Just another post-mortem: Echolon
Entry: Echolon (check it out here)
Timelapse: http://www.youtube.com/watch?v=U_toZ_YeZ6U
Summary:
Nothing is a more effective motivator than a hard dose of reality and time constraints. Also it’s way easier doing something you’re familiar with than trying something new in 48 hours.
What Happened:
Developing what eventually became Echolon was a tale of two games. Saturday was spent on one idea that failed to pan out and Sunday was spent primarily on what ended up being my final submitted game for Ludum Dare. This essentially means that I gave myself around 9 hours to go from near-scratch and create Echolon.
This was my 3rd attempted Ludum Dare and my second consecutive successful one (along with caverns from Ludum Dare 15). There were two points during the weekend I was unsure if I would finish, but some radical redesigning and acceptance of certain limitations allowed me to push until the end.
The initial game attempted to approach exploration in an indirect way. The basic idea was that you as the player never went exploring yourself. Instead you used a rover/robot that would explore in your stead. There would be little to no graphical component, most of the exploration would be conveyed in text reports and other feedback. As you explored the landscape outside you would find equipment to modify your rover and allow it to explore new areas, carry new items, and extend its capabilities in general.
I spent most of Saturday working on the background mechanics and trying to figure out the best way for the player to interact with the rover. The game ultimately started becoming heavily UI-dependent which was the worst thing possible. In my experience user interfaces tend to be time-consuming and tedious to craft and ultimately are the biggest time sinks in any project I’ve done. The trade-off between UI creation and time is pretty piss-poor.
The first game on the left and an early image of Echolon on the right
This wasn’t help by overall poor time management on Saturday. Part of Saturday night was spent going to the movies with my family (Princess and the Frog) and doing other tasks unrelated to Ludum Dare (you know, the important stuff). So by the time I had crashed on late Saturday night/early Sunday morning I basically only had the fragments of a game in hand.
Waking up Sunday it quickly dawned on me that at my current rate the first game would never be complete on time. After this realization my mind quickly scrambled for a second idea that could be implemented in a short amount of time. I ended up thinking about a game where the player explores a cavern with no lighting except his own bullets. I added on the idea of illumination through shooting along with some basic reflection mechanics and I had a game idea to implement.
Afterward development was fast and furious in the final 9 hours of the competition. The basic game controls are pretty much Asteroids and Flixel pretty much allows that kind of gameplay to be implemented immediately. Things like shooting and reflection were also implemented with little to no problems and basic enemies quickly followed.
Echolon in its near-final state
The biggest obstacle in that 9 hours was spent on creating a procedurally-generated level. I originally had planned a more ambitious generator that would construct various rooms of differing sizes that would be connected with hallways at various points essentially creating a large maze. Over time this became less and less feasible as I had a hell of a time figuring out how exactly to create “screens” that didn’t conflict with each other. I spent roughly an hour on this before abandoning it near the end for a much simpler solution. What I ended up implementing instead was one large room that had random blocks placed inside. There’s no real smart logic to the placement, but overall the result was passable for the purposes of the competition.
To demonstrate how much I was pushing time in this case level generation was complete with about 45 minutes left in the competition. The final 45 minutes I had to create a pretty terrible menu screen, victory/game over screen, add the final sound effects, finish up the UI, and make sure everything sorta worked. I pretty much finished with 2 minutes left before 9 PM.
The big question in my mind is if things had gone exactly right would I have been able to complete my original idea? Possibly if I had worked the full 48 hours, but even then I feel like nothing of the scope I had imagined would have been completed. The idea was a radical departure from any thing I’ve done before – attempting to create a limited simulation game with a heavy emphasis on stat manipulation and menus. When this failed to pan out my second idea relied on simple game mechanics I’ve done before and could easily mock in before focusing on newer ideas. I still really like the idea however and would like to return to it in the future.
Overall my feelings on Echolon are pretty muted. It’s not very fun nor very playable and it comes off slight (although for 9 hours I don’t know if I personally could have done better). I like some basic ideas, but having even an extra hour to tweak some of the gameplay variables would have been nice. I had some additional ideas I would like to implement and I wouldn’t bet against me returning to the game in the future.
What Went Right:
- There was a game. After everything I managed to submit a final game. That by itself is a win in my book.
- Accepting failure. After realizing the first game would never be completed on time I completely shifted gears to developing Echolon and never looked back.
- Not rebuilding the wheel. Echolon was a game well-suited to the tools I was using and also similar to other games I’ve created. Getting to first playable took little time.
- Rush rush rush. With 9 hours pretty much uninterrupted for coding I managed to get in the zone and almost never left it until the deadline.
What Went Wrong:
- The first game. Oh god did that not work out.
- No playtesting. Echolon as you may note is not balanced in any sense. Chalk that up to limited dev time.
- Remaining overambitious the second time around. Things like trying to implement the full-fledged procedural level generator and an upgrade system was far beyond the possible scope of a 9-hour game.
What I Ate:
4 Eggos, a Jimmy Johns sub, some chips, popcorn, lots of tea, some lemonade, and a roast beef sandwich.
Finally:
Still an awesome experience. Plus I made a game! Score!
Tags: postmortem