Written 06.03.2008 - Uploaded 15.09.2008
This short article takes a look into the intended storytelling method of the game. It was written to point out the problems that might arise but also to show that things aren't as bad as they initially look like. This is actually much more megalomaniac than the actual combat part of this game. Anyway, on to the article itself.
---
While the game's focus is, from the game perspective, more in its tactical combat system, I feel the story is somehow even more important. Sure, the game mechanics have some new elements but it may turn out we can't even implement all those plans. In my vision of this game, its the story that keeps players hooked. This article however is not about the story itself – we'll save that for later – but instead goes into the storytelling methods and specifics.
The primary question is handling character deaths, since permanent character deaths are a part of my vision at the moment. It does sound like an impossible task to take all these deaths into account and branching dialog accordingly so we should look more into this. There's also the handling of key characters to consider, and for that there are two alternative methods which are discussed later in this article. Let us, for now, assume that key character deaths cause game over because the plot cannot continue without them. Any piece of dialog in the game should be designed keeping this in mind, so we have to be careful here.
The method I am currently favoring is writing several versions of each dialog, as many as there are possible combinations of missing characters. This of course includes participants and any characters that might be a subject in the conversation or affect the emotions of some participants. Now I do realize this sounds like a lot of work, and it probably is too, but there are a few relieving factors. First of all, if key character deaths mean game over, there's no need to take their deaths into account. Since much of the dialog (especially core dialog) actually involves some of them, this should really help. Second, typically at least one participant in a conversation is the one who initiates it. This means that, if the character is already dead, the conversation just never happens. Third, there's some emotional mapping to be done, which tells us which characters really care about which characters' deaths. It's as not many as you might think. Naturally deaths typically also spawn new dialog, at least immediately after. If the character in question was not important to anyone then probably that's about it.
Of course, it might still turn out there's too much work in this department. Even if we have several writers (which we WILL have, I'm so not going to write this metric ton of dialog alone) there might be reason to consider other options. However, for now we'll leave this as it is.
Key character deaths can be handled in one of two ways. The game over method is one, and the substitute method is another. Probably a mix of these two can be used for the best outcome. The game over method should be rather obvious. As long as a character is marked as key character, his death causes a game over (notice that after their key event(s) have passed their key status is lifted). The substitute method could work for more loose key events, where it is not crucial to have a specific character but instead one from a group of characters. There will be some kind of inheritance order. If the primary is dead, use the secondary etc. The chain should always end with a crucial key character whose death is not allowed. This way we can be certain that the event does take place without us having to go through every single character alive to find a substitute. There's always some character to fall back to.
Ironically, the key event for some characters actually involves them dying.
Next, I'd like to share some light into events. Typically plot and private events take place before a battle and afterwards, and the amount of events may vary rather wildly. Some events may also take place during a battle. Plot events are the ones that take the story onward, and they should always happen in one way or the other. Private events on the other focus on giving depth to the characters and these are typically subject to cancellation or substitution due to character deaths. Finally there's some kind of special events which are only triggered by certain conditions, such as a character dying.
So typically the game proceeds like this: First there's some events, mostly private, and some plot events where the next destination is planned (or options). Then the player is shown a map and a list of places to go to (this is usually one, but we may want some short branches in the plot. Also, there might some triggered side quests available.) Once a destination is picked, more events occur, this time more related to this specific scenario. Finally after all these events have been watched through, the actual scenario begins. After the scenario, there's typically some aftermath events and then we're back at the beginning of this chain.
So how long is this game? Long. The plot alone I've been planning for it is rather winding and then there are all those private events (assuming the player wants to view them!) I don't see this as a problem however, because we're doing this in the TV series release method, one episode at a time. Typically one episode should include several scenarios and the game should be released in reasonable chunks, giving the players a lot of content but also within a reasonable schedule.
Finally some notes about the writing process which becomes more complicated if there are several writers involved. At this point I'd suggest that characters are assigned to writers who then familiarize themselves with these characters and are responsible for keeping them true to their natures. This doesn't necessarily mean they write all the dialog for these characters. The writing process starts with the outline phase, where the writers more or less collectively design and approve outlines for dialogs (these should be rather short, including just participants, subject, mood and approximate outcome). Then each dialog is assigned to a writer (preferably the one whose characters are most involved in it) who then writes it once, using optimal conditions (everyone alive). Then each dialog is (spell- and character-) checked and after approval other versions can be written, which are then also validated etc. Once all dialogs are written and approved, we can move on to the next bunch of events. Yay?