For the production of Anything But a Gun, otherwise known as ABAG, I was the producer/team lead and the design lead. The project was our senior project for Game Workshop, a class that lasted two semesters instead of one. That gave us roughly nine months to make a game from start to finish, with a team of 14 people.
The semester before ABAG, I had worked on another project called Thermoflux. One of the biggest problems we had was a lack of leadership. That experience pushed me to take on the role of team lead for ABAG, not necessarily because I wanted to be a leader, but because I felt like somebody needed to do it if we were actually going to stay organized and finish a game. While working on Thermoflux, I realized that I was always the person who knew what was going on with the project and where we were in development. I figured that probably made me the best person to take the reins.
Once our group was formed, we had to figure out what kind of game we were actually going to make. I wanted to make the process as democratic as possible, with everyone's vote being weighed equally. My thinking was that people need to actually like the game they're working on if you want them to stay motivated. I polled the group for which genres, aesthetics, and art styles they like most.
During brainstorming, a lot of us were interested in making a fast-paced 2D/3D hybrid shooter. We had one team member who was pretty reluctant to make a shooter, though, and our engineering lead kept pushing him to figure out what exactly he didn't like about the idea.
It turned out that he didn't actually dislike shooters. He was just tired of the same generic realistic military setting that seems to show up in so many of them. He said he would be fine with making a shooter as long as the thing you were shooting wasn't a gun. It could be an ink brush, a lava lamp, or basically anything else.
And that's where the name came from.
Anything But a Gun stuck from the very beginning and never changed.
After that, the rest of the brainstorming process went surprisingly smoothly. Sometimes almost too smoothly. Someone would throw out an idea, and three other people would immediately jump into the conversation and start adding to it. There were very few "no"s. Everyone was laughing and getting excited about the weird weapons, wacky lore, and comic book visuals we were coming up with.
That was a pretty big contrast to another game concept we had prototyped, where we were constantly struggling to make the gameplay and lore connect in a way that made sense. ABAG just clicked.
Our first prototype of Anything But a Gun was pretty barebones. We didn't have the outlines yet, there were only two enemy types and three weapons, and the demo level was small and mostly empty. There was a lot missing.
What it did have was decent gameplay.
From the beginning, we had the floaty jump that let players jump onto rooftops, as well as the fast-paced movement for both the player and enemies. These weren't things that happened by accident. They were some of the most important parts of the game that we wanted to get right early.
Our class had regular demo days where other groups would play our games and we would play theirs. Those became some of the most important deadlines for me during development. Getting an A in the class was relatively easy, but getting the game into a state where we could actually put it in front of other people and get useful feedback was much more valuable.
Because of that, I cared more about being ready for those demo days than I did about some of the actual class due dates.
For that first prototype, we focused on the things that mattered most: shooting, player movement, enemies, and, to some degree, the level layout. The goal was to get the game to a place where we could test if its primary concepts were actually fun in practice, and they were.
One of the bigger changes we made during development was moving away from the halftone-heavy art style we originally had.
I put together a presentation for the team explaining why I thought some of the halftones and other elements were making the game feel more cartoony than I wanted. I thought the game could still be funny and weird without looking like it was specifically made for kids, and I wanted to push the art style toward something a little more mature.
This ended up being a pretty big discussion with the team. There were good arguments on both sides. A lot of people liked the cartoony look because it fit the ridiculous weapons and the general personality of the game. And honestly, they had a point. We have a fart gun for crying out loud.
My concern was that we still wanted the game to feel like a serious, high-quality shooter underneath all of the ridiculousness. The gameplay was fast-paced and required a decent amount of precision, and I didn't want the visual style to make the game feel like it was aimed at a much younger audience than it actually was.
Eventually, we landed on a distinction that I think helped solve the problem: we weren't trying to make the game cartoony, we were trying to make it comic-book-y.
That let us keep the exaggerated weapons and humor that made the game fun while moving the art direction toward something a little sleeker and more contemporary. The halftones had originally been our way of referencing retro comic books, but they ended up being one of the things we removed as we refined the style.
