Arnulf wrote : I do not quite agree with tlaero and her rules, because such a game is individual everytime and the game-creator should have his free decisions, how to make the game-story. Nevertheless it is not easy to understand all of the single steps and why they are working and why not. But believe me, in short time the main points will have been cleared and then for the short-thinkers there will exist something like a walkthrouguh!
Even when following these 'rules' the creator still has plenty of space for free decisions. It might be better to refer to them as 'guidelines' though, but still...
And I'm not sure what you mean by "the main points will have been cleared", but it's virtually impossible to 'think' your way through the game in its current form. The fact that people come here and give hints/walkthroughs really has little to do with the game, and the game really should not have to rely on that to be able to be solved by the average player.
Anyway, I do have some things to add:
In rule 1, feedback does not have to be immediate.
Let's create a little scenario to illustrate this: Say you're walking the girl home, she's wearing a tiny dress and it's cold. Let's say you have some dialogue options (in which she might actually say she's cold), as well as the option to offer her your coat. Now by all means, giving her your coat is the best option. But, not giving her your coat does not justify direct feedback: if she'd say something like "Why are you not giving me your coat?", it'd come off as very pushy. A much more realistic alternative is that she begrudgingly accepts that you might not be a true knight in shining armour without saying anything (happiness -= 1, imo :P).
But now we're missing the feedback, so enter part two of the scenario: you're back home and you made her a nice fire in the fireplace (happiness rises again). But, as a twist, because she was so cold earlier she refuses to go away from the fire (to more interesting locations?), clearly stating "No, I'm too cold, I want to stay here". And there's your delayed feedback, it does not have to be a game-ender, but it will certainly make the player wonder if he could've prevented her from getting cold.
In my opinion, a scenario like this is much more pleasant than a direct "I want your coat" type of feedback.
And while not really on the topic of actual design, I can add another 'guideline':
As a designer you're not the best tester. Now of course a designer can do plenty of testing needed to make it work (verifying paths works as planned, every works as per design, etc.). But, it is very hard for a designer to consider his own game as a 'black box', because he knows what's inside. He can never play it with the same naïvity as the actual players. As such, the quality of a game can really benefit a lot from some testing by a few motivated players before releasing it (of course these people shouldn't be afraid to show discontent with certain things, just saying "good game, I like it" is not gonna help much

).