For combat-specific areas of the game, we want to start with a simple approach and build up from there. The key reason for this is to keep things light and lean for quick iterative testing as we focus in on what works.
As the combat designer, this is where I usually start to come in and throw a monkey wrench into the equation -- um, I mean, help make things fun. Working in conjunction with the point designer for the level, as well as with the game director and leads, we'll figure out a starting direction to try based on the requirements for the encounter. The goal at this point is not to do any fancy or complicated combat scripting, but to just put in the most basic pass we can that gets the essence of the desired direction across.
From here we'll do several iterations to try different ideas and find out what will be the best fit. Sometimes it's small changes and enhancements to the basic first pass; sometimes we scrap the whole thing and try a different direction. This is why it's really important not to get too complicated from the start, because we want that ability to scrap what we have if need be without remorse.
It's very difficult and rare to come up with something from scratch that is going to be perfect on the first try, so you need that ability to quickly iterate, along with the willingness to try different ideas. As we get closer to what we're wanting, the scripting will start to become more complex, but only once we know we're on the right path.
We also need to have a critical eye and awareness of analyzing the fun factor. As an example of what I'm talking about, if you play through something and it seems "alright", don't just settle for that. Dig down and try to figure out why it's just "alright" and what can be done to improve it. Granted, these are the first steps towards finding the right direction, so it's not going to be absolute perfection, but you should still be able to feel out if you're just putting in something that's mediocre, or even worse, frustrating.
Also, don't forget to compare it to the goals of the space and story. For example, if you're trying to give a sense of being oppressed and pinned down, if you feel like you have the ability to roam around without much threat, then you're probably not going in the right direction yet. An eye for the details can go a long way in turning the encounter into something awesome, but the trick is to not get bogged down in the polish details at this phase.
Once I've gotten something that I feel is working and going in the right direction, I'll start to have the point designer, lead designer, and game director play through it to see what they think. We'll probably start with a few rounds of iteration amongst this group, but eventually we'll start to grab whoever is available in the area.
Our office is arranged in giant open spaces where we keep the different disciplines together. Our desks are arranged in groups of four with half-height walls, so it's really easy to hear and see everybody. At this point I'll literally pop my head up and grab anybody that's close by (which would be fellow designers, mainly) to come over and try the encounter.
The standard focus testing rules apply when observing someone play (i.e. you keep your trap shut). However, it's alright to give them a brief synopsis of what has happened up to this point in the story, so they have some context when playing through the encounter. A few key questions to keep in mind:
Once a person is done with a playthrough, we ask them for their thoughts so we can get their raw feelings about what they just experienced. After gathering all of their feedback, we'll start to grab several other people and run through it again.
Our fellow coworkers are one of our best resources, so we want to get a variety of opinions involved. We also want to use people that aren't just the "core" gamers, as we want to get a good smattering of different skill levels. Some people are going to like head-on conflict and combat, others will prefer stealth and being crafty, and we want to see every approach, and what lies in between.
Along the way we'll make adjustments to the encounter and fine tune things, but the goal is to find the right direction for the space. Once we've accomplished this, then it's time to hand the space off to the artists for beautification. From here on we'll be working with the artists to oversee the space, making small adjustments along the way as the blockmesh starts to change and become more organic.
Something that's always a bit of a struggle is the maintaining of standards during this process. While those blocky shapes are absolute gameplay perfection, it's not exactly a Rembrandt -- a child's crayon interpretation would be prettier.
So as the artists start to turn the gameplay masterpiece into something visually stunning, we need to work closely with them to make sure certain gameplay standards are adhered to (cover heights, jump distances, etc.) We want to find the best compromise for what works. Even though we'll favor gameplay for critical areas, we still want it to look its best and let the artists strut their skills. It's always a collaboration.
During this time I'll start finalizing the scripting and gameplay for the encounter -- making sure the right visual variety of enemies are present, adding weapon and ammo drops for the player, polishing up the scripting and making it bulletproof, etc. Then we'll start the entire process again on the next level, usually having several running concurrently over the course of production.
So there you go. That's a brief walkthrough of how we approach the process of encounter design here at Naughty Dog. In the third and final installment of this series I'll be covering the technical side of how our AI works, techniques we used to make our combat fun, and then lessons we learned along the way.