Writing game plots: interest and emotion

What makes a good game? The graphics? Probably not. Graphics tend to impress the in the first minutes of play, but an hour later we no longer see that nice cloud of dust, that detailed zombie hitting you with his forearm or the realistic scenario. Gameplay and storyline are the key elements that endure the whole game playing lifetime, and I will talk about the second here.
What makes us keep playing the game until the end? If you say interest, you are right. So, if you want that players actually finish your game, you have to build interest. Want more inmersion? Then you need emotion.
Emotion and interest are not the same. You can have the second, but completely miss the first. Fallout 3 creates interest. Mass Effect creates emotion. When there is interest, you look at the clock and say "it is 11 PM, I will continue tomorrow". When there is emotion you say "it is 1 AM, but I have to finish this quest before going bed!".
Gameplay can generate interest, but emotion is only generated by the plot.
There are some well known elements to introduce emotion and interest in your plot, they are easy to spot if you read a best seller. Take a look at Harry Potter books. They have an almost predictable structure: there is a mistery, of course; invariably there is somebody we will hate and the rest is filler, what makes the book diferent from the others. Easy, don't you think? But, being a writer myself, I tell you that sometimes we can miss these elements.
A game plot is like a book plot with alternative pages. You can have a mistery based plot: the player is embarked in a quest to discover something, but won't achieve it until the end, or will reveal it progressively, piece by piece, in some cases finding that the object of his pursuit is totally different to what he thought in the beginning of the quest. Mistery can be built around anything: an item, an event, a character. Hide to the player why the item needs to be recovered, what really happened in that event or what are the true objectives of a character, or what are his own character true story.
The other plot element: a nemesis, somebody we will hate. This hate can be built through his/her actions: betrayal, murder, etc. Same as mistery, you can escalate by providing another more evil enemy when you reach the first one. A very useful element that fits in any plot, and in fact, it is found in virtually all plots. At the same time, the opposed figure can be used: a friend or somebody to admire.
Decisions can be another way to create emotion. Confront the player with a choice: should I save the village or just the girl/boy I love? My best friend has taken the wrong path, should I kill him? There is a wide range of possibilities. The interactive nature of a videogame makes this a very interesting resource, but do not overexploit it. Difficult decisions are good, but not to trust the whole emotion building process to them. You want to give the player something fun, not a greek tragedy.
Avoid excesively open worlds. Too much free roam disconnects the player from the storyline and dilutes the interest. It is good to have several paths to reach the end, but not too many paths taking you away from it. Introduce some reminder to bring the player back to the main quest.
Action (in the sense of fight) is another source of emotion. But not all action weights the same, it is more effective when you fight a particulary important and strong enemy. The rest is just interest, not emotion builder.
Build complex characters, both player and NPCs. Take some time to create a background for each one, and plan how to progressively revealing that information. Establish links among the player and NPCs, make the player feel that the party members are there for more than helping kill monsters.
That is a short list of techniques to create some emotion in your game, but Im sure it is far from complete. Like a professional writer told me once: you will learn to write only by writing. Go and experiment!