[In this comprehensive postmortem, former Volition developer Luke Schneider talks about how he went indie and assaulted the Xbox Live Indie Games market with an array of games built on the same ideal -- and how each worked, as well as the fate of the overall project.]
I had worked at Outrage Games for five and a half years when it closed in July 2003. Because my wife and I had bought a great house two weeks earlier, we had a choice to make: I could start looking for another games industry job out of state, or I could try to turn a Game Boy Advance prototype I'd been playing with into a full-fledged game. Making a game by myself was a challenge I couldn't resist, so I chose the latter.
The game was completed in a few months, but finding a publisher was a more difficult and demoralizing process than I expected, and a half-completed Xbox port fared no better. So I made the move to Volition in 2004. Though my six years at Volition were enjoyable, the desire for control over my destiny eventually overwhelmed my need for stability. In early 2010, I left Volition to form Radiangames.
When considering the development options for Radiangames, I felt a strong urge to do the opposite of what I'd been doing recently. Instead of taking five years to make one game, I thought it'd be cool to make five -- or more -- games in one year.
My plan was simple: I would create a monthly series of arcade-style games for Xbox Live Indie Games. I'd seen Halfbrick and Arkedo, two small independent studios, do their own series on XBLIG, and I admired the Pixeljunk series on PSN. I also knew that the XNA tools and focused development on a single platform would make it easier to achieve my goals.
Eleven months later, I completed the seventh and final game in the series, with the last five being released in consecutive months. The titles in release order: JoyJoy, Crossfire, Inferno, Fluid, Fireball, Crossfire 2, and Ballistic. Even though I've moved on to a more traditional indie route (larger games, more hype, publisher support, and a higher price) it was still an invaluable experience and an accomplishment to be proud of.
Throughout the article are interspersed "what went right"s and "what went wrong"s from each of the games in the series.
What Went Right
1. Scope Control
Large games take more time to make than small ones. Good games usually take longer than bad ones. So the number one priority in deciding which game ideas to work on: scope. It's amazingly hard to keep a game simple and not to turn it into something epic, so the smaller and more focused the original idea, the better.
With the Radiangames series, the struggle between keeping the games small and making them good was never-ending. I always started small with my ideas, then spent the rest of the time trying to make it good. That invariably required adding more to the core concept, or changing the core concept to be something bigger, but in the end it worked out.
To help keep focused, I avoided hard technical problems whenever possible. The games were all 2D and had no online multiplayer. With development on a single platform, I was also able to limit the games to one resolution (1280 x 720) and a fixed-step framerate.
Applying good scope control also meant choosing game concepts that could build off the elements and ideas in previous games. Every time I started a new game, I'd make a copy of the project it most closely resembled and build from there. Though I didn't have a singular engine I was building, keeping a similar game structure (with improvements only where absolutely necessary) meant I could focus more on game design and less on game code.
In the end, sticking to my original plan of creating simpler games was the main reason I was able to release so many quality games in a short timespan.
Right or Wrong: JoyJoy
Right: Safe Design. Being the first game in the series, I went for a safe core design with JoyJoy to reduce risk and focus on getting the game done quickly. The safe design also allowed me to work harder on writing code that was solid and re-usable. Though there were a few major improvements made in subsequent games, the basic code structure was used throughout the series.
Playtesting doesn't make a game great by itself, but it essentially guarantees your game will be enjoyed by more people than it would have otherwise.
With the Radiangames series, I employed a simple but effective two-pronged approach to playtesting: I would occasionally get a few new players to play the game while I observed, and I would send builds to other XBLIG developers for feedback. There's nothing unique about this approach, but I found the two methods combined to be effective at catching a wide range of issues.
The other half of playtesting is listening to the feedback and figuring out what to do with it. My experience with retail games made it much easier to deal with any negative or confusing feedback.
As most of the playtesting comments went directly against scope control or were painful to hear, I'd take a little time to contemplate the best way to respond or address the issues rather than reacting defensively. Usually I'd end up agreeing that something had to be done to make the game better, and that's what playtesting is for.
3. Restricted Art Style
It's difficult to be good at all aspects of game development, but as a solo developer you have to do your best. Since I knew art was one of my weaker areas, I needed to find a way to create compelling-looking games without a lot of actual drawing skill. Through a series of not-entirely-planned steps, I eventually settled on an art style that required nothing more than white bitmaps that would be colored in code. The bitmaps had some detail in the alpha channel to give them shape, but the most common graphical elements in my games were usually solid and gradient primitive shapes such as circles and squares.
This simple art pipeline worked well on its own, and I was able to add some additional layers over time, including a bloom shader and a refractive/reflective shader (using normal maps generated from xNormal). I also found a vector drawing program (Draw Plus) that felt natural, allowing me to add more subtle style to the player and enemy graphics in the later games.
Though finding the exact rules, restrictions, and workflow was a major struggle, in the end each game's art had its own twist on the Radiangames style.
