David Gaider on the Nuts and Bolts of Writing

David Gaider has recently took to typing about the writing process at BioWare on his personal Tumblr, and given his ruminations were thorough enough to span no less than four posts (1, 2, 3 and 4) I thought it would be worth it to report them here. Here's a snippet on quest writing:
Where does it start?

First thing you need to do is figure out where the plot begins. What's the player's entry point? Is there more than one? Do they come across the quest's beginning in the middle of their travels, undirected? Are they given the plot by an NPC? Is it part of a chain of plots, and they arrive at this plot from the clues of a previous one? That gives you a (point A) from which to launch.

Where does it end?

This is good to figure out before you start plotting out the rest. What's the endgame for the plot? If you intend to have multiple choices, it's good to decide them now and reverse-engineer them into the rest of the plot.

Why? Imagine this: you want a plot that ends with the player deciding whether to kill a man or spare his life. You could just insert that at the end of the plot, after the battle with said man is done. but what if you haven't given the player a good reason to do either of those things? You might want to insert places in the plot where the player could get info on why this man deserves to die, or what he's done to deserve forgiveness. or provide consequences to the player's decision which they'll know about for certain by the time they get to that point. If you don't do that reverse-engineering, you'll go through your plotting and likely discover the equation doesn't add up like you thought it would.

What does the player do?

Here's the biggest mistake that new writers make with plots they don't track exactly what the player is doing at every point. They have a solid (point A) and it connects to points B, C, and D. maybe there are even decision branches along the way. But the connection between some of those points becomes. vague.

"The player finds the artifact and then goes to (E) where they discover the zombie eating the corpse" .wait. How did they go there? Why? Did they have no choice, as in it was a hallway and they just had to keep walking in one direction until they came upon the zombie? What if they go the other direction? What if they're in a city and decide to run off can they come back to it later? Can they miss the zombie completely?

You need to account for every possible action the player might take. You can close off other paths, but if you close off every path you end up with a hallway and likely a loss of agency. That can even work fine, so long as the player still has some choices to make. and you have to spell those out. Another designer looking at the document you create for the plot can't be left wondering what's supposed to happen at any given point. Worse, what if they don't wonder and implement the plot exactly as you've laid it out. and now a QA tester is playing through the plot and does something unexpected, or has no idea what to do next?

Another mistake that often occurs for new writers is that they spend too much time thinking about the story as a thing which just occurs they have an idea of a plot which revolves completely around the NPC's ((so there's this princess and she loathes her father, but her father really loves her deep down and wants the best for her, but that's complicated by his wife who thinks.)) and no place for the player to insert themselves. Or they have a story which just is ((a meeting between old friends at a bar which gets out of hand)) with no actions for the player to actually perform other than to be present and witness it.

We call those (high concept) plots, and generally they have a story which would make for a wonderful book but don't make for something that's fun to actually play.

What will the player feel?

Important! This sometimes gets lost somewhere between coming up with the story and plotting out the exact points where the player experiences it. We need to keep in mind what the player is supposed to feel at the end of all this.

Often we come up with what we call the (razor) a simple phrae or sentence that describes what the plot's about. (Harrowing race through a demon-infested tomb), (Indiana Jones meets Hellraiser), (Sophie's Choice between two beloved town leaders). all of these could work. You just need something against which to judge stuff you intend to later add. Does it work towards that razor or detract from it? Often you'll be tempted to keep adding stuff in, and then be surprised that suddenly your pacing is shot ((how is this '˜harrowing' when the player wanders around lost for 30 minutes?)) or that you've lost the thread of your story ((you said Indiana Jones, but there's not even a single trap in this dungeon)). If we think about what the player's supposed to feel, we give some thought into what will make them feel it. and also why the player might not feel it, and what to do about that.

Ask yourself (why?)

First thing any good QA tester is going to ask is (why?) Why am I doing this? It might be a simple plot where the player's explicitly accepted the quest. but do they even have a reason to accept the quest in the first place? Is this something they want to do? Are there multiple ways to feel about it? Try putting yourself in the shoes of players of multiple types why would a player who's just trying to be the good guy do this? What about a player that just likes to feel clever? What about a player who likes action? If it's an optional quest the player could miss, the answer might very well be (they won't) and that's fine, but you're better off giving those different players different reasons to do it and/or different ways of doing it rather than cutting them out of content completely.

Same question applies to the various plot points: why did that NPC do the bad thing? Why would he give up his special artifact? Why was the artifact special? You don't necessarily need to explain the answer all the time in fact, doing that constantly (and assuming the player will be questioning every single event and trying to counter those questions before they ask) will lead to the sin of over-explanation, but you do need to know the answers yourself. and if you're not going to provide answers you do need to make what happens plausible enough that the player's not left stumped and pulled out of the story.