Restrictive Design Standards = Vast Fields of Opportunity

September 27, 2007

The developer commentary included with Team Fortress 2 has one interesting segment on a conundrum they faced while designing the Spy class. They wanted players on the Spy’s team to be able to identify them as friendlies, and they wanted to do this without adding another 2D element to the HUD. The solution they came up with was inspired: have the Spy put on a paper mask with a picture of the class he’s masquerading as on it. The lesson here was that holding one’s self to a strict design standard can force innovation by removing the easy way out. It is said that restriction’s are an artist’s best friend, and their Spy solution is a textbook example.

I think this is a great concept, but it should be pushed farther. So how about making a game of a certain genre that is known for primarily revolving around a certain kind of play mechanic, and then refusing to use that mechanic at all? This would force the developers to come up with new ways for the player to interact with the game and their environment, which would create new styles of gameplay and inject a big fat slug of innovation into the games industry. The play mechanics that would be developed to fill the sudden vacuum could later be refined and applied to other games all over, potentially opening up entirely new genres and game types.

And the most overused type of gameplay is combat. When a developer doesn’t know what else to do, they hand the player a gun, put them in a room full of monsters and let them fight it out. This is true of several genres, almost by definition, but it doesn’t have to be. How about an RPG with no combat whatsoever?

With no combat to balance, development resources could be focused elsewhere, such as dialog and writing, and here’s where an idea really grabbed me: make a game that only has 8 or so NPCs, but each of them has a thousands and thousands of lines to dialog. Instead of making a whole lot of little NPC interactions, focus on a very few but very big and deep and complex NPC interactions, with hundreds, or even thousands of permutations.

And to give players more control over how their half of the conversation plays out, perhaps they could be given two option categories they must pick from during each round of dialog. One category might be answer content, with options such as “yes,” “no,” “I don’t know,” etc, and the other category could be a tone of expression, such as sneering, humble, proud, direct, and so on. Different NPCs would not only react to what the player said, but how he/she said it. And this would be true of every dialog option in the game.

With the current techniques involved with writing NPC dialog, this is probably an unworkable goal. So of course new techniques would be developed to meet this challenge. For instance, maybe the developer could create a special AI to aid in this process, one that has a set of preferences and then looks at the player’s current state, the player’s past actions, the state of the game world, any story flags that might be relevant, and so on, and then chooses the line of from a list of responses that best fits the context of the conversation. Or maybe an AI wouldn’t be much help, and the solution would just be an improved method of scripting conversations. In any case, the main goal would be to move beyond simple dialog trees, and more towards a simulation of a real conversation.

If every answer had a real and immediate effect on how the rest of the game played out, the replayability of this game would be astounding. This quality could be enhanced further by deliberately writing the script so that you could only get part of the story during a single play-through. A single play-through would contain enough information for the story to make sense, perhaps, but the NPCs could be set at odds with each other so that to get all the information an NPC could provide he/she would have to earn one NPC’s trust at the expense of another’s. A player would never be able to get everybody to tell the whole story in a single game. And of course the ending would change depending on the player’s interactions with all of the NPCs, so even if a player decided to stick close to NPC X during two different play-throughs, the ending might still be different depending on how the player treats NPCs Y and Z.

This is just one of the ideas that came to me while pondering VALVe’s example of how a commitment to design principles yields big results. I have more, which I may post about later. The daydreams this will be running around in my head for months, I can already tell.


One comment

  1. […] wrote an interesting post today on Restrictive Design Standards = Vast Fields of OpportunityHere’s a quick […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: