RPG Conversation Redesign
So I was dwelling on Noah Falstein's article in this month's "Game Developer" magazing about "Emergent Complexity." (See his 400 list, item # 105, if you haven't read the article). Muddled with that was a couple of talks I heard LONG ago by Chris Crawford on storytelling in games, and some ideas I have had in my brain about creating "non-combat challenges" in Computer Role-Playing Games (CRPGs), and the trickiness of doing conversations in said games. And risk-reward structures, etc.

Am I the only guy who ever muses on these things? This probably reveals way more about my lack of mental stability than I should be admitting in public. But oh, well..

Anyway, so I'm ruminating on this collection of topics, and I have a "Duh!" moment. Which is sort of like an, "Ah-hah!" moment, but more of a realization of how I've been handling things all WRONG while simultaneously discovering the blindingly obvious. Why I was having a tough time working out how to make conversation work as a game challenge. Part of it was that I was still thinking inside the box of how conversations are handled in other RPGs.

And quite frankly, in about every Computer RPG I can think of, Non-Player Character (NPC) Conversations are simply a variant of terrain.

Huh?

Think about it. What is the purpose of terrain in an RPG? It's not really part of gameplay. It's mainly something to navigate and explore, something which creates a natural separation between game segments. It gives you different visuals and pretty stuff to look at. But there's not really an option of "failing" to navigate the terrain - it's either possible, or not given your current game state.

NPC Conversations serve the same function. You navigate the conversation trees, which dispense knowledge about the game world (scenery). Conversations separate the game elements by unlocking options. If there are any random elements in conversations, it is restricted to purely optional content. But everything else is automatic, you just have to walk your way through the "conversation path" to access it. To quote Yoda, you either "Do, or do not. There is no try."

In order to turn the 'conversation engine' of an RPG into a full-fledged part of gameplay, you need to make a few fundamental changes:

Risk / Reward:
You need to add a risk / reward structure for all conversational options, including the (possibly invisible) option to not engage in conversation whatsoever. The "reward" is the easy part. Gifts, payments, or access to additional content are already common. The trick is forcing the player to invest some kind of game resource into the conversation. The player character has nothing to lose by talking to everyone available and exploring every conversation tree as fully as possible.

In The Sims, this investment really comes down to in-game "time." Poor performance in a conversation can have a negative effect on the relationship between the characters, which means more time will have to be spent later to "repair" the relationship.

Most RPGs don't really have "time" as a gameplay element. Some games will have day and night cycles which might have some effect on availability of NPCs and shops and so forth, but time is RARELY plot-critical. The imminent invasion by demonic hordes will almost always pause in their plans until you've cleared out the newbie dungeons and talked to everyone in town six times.

But time doesn't have to be the resource to be invested by the player. It could be something as abstract as a "number of conversations allowed in a row," or preferably something closer to how Real Life works. But the player has to be required to put in-game resources at risk somehow.

Failure Must Be An Option
So the next thing that needs to happen is that the player needs to be allowed to "fail" conversations.

I think the lack of failure as an option at all is exactly why most combat is generally the most common in-game challenge: Failure effectively ends the game (with the death of the Player Characters), so the designers don't have to deal with the explosion of branches that could occur. Like what would happen if the players FAIL to rescue the princess from the dragon? The entire rest of the game would have to be written to handle both the "Princess-Rescued" and "Princess-Was-Devoured" outcomes. Multiply that by hundreds of such outcomes, and it can be tough in a story-based game.

Non-story based games, like "Civilization," take this kind of thing into stride. Failures and successes are simply abstracted into the running game state. But the flip side of this is that individual successes or failures lose importance before the aggregate whole. We'd really like the fact that the player botched the princesses' rescue attempt to MATTER in the course of a game.

But bringing it back to NPC conversations. Traditional "conversation tree" design wouldn't allow something like this to be bolted on. The game would screech to a halt if the player had to make a "conversation" skill check on every branch to continue exploring the tree, and somehow managed to fail every roll with major NPCs. The player might find themselves unable to leave the newbie area of the game. Instead, a much more "organic" system would have to be put in place that could aggregate individual successes and failures, never leaving the game in an unwinnable game state.

And then....?
Hmmm.... I thought I had a third thing, but it's eluding my sleep-starved brain right now. So maybe it wasn't as blindingly obvious as I thought.

So that's what I've got so far. This doesn't solve the problem, but it does point out some directions to look for a solution. So am I on drugs, or shouting from the lowly heights of Mount Obvious? Or both? Post and let me know!
- posted by The Rampant Coyote @ 11:00 PM (Permalink)