creative process Deep-Dive: ¡QUIXOTE!
This page covers my approach to:
- Early project planning.
- World building and collaboratively defining creative vision.
- Defining an aesthetic and creating the game's visuals.
- Designing and iterating on new systems.
Matthew Weise: Designer, Writer, Scripter, Project Co-Creator
Brett Camper: Designer, Coder, Project Co-Creator
Yue Li: Concept Artist, Pixel Artist, Fencing Consultant
PROJECT GOALS - WHY MAKE A procedural swashbuckling sim?
Our goals were to create a game that:
- used map data of everyday, modern spaces to drive its core player fantasy
- was unpredictable and different everytime you played it
- allowed you to play with people and explore transgression, but without extreme violence or murder
how IT began
Brett had created an algorithm that took local city map data and expressed it as 1980s-style pixel art "world map", like those in old Nintendo games like The Legend of Zelda, Final Fantasy, or Dragon Quest.
Brett and I decided this could be the basis for an interesting video game, a modern-day spin on old 80s games that used the real world as content.
But how should we frame it for a modern audience?
FINDING THE RIGHT METAPHOR
We wanted a core player metaphor where the fact that the game:
A) was based on real world places
B) looked like 80s pixel art
informed the design, the presentation, the art, everything.
We settled on the idea of a "Modern Day Don Quixote" where video games, not novels, were the source of the protagonist's fantasy, his belief that he is a "knight", or, in this updated verion, a "video game hero".
This was a bit of a revelation. We realized that the Quixote concept actually has strong analogs in video game culture, from cosplaying favorite game characters to comedy shows like Mega64.
Our core concept became: what if you were in the real world, but to you it looked like a video game? What happens when a video game hero starts trying to interact with the real world?
Initial design mock-up
Here there will be a slide-show of the first storyboard I did which walks you through the second-to-second player experience flow.
The first challenge was map scale. How big is too big? How small is too small? The algorithm we had could create the map of New York at any scale, but everything always looked like the same 8-bit tiles.
At the largest scale, a single tree tile represented 16 city blocks. At the smallest, it represented the size of an actual tree.
After much experimentation, we found a middle ground we liked.
ADD: SPECIFIC NUMBERS FOR THE DIFFERENT SCALES, A BIT MORE DETAIL ABOUT WHY WE CHOSE THE ONE WE DID, WHAT FACTORS MATTERED, SCREENSHOTS OF DIFFERENT SCALES + A COMPARISON PHOTO OF THE ACTUAL CITY THAT MATCHES THE MAP
WHAT THE PLAYER CAN DO
What would a person who thought they were a video game hero want to do?
We started with the main character, our "Quixote", and thought what behaviors are Quixote-like? What sorts of things would players want to explore as a Quixote avatar?
For us this was: bravery, mischief, cruelty.
Players want to get into situations where they can act brave, but also, because players know this is a fantasy, toy with the people they find in humorous ways. We also felt that, even though Quixote isn't cruel, that preventing the the player from acting cruel would make Quixote's heroism meaningless. Players need to have the opportunity to be cruel and choose not to. This makes their heroism real... well, as real as it came be in a game.
With this in mind we designed our core player verbs so that, when mixed and matced in different contexts, they would result in such choices.
Verb: move, take, threaten, attack
This is where we began: a game where the player could move around, pick up whatever they found on the ground, and then pretend whatever they found was a sword. They could then use that "sword" to intimidate people (threaten them by pointing) and attack them if necessary.
How would everyday people react to a person running around thinking they were a video game hero?
We wanted a simple-yet-consistent artificial intelligence system for our Non-Player Characters, one that encouraged player experimentation and would result in fun, humorous encounters.
Rather than traditional combat games, we were more inspired by stealth games -- games about subtrifuge and sneaking, like Metal Gear, Thief, or Hitman -- for our A.I. paradigm. This is because stealth games tend to have high degrees of simulated awareness in their NPCs -- NPCs that are not trying to kill you initially, but are just going about their daily lives and react you to shattering their sence of normalcy.
We wanted our NPCs to embody: curiosity, fear, surprise.
This is why our NPCs have limited vision cones and, so you can sneak up on them. This is also why they have a basic A.I. routine of minding their own business, which switches to bemusement, fear, or anger, when the player interacts with them.
NPCs become curious about Quixote ("Who is that weird guy running around?"), will investigate if the situation escallates ("Did he just disappear behind that tree?"), and react if threatened ("Ah! Get away from me!")
We arrived at all this rather intuitively, based on the sorts of behaviors we felt were believable given the situation. However, once we had our basic prototype running we realized we'd created "a simulation of acting crazy in public", where the core gameplay is about breaking social codes and seeing what happens.
Isn't that, at some basic level, what Don Quixote is about?
Creating the player character & Visual style
conversation system (talking!)
Here I will discuss the evolution of our conversation system, why we chose to have one, how we approahed it, our interations, roadblocks, and results.
COMBAT SYSTEM (FENCING!)
Here I will discuss the design and implementation of fencing, our inspirations for the system (original Prince of Persia, plus the novel of The Princess Bride), how we saw it as an organic part of our social simulation, and the implementation we arrived at.