Peter Stock, creator of Armadillo Run, was kind enough to field a Fun-Motion interview regarding his debut physics games. He covers the project’s inspiration, his design process for developing and testing levels, and his future plans. It’s a great read, and thanks again to Peter for his time.
Fun-Motion: Congratulations on Armadillo Run’s release! I hope the game does well for you. Thanks for taking the time to do this interview. I know a lot of people are curious about the game’s development and eager to get a look behind the scenes.
Peter: Thanks - the launch of the game has gone really well and I’d like to thank everyone for their support.
Fun-Motion: What inspired you to create Armadillo Run? The game’s concept-building things-bears obvious similarity to other games in the genre. What motivated you to create your own title?
Peter: In 2001, a friend of mine showed me Bridge Builder, which we both thought was great fun. A little while later, he introduced me to Stair Dismount - probably the most fun I’ve had pushing anyone down a flight of stairs. While these are totally different games (one’s about construction, the other’s pretty destructive), I think they’re both fun because of the physics - trying out different things and seeing what happens is fun.
Around this time, I thought that I could create something unique by taking the concept of construction (Bridge Builder was essentially the simulation of a static structure) and adding some dynamics (which was what made Stair Dismount so great). During my university course, a student implemented a physics simulation of a ‘marble run’ for their third year project, which I also drew some inspiration from.
I didn’t do anything with the idea until mid 2005, when I left my previous job and followed my wife across the globe - I decided it was a good time to try out my idea. Processing power and 3D rendering capabilities have advanced so much in the last few years, which makes games like this feasible - it seemed like a great opportunity.
The design of Armadillo Run is very interesting in that it encourages experimentation and a certain sloppiness in building contraptions (at least, it does for me). Was that intentional? Did you start out with a certain “feel” in mind when you began development?
Right from the start, I tried to create a game that would allow different solutions, so a level could be completed in ways I hadn’t expected. Any game that uses physics simulation as part of the gameplay naturally suits this concept really well, and, like you say, experimentation allows players to learn more about how the game world behaves.
All games need some kind of reward/penalty structure, which sets them apart from toys. I thought that making the more advanced materials cost more would be a good way to offer a trade-off decision when choosing how to construct levels. This does mean that some of the more optimal solutions are a little under-engineered!
Some people had suggested that I should make it clearer how to complete each level or restrict the number of choices, but I didn’t want to do this because it would lead players to a single solution rather than encourage them to find their own. I tried to allow players a free reign in their use of materials (the only restriction being the level budget), which I hoped would encourage experimentation. The wide choice does allow some sloppy solutions, but I like the elegance of setting down a simple set of rules - if a solution passes the rules, it’s as valid as any other.
The available materials in Armadillo Run are much more varied than in other physics building games. Did you try out any other objects/materials that didn’t make the final cut?
I decided on the materials by working back from what I wanted to achieve. I wanted to keep the simulation 2D (to keep the interface simple) but I also wanted more flexibility than having everything occupying the same plane. I created what is essentially two planes - some materials go in the centre and collide with the armadillo, others are placed at the edges and don’t collide with the armadillo. The two basic types of materials are the flexible ones and the rigid ones - combining flexible and rigid with two planes gave the rope, cloth, metal bar and metal sheet materials.
I wanted to make more use of dynamic structures, so I needed some materials that would be able to store and release energy (elastic and rubber) and something that would create energy (rockets). I wanted to keep the set minimal, so I didn’t develop any others. Combining the materials with tension and timers seems to be enough to create many different things.
What was the creative process like for creating new levels? Did you start with a particular solution in mind and build backwards from there, or generally just experiment with various starting points and configurations?
I had a few general concepts that I wanted people to play around with (force, momentum, balance, leverage, energy transfer and designing a stable structure) that I incorporated in many of the levels. But I also wanted some more freeform levels, which were totally blank except for the anchor points (the ‘void’ levels). For the normal levels, I usually worked backwards from a solution, sometimes involving some sort of gadget, but I tried to keep multiple solutions in mind. For the void levels, I tried to find challenging configurations that were significantly different from each other.
How hard was it to balance the game’s difficulty and level progression? Did you utilize a stable of beta testers or mostly go it alone?
Apart from the early levels, I didn’t create the levels in order. I played around with an idea to see if it worked and if it did, then the level got put aside for inclusion. I carried on until there were enough levels then I tried to put them in a sensible order.
Later on in the development, I used some of my friends as testers - I used their feedback to alter some levels and do some re-shuffling.
Did you have a particular market in mind when you created the game?
Not especially. I saw that there were some people (like myself) who liked the physics-based games that were currently available (these people seem to be avid gamers). I definitely wanted to appeal to this market, but I also tried to keep the game simple enough to appeal to occasional game players too. I didn’t specifically target a young, old, male or female audience, but I tried not to exclude anyone.
What was the development timeline like? Were you able to work on it full-time at any point, or was it a part-time hobbyist effort?