When Inferno eventually morphed into "Gauntlet meets Geometry Wars", I kept the powerup system from the previous version of the game. That system resembled Gunstar Heroes' powerups. The player had two slots that could each hold one of seven always-active powerups. I thought the system provided nice gameplay variety without being too time-consuming to implement.
But after receiving overwhelming feedback from both local and online playtesters, I switched to a more traditional upgrade system with a week left in development. The new system gave a better sense of progression, a reason to collect gold other than score, and a more obvious view of what was happening with the ships' abilities. I resisted initial feedback because it meant I had to add a new upgrade screen (that would work with 1 to 4 players) and rebalance pretty much the whole game. I thought it'd be a lot of work -- and it was -- but the result was well worth it.
Inferno's final upgrade system was such a key to its success that I made a large upgrade system the main new feature for Crossfire's sequel. Then when Ballistic's gameplay felt too one-dimensional, it received a last-minute upgrade system as well.
In the end, I think the upgrade systems' popularity proved to me that most modern gamers want more than a pure score attack experience. They want something that rewards them and keeps them invested in playing, something beyond the intrinsic rewards of getting better and competing against your own (and others') best scores.
Right or Wrong: Inferno
Right: Overscoped. Inferno just kept growing the more I worked on it. It ended up being the largest game in the series, and also the best-rated and best-selling. Even now I can't help but play the game and see lots of ways to expand and improve on the basic idea. Some game concepts are naturally larger than others and can't be chopped down to fit a schedule.
5. Passion vs. Innovation
If there's one thing game journalists and fans love to talk about, it's innovation and originality. A game is only as good as it is innovative, or so we're conditioned to believe. I even used the word in my press releases on occasion, and I often agonized (in a good way) over how to make my games stand out more.
But I was trying to make a game in a month or two, so innovation could not be my primary goal. Instead, my focus was on games I wanted to play, and making them the best I could as quickly as possible. If the games had some innovation in them (Crossfire, Inferno, and Fluid did, in my opinion) that was a bonus. But I knew that getting a game done is hard, and by making games that I had a passion for, I'd be more motivated to play them, improve them, and most importantly finish them.
Right or Wrong: Fluid
Wrong: Missed Potential. I entered Fluid into a couple indie game competitions, and talked to at least 3 publishers about the game at some point. But I wasn't fully committed to turning it into something bigger, and the game never fully realized its potential. I still can't shake the feeling that I missed something special with this one.
What Went Wrong
1. Pricing Indecision
There are many factors that go into choosing the optimal price for a game. When I set out to make the monthly series, I thought I would be best off going with a $1 price in order to create a bigger fanbase that would steadily build. By getting each subsequent game into more gamers' hands, I'd increase the chances of the next game being an even bigger hit. Or so the theory went.
Instead of sticking with my plan, I wavered with the second game, Crossfire, and went for a $3 price. JoyJoy had been selling okay, but not great, and I could see it taking a very long time to make money with $1 games.
Though Crossfire's revenues were higher at launch, I could tell the sales curve was going to drop off quickly due to the game's placement on the Top Rated list. So I switched back to a $1 price for Inferno (the third game) and dropped Crossfire's price to $1 as soon as possible.
However, the reason some $1 games do well on XBLIG is because they get into a positive feedback loop on the Top Downloads list, and my games tended not to benefit from this phenomenon as much as I'd hoped.
The XBLIG marketplace favors games with an off-the-wall or unique hook, while hardcore arcade-style games are more of a niche than I realized. If 360 players want arcade-style games, they tend to look on XBLA first, where there are plenty of high-quality arcade titles.
Given all that I know now, I would have released all the games at a higher pricepoint.
Right or Wrong: Crossfire
Wrong: Difficulty Curve. Crossfire starts a little slow to ease the player into the unique warping mechanic, but then overwhelms most players a little over halfway through the game, and defeating the last two waves in the game was an exercise in frustration. This uneven difficulty curve drove off all but the most dedicated players, and went completely against the player-friendly features that were so prevalent in JoyJoy.
2. Mental Beatdown
Making games is hard in so many ways. It's hard to choose which idea to work on next. It's hard to hone a prototype into a solid direction. It's hard to bring all the different elements together into a cohesive design. It's hard to write music, create art, and code. It's hard to hear people say your game sucks when you thought it was almost done.
It's hard to come up with good box art. It's hard to wait for sales numbers and press coverage and user ratings. It's hard to start a new game just after launching one. And most of all for me, it's hard to work all the time knowing you're slowly losing money and will have to give up on your dream if things don't go better.
The last one got to me because of my family situation (i.e. I have one -- and a house), but every one of those challenges takes a toll on a person. Going through them every month or two with only yourself to blame is a good way to drive yourself crazy. Granted, if the games had been financially successful, I probably wouldn't have cared so much, but I probably would have slowed down a little more.
