Difficult part about defense is evaluating when groups should be defended. Often there are sequences that are so-called "rightful sente" for the player that has them. For example, the opponent can rightfully peep a between one-point jump and the defender should connect. Pre-emptive defense here is not to connect those two stones before peep, but some other move altogether - if it's even needed.

From attacker's point of view, an attack is success when it forces opponent to live small and gives the attacker some kind of benefit (points of territory or thickness). But from defender's point of view preventing that successful attack is not necessarily good defense!

One very simple example of what I'm talking about:

Click Here To Show Diagram Code

[go]$$W White plays a rightful sente
$$ ---------------------------------------
$$ | . . . . 2 1 O . . . . . . . . . . . . |
$$ | X X X X X X O . . . . . . . . . . . . |
$$ | . O O O O O . O . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

Click Here To Show Diagram Code

[go]$$B Black plays poor 1 point gote
$$ ---------------------------------------
$$ | . . . . . 1 O . . . . . . . . . . . . |
$$ | X X X X X X O . . . . . . . . . . . . |
$$ | . O O O O O . O . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |[/go]

Defense is something that makes eye space or eye shape. In this purpose it prevents attack:

Click Here To Show Diagram Code

[go]$$B 2 defends, 3 attacks
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X X O . a . . 2 . 3 . . . . . . |
$$ | . . X , O . O . . , . . . . . X . . . |
$$ | . . . . . . . . . . b . . . . . . . . |
$$ | . . 1 . . . . . . . . . . . . . . . . |[/go]

In above diagram white should defend invasion of "a" by jumping at "b". Was this joseki choice poor for white, because he has to defend?

Not at all. Black's shape is not without problems, and white's shape is very strong afterwards. White might choose not to play like this way for very good reasons, but it is acceptable (in general) to play like this.

White also doesn't have to answer black if he doesn't want to, the black invasion at a has counters which, whilst losing a little territory, keep white connected and acceptably strong. It is a blow, but tenuki is not unthinkable.

Click Here To Show Diagram Code

[go]$$B Fine for W?
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X X O . . . . 2 . 3 . . . . . . |
$$ | . . X , O . O . . , . . . . . X . . . |
$$ | . . . . . . . . . . 4 . . . . . . . . |
$$ | . . 1 . . . . . . . . . . . . . . . . |[/go]

I like this position just fine for W. B will probably tennuki, but there's plenty of room for white to play on the right side. However, if you don't like that you could try playing this way:

Click Here To Show Diagram Code

[go]$$B approach first
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 3 1 2 . . . . . . . . . . . . . |
$$ | . . X , O . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 4 . . |[/go]

Now if B blocks the normal way your shape is a bit safer, as black is further away:

Click Here To Show Diagram Code

[go]$$B low knight jump
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 3 1 2 . . . . 8 . . 5 . . . . . |
$$ | . . X , O . 6 . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . 7 . . . . . . . . . . . . . 4 . . |[/go]

And if B goes the long knight's jump anyway, you can jump in to invade:

Click Here To Show Diagram Code

[go]$$B approach first
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . 3 1 2 . . . . . . 5 . . 8 6 . . |
$$ | . . X , O . . . . , . . . . . X 7 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . 4 . . |[/go]

I think either of these are good for W also, maybe even better than the original diagram.

EDIT: as always, my rank is my disclaimer. I think all go is very difficult, defending and attacking

I thought the idea of 4 in your second diagram ED was that if they played the large knight extension you could play the tight connection with 6, and still end up further away than they'd like:

Stable's take matches my own (not that I'm a good judge, but someone in my club relayed professional comments on the position a few years back).

_________________
Occupy Babel!