The total development time from the project start to the release was nine months. The development of the prototype took about two weeks, after which I decided that the project was feasible. The physics engine took about two months, but it was another six before the game was on sale - I seriously underestimated how much time it would take to turn an interesting physics simulation into a game!
Right from the beginning it was a full-time project. I had to be both the developer and the project manager, so I drew up a development plan and assessed the project from a business perspective (answering questions like “is the project feasible?”, “what are the risks?” and “how’s it going to stand out from the competition?”). I didn’t find it too difficult to switch roles and assess the game with a cold business eye, because I was providing the funding!
Your background is computer science. Is Armadillo Run your first game? If so, did you find anything about game development particularly frustrating? Did you find any physics game programming tutorials or references especially useful?
Yes, Armadillo Run is my first game (apart from an unreleased kite-flying simulation and a clone of Tetris for the PS2 using the YaBasic interpreter). In my previous jobs I have done a fair bit of programming and have also had some experience in computer graphics and code optimisation, so I didn’t have too much learning to do during the project.
Building on established libraries like OpenGL and FMOD meant great performance, reliability and documentation. I did find some portions of the OpenGL API where some drivers didn’t follow the spec and this was a little tricky to track down.
I did a lot of research on physics before I started the implementation - on the Internet as well as in books at the local university library. I found Chris Hecker’s game physics articles particularly useful as well as ‘Physics for Game Developers’ and the ‘Game Programming Gems’ series of books (they contain small chunks of information rather than in-depth tutorials). GameDev.net is a great development resource, which has a lot of valuable information on all aspects of development - physics, graphics, sound and game design.
Finding this site showed me a load of physics-based games I hadn’t seen before, which was good for doing background research - and playing them was a lot of fun!
Your website has some community functionality for users to upload new levels and solutions to levels. How was the community response been? Have you thought about running competitions or do you have any plans for other community events/features?
I decided that this functionality would be expected, and it’s beneficial to me because it extends the appeal of the game by offering new content to players. The community response has been good (some of the levels are very impressive) and someone’s even created a forum for people to discuss the game. I’ve received a lot of valuable feedback from players including suggestions for new features, some of which I will incorporate in updates to the game. I plan on running some competitions in the near future.
How have sales of the game been? Has it met your expectations? What are your goals with the game’s performance in the marketplace?
Sales have been better than I expected, although my expectations were maybe a little pessimistic. I don’t really have a goal, apart from people enjoying the game. I decided at the beginning that I would continue as an independent game developer if Armadillo Run was a success, otherwise I’d go back to normal employment - I consider the game relatively successful, so I think I’ll stick to game development for a while longer.
Any sequels or other games (physics or otherwise) in the works? What’s next for you as a game developer?
I’d like to do something else that’s different. I think that’s the only way for people like me to make games these days - there’s no way a one-man-band like me can compete by trying to out-do the competition at what they do very well (like the RPG, FPS and RTS genres). My development strategy is to make games based on novel gameplay mechanics - to make games that don’t have any direct competition.
If I were to make a sequel I think it would have to be more than a minor update to justify players making another purchase - I don’t just want to tack on something small. I don’t currently have any plans for Armadillo Run 2 because I don’t have any ideas for expanding the game that are good enough, feasible and that would improve it enough to justify it. I’m going to incorporate some updates in minor revisions that will be available free to everyone who’s already bought the game.
Having said that, I do have a couple of ideas for my next project - one is another physics game based around liquid, the other’s a music-based (Bemani-style) game with elements of traditional platform gameplay. These ideas might change - the early part of development is more about exploring and refining ideas than actually making a game for me. I’m looking forward to trying out some new things and seeing what happens.
Do you play many physics-based games? Any favorites?
Absolutely! Jetro Lauha’s games (including Stair and Truck Dismount) are firm favourites of mine. Since finding your site a year ago it’s given me a whole lot more to try out - I really enjoyed Elasto Mania and Ski Stunt Simulator, which I started playing just before I went skiing for the first time - I’m sure the experience on the game gave me a head start on the jumps!
What inspires you (favorite games, activities, or whatever it was that started you on the path to game development)?
I’ve always enjoyed games, particularly some from the 16-bit days. If anyone hasn’t played Super Mario World, they should pick up a SNES from eBay and play it - it’s a timeless classic. My other favourite games include Devil Dice (an excellent puzzle game for the PlayStation), Gran Turismo 4, Final Fantasy IX and Freecell.
I like maths and general problem solving, which is basically what programming is about. Having an interest in games attracted me towards games development. Outside work, I like to spend time outdoors hiking, cycling and (my new-found love) skiing. If I can ever afford it, I’d like to do Gran Turismo in real life, if only for a day!
Thanks again for taking the time for this interview!
[Editor's note: Peter has also penned an excellent postmortem of Armadillo Run for Gamasutra]
- Heading Down Under!
- 2007 IGF Finalists Announced
- Builder Physics Games Evolved: Armadillo Run
- Indies Have Opportunity with Physics Games
- Beautiful, Frustrating Puzzle Physics