Right or Wrong: Fireball
Wrong: Ignoring the Market. Exhausted from Inferno and Fluid, I made the game I wanted to make: An expanded version of Pacifism mode from Geometry Wars 2. Due a combination of issues, including a garish art style and unforgiving gameplay, it never got any sales momentum and was my biggest flop. It's my wife's favorite game though, so it wasn't all for naught.
3. Minimal Personality
There was a great article in Game Developer magazine this past year about how to inject personality into your games. If you've played my games, you may have guessed that I never read it, but you'd be wrong. I just rarely spent time thinking about it.
As I worked on the games, my brain became wired into improving efficiency, solving the day-to-day design issues, and creating a cohesive visual and audio style. I instinctively avoided worrying about how to enhance the gameplay experience with "superficial" animations or quirks, and I didn't even consider the context for my games as a potential selling point. Yet it wasn't just an accident that those things happened. Cutting out those extras also supported my fast production cycles, and not doing them let me avoid things that didn't come as naturally.
But obviously, those "meaningless" things do matter to games, particularly to drawing in players who don't necessarily care about the core gameplay experience above all else. I tried to address these issues to some degree by sticking to a visual and audio style for each game, and that definitely helped gain some fans. But overall I failed in any attempts to inject sufficient personality in my games or appeal to a larger audience, and my choice of audio-visual style may have actually turned off more potential fans than it helped gain.
4. Branding Baggage
One of the most common questions/complaints I get is regarding the names of my games. To be more specific, the games all have "radiangames" as the first part of the name, so when you look for them on the Xbox, you'll see "radiangames JoyJoy", "radiangames Crossfire", and so on. While this helps people who want to find all of my games by looking under "R" (and there's data to support that it helps a few people a day do this), I heard many complaints about players having trouble finding specific games because they were looking under "I" for "Inferno", for instance.
I wasn't prepared for the awkwardness it created, particularly with regards to the press. It felt odd to continuously be writing my company name before my games' names, and the press didn't really like it that much either. Some press members even asked how they should refer to the games in their coverage, and that's not something they should have to think about.
Another thing I hadn't considered was how much more important the box art logo was than the name prefix. When looking through the lists of games with box art (the most common way to browse XBLIG), it's very obvious which games are from Radiangames because of the blatant orange logo, making the name prefixing pretty much unnecessary.
Given a choice to do again, I would definitely let the games' names stand on their own instead of branding them.
Right or Wrong: Ballistic
Right: Applied Knowledge. With Ballistic, I stopped trying so hard to do something different, and instead focused on using what I could from previous games to create a very polished game that was appealing to a large audience. In that regard, Ballistic was a big success and it had an Inferno-like first month of sales.
5. Too Many Modes
Even though I know that most gamers play the primary mode of a game and nothing else, I still spent too much time on secondary modes for most games in the series. When thinking about game design constantly, it's easy to come up with variations on a game's main mode or small spin-offs that use similar mechanics. So I often gave into temptation and would turn those variations into bonus modes or challenges that weren't part of the main gameplay experience. They were big bang-for-the-buck value-adding features... or so I told myself.
What's more difficult is combining those somewhat disparate modes and variants into one cohesive experience. With every game except Inferno and Fluid, I added extra modes that took somewhat significant amounts of time to create and test, but didn't fit within the main gameplay experience. Even now I'm not sure which modes or gameplay ideas I could have combined or cut, but the fact that I never really considered those options thoroughly is something I need to improve on in future games.
Right or Wrong: Crossfire 2
Wrong: Rushed Release. Crossfire 2 was rushed to release during the Indie Games Winter Uprising, a developer-created promotion designed to draw positive attention to XBLIG. Though the game improved on the original Crossfire in many ways, I missed one major issue in playtesting: the upgrade screen appears too frequently, breaking the flow of the hectic gameplay and hurting the game for players who wanted more of a challenge.
It's hard to say if choosing XBLIG as my primary platform was a mistake or a wise choice. Though the platform has problems and is rarely financially rewarding for developers, I don't know how else I could have made the monthly series. The basic concept of XBLIG is fantastic, and the tools are great too, so it's a bit of a shame it's been relegated to the dark corners by Microsoft and the gaming community in general.
Though the Radiangames series never became the financial success on XBLIG I was hoping for, in terms of fan reception, support from the community, and actual output I couldn't have hoped for more. Not only did I create 7 games to be proud of, I also learned a lot about myself and how to make games, and found a great group of developers, fans, and dedicated press that I'll always share a bond with. I may be done with Xbox Live Indie Games, but the spirit of the first (and hopefully not last) Radiangames monthly series will continue to live on in bigger and better ways.
Release Dates: May 15, 2010 - January 28, 2011
Length of Development: 11 Months
Lines of Code: 116,019 (7 games, not including 3 level editors)
Development Tools: Visual Studio C# Express 2008, XACT, DrawPlus, Corel Photo-Paint, xNormal, Stomper, Reason, ReBirth
Projects Cancelled: 1 (Blaze, a Sonic-like platformer)