before I give up on a puzzle from a book, I'd look at each of those puzzle-like moves, pretend it's a real game, and calculate why I think each of them doesn't work. Then when I look at the answer I can better understand my thinking process and what I saw in the position and what I missed in the position.

Going over your solution and the real solution is really important, and sometimes a lot more work than the initial attempt to solve the puzzle. Online you can do this too of course, but like you said I think a lot of people just guess and then move on to the next one.

Even if you never had a tactic in a real game, they would still be worth doing, because they help train your ability to calculate. And your ability to calculate is used just about every move outside of the opening, and some end-game positions that you have memorized.

Also, I think how you do the tactic matters too. Preferably you don't just move random pieces until something looks like it works. Instead, look for ideas in the position. Is there a weak diagonal leading to the King? Is there an overloaded piece? Is there a piece hanging if you can get another piece out of the way? Then start looking for moves that take advantage of that idea.

Again, that is much more like how you calculate (or should calculate) in a real game.

"The problem with learning tactics is this: you know that the tactic exists."

The point of learning tactics, is to understand the "why and how" of the position. 

Why is their a tactic?

How do i go about executing that tactic?

Take this position...

If we use the different strenght of chess machine, could the final result be so different ?

"

learning to analyse a position is an important skill that will help the student to improve his tactical awareness much faster.

      Overall, proper analysis of the position , the ability to read your opponent's moves and the ability to find candidate moves are skills that are more important for beginners , yet they are neglected because people believe that tactics will solve all their problems."

There's gold right here from Deirdreskye.

Tactics are how chess games are won and lost, it's as simple as that. Positional evaluation, pawn structures, strong and weak squares, these are important, but only if there is no tactic in the position! Strong players look for tactical opportunities--both for themselves and their opponents--on every move. In fact, the are also aware of tactical possibilities that don't yet exist in the position, but might if the opponent makes the right moves. So learn tactics!

The great chess writer C.J.S. Purdy once wrote "The man who is a fine positional player, but not a fine tactician is doomed to a life of disappointment."

Of course they're important--if there is no tactic in the position! If there is a mating combination, a knight fork that wins the queen, a double attack that snags that rook in the corner, then you can wave your copy of My System around all you like but the game is over! That's why strong players examine the forcing moves first. Most of the time, there is no decisive tactic available, then all the fine points of strategy become important, but it's always worth looking for the knockout punch, and knowing how to deliver it. 

Of course tactics are used to achieve positional domination too, but that's all the more reason to study them!

I agree with @laskersnephew. At least for sub 2000 players, tactics win or lose the game. Positional moves are natural moves for the tactician to prepare for the tactical blows. Or prevents opponent from delivering tactics. The tactical gain could be as minimal as a pawn or an advantageous endgame.

Look at the games of Alekhine, Fischer, or Kasparov. they outplay their  opponents positionally, but they often dispatch them with a tactical thunderbolt. And even when the game doesn't end in a tactical catastrophe, the threat of one often forces the opponent into a losing end game.