Looking back, I think that discussion was important because it forced us to figure out what we actually meant when we said we wanted the game to look like a comic book. It wasn't just about adding comic book effects. We had to decide what parts of that style actually supported the identity of the game.
One of the biggest technical problems I dealt with during development was figuring out how to get outlines onto the environment.
The first method I tried was the inverted hull method. I quickly ran into a problem, though: outlines on the edges of walls and floors wouldn't show when those surfaces were flush against each other.
So I tried something completely different: a fullscreen render pass shader.
This method used the normals of the surfaces on screen to figure out where the outlines should go. It worked pretty well for the edges of walls and floors, but it couldn't properly outline changes in color.
Technically, it could do that because the tutorial I followed included that feature. The problem was that turning it on caused thick, splotchy outlines to appear all over our 2D enemy sprites anywhere the color difference met the threshold.
The render pass also used a depth map of the scene to figure out where the outlines should appear. Since depth maps don't account for sprites in the same way they account for 3D geometry, the shader would sometimes draw outlines around objects that were actually behind enemies. This made the enemy sprites look strangely transparent.
I did a lot of research and followed pretty much every relevant tutorial I could find on YouTube. Eventually, I came to the conclusion that there probably wasn't an all-encompassing solution without custom HLSL code, which was beyond my skill level at the time.
So I shelved it for a while.
There were more important things I needed to work on, especially level design and gameplay. When I eventually came back to the problem, I decided to give the inverted hull method another, more in-depth try.
This time I followed a different tutorial and got a much better result. Instead of using a fullscreen render pass, this version used an actual second shader material that was applied to the objects in the game.
I was initially pretty concerned that essentially doubling the polygon count of the environment would be too expensive. Fortunately, we'd done enough optimization leading up to that point that it ended up being fine.
It wasn't perfect. There were still some objects and corners where the outlines didn't render correctly, but by that point I had a solution that worked well enough. For some rooms, I even manually created lines in the corners where the shader couldn't get them right.
It wasn't the cleanest solution in the world, but it worked.
Level design ended up being one of the more interesting challenges of the project. It was fun to observe patterns in playtesters playthroughs and try to identify why they played the same. I'm proud of some of the solutions we came up with.
One problem we noticed during playtesting was that players would get lost or run straight past things that we expected them to notice. If someone ran past a reward or a machine, we couldn't just blame the player. If a lot of players were doing the same thing, there was probably something wrong with the level.
I eventually came up with a theory for why this kept happening.
The game is, at its core, a speed-running shooter. Once players got into that mindset, they wanted to keep moving quickly. If they entered a room that we knew contained a reward, that didn't mean they were suddenly going to slow down and explore it. They wanted to keep going, kill more enemies, and move on.
That meant that we couldn't just put something important in a room and expect players to notice it. We had to make them look at it.
This changed the way I thought about our level design. I started paying much more attention to the player's sight lines and how long they would actually have to look at something based on the flow of the map.
From there, we could change the length of corridors, move entrances and exits, and use turns in a room to make players look in specific directions.The goal wasn't necessarily to stop the player from moving quickly. It was to design the level around the fact that they were going to move quickly.
Designing levels for ABAG sometimes felt like playing chess against the player. Only I wasn't controlling the pieces. I was controlling the board and trying to predict how they would move through it.
One of the parts of the project I had the most fun working on was the VFX. I made all of the VFX for the game myself, from blood splatters to explosions, using Unity's particle system and my own art skills.
The craziest part, though, was that I also made all of the weapon reload animations using particle systems.
There was actually a good reason for this.
When we started figuring out how we were going to handle reload animations, we knew we couldn't make one animation that worked for every weapon. The weapons were way too different from each other. We would have needed a custom animation for almost every weapon.
We also didn't have an animator on the team. Even if we did, modeling, rigging, and animating hands for all of those reloads would have taken a huge amount of time.
One idea we considered was scripting the weapon's transform so that it would move around and look like the weapon was being reloaded just off camera. That probably would have worked, but our engineers were busy, I really liked working with particle systems, and I knew that you could use 3D objects as particles. So I wanted to try something a little ridiculous.