LD24: Evolution Game Post Mortem #LD48

My last few Ludum Dare entries have missed the main compo deadline and had to be submitted to the Jam, but my entry for Ludum Dare #24 was finished in time to get judged along with the other 1,000+ entries. That was a big win for me, since my game development in general has felt quite slow.

The theme for LD24 was Evolution, which always makes it to the final round and has never won as a theme until now. I ended up creating a simplistic side-scrolling shooter. It definitely wasn’t my best entry, but health problems made it difficult to do better in the 48 hours.

What Went Right

1. Leveraging existing tech

   Since LD#18 (“Enemies as Weapons”), I’ve been slowly building up a game engine for Stop That Hero!, my casual strategy game that lets you summon minions to fight heroes who want to put a stop to your evil empire. As I am using test-driven development to work on that game, I ended up with code that is relatively loosely coupled and cohesive. It was quite simple to gut out the STH! parts and leave behind a way for me to immediately create a menu, game play screen, and ending. Instead of spending the first six or twelve hours trying to get an SDL window to shut down properly, I could concentrate on more important things.

   My engine isn’t super powerful. People using Unity or Flash had a huge advantage since so many components are fully-formed and well-designed. Still, I had code that I was able to use in multiple games, and I knew how to use it. I was able to put together my game fairly rapidly.

2. Better familiarity with tools

   My art program of choice is GIMP. While I’ve been known to doodle on paper since I was a child, I haven’t been very practiced with pixel art or image manipulation. I don’t know how to do 3D modeling very well, but if I did, I would use Blender.

   Over the years, I’ve learned how to use GIMP to create some decently functional art and ads. Selecting shapes, making use of layers, and knowing how to manipulate color goes a long way. What I create is not production-quality, but it works for an LD48.

   

   Sound effects were made using DrPetter’s sfxr. Even though it is really easy to use if you just want to generate random sound effects, knowing how the various parameters can be tweaked helps a lot in getting an effect to sound just right.

3. Iterating fairly well

   My biggest victory came from iterating, even if I could have approached it more intelligently.

   After working out some ideas on paper, I had a basic design for a shooter.

   

   I wanted the controls to be simple. I came up with some different ways for how enemies would evolve, as well as ways the player’s tank might evolve.

   I wanted different kinds of enemies that came in different sizes, used different movement patterns, and attacked the player in different ways. I even had an idea for a boss character.

   

   I also thought of Evolutionary Upgrades, aka Power-ups, for the player. Some impacted the tank’s weaponry, such as a spread gun or homing missiles, while others affected the tank’s size or armor.

   

   Once I had an idea of the kind of game I wanted, I set to work. My initial list of tasks:

   – get the player’s character in the game
   – make it controllable
   – add obstacles (most likely boulders)
   – make collisions between the player and obstacles deadly
   – add an enemy
   – create a wave of enemies
   – create a way to modify the wave of enemies so each enemy evolves in some way

   First, I got a scrolling background. In hindsight, maybe this part could have been left until later. My next goal was to get a controllable tank on the screen, complete with the ability to fire bullets. Having something controllable that early meant that throughout the development of the rest of the game, I could get a feel for the controls. As the game came together, the tank’s controls were updated a few times. I originally had the tank’s movement a bit slow to accelerate, as switching a tank’s directions is probably really hard, but I found it was more annoying than fun. It felt too sluggish. I made it more responsive in the end, and it was better for it.

   Next, I added boulders, followed up with collision detection between the tank and boulders. Now tanks have to avoid obstacles, making the side-scrolling environment a bit more maze-like. In terms of the theme, however, perhaps these obstacles were not the best thing to add earlier, although I did have plans for enemies to interact with boulders by pushing them toward the player if they collided with them.

   Finally, I added killable enemies. The enemy was pretty basic. It moved in a straight line toward the player, didn’t attack, and moved through boulders.

   I realized that I was not going to get all of the enemy types and upgrades in, so I focused on making sure what I had was as finished as I could make it. I added enemy waves, which added more enemies and made them harder to kill as the player progressed. I added a score so you could see how well you’re doing, and I even had time to make some points visibly pop up when you kill an enemy. That last bit was a small aesthetic change, but I think it polished the game up quite nicely. Great bang for buck.

   Now, there are some things I spent time on that I could probably have left until later. The scrolling background wasn’t really necessary, was it? And the enemies were supposed to be the main focus of the design, so why did I work on boulders first? I definitely could have prioritized much better.

   Still, what’s key here is what I didn’t spend time on. I didn’t spend time making tank upgrades, which is good because I didn’t have a need to upgrade the tank. I didn’t spend time making lots of enemy types, which is good because I didn’t have the time to intelligently figure out how they should be introduced. I didn’t spend time to figure out boulder/entity interaction, which is good because who knows if it would have added anything?

   By getting something playable and iterating, I was in a position to reduce scope to finish the game by deadline. Along the way, I almost always had something playable that I could submit by the deadline.

What Went Wrong

1. Dealing with Back Pain

Shortly before the theme was announced, I was working to get the next release of my casual strategy game Stop That Hero! out the door. Between that project and my responsibilities as President of the Association of Software Professionals, I was sitting at my desk a lot, and it was taking its toll on me. I could feel some tightness in my hip, and so I decided to try to setup my environment so I was standing more.

I placed a container under my keyboard, and it was raised to the perfect height to let me stand while I work. Everything was great, until I had to use the mouse for some reason, which was still on the desk. I leaned to the right to reach for it, and I didn’t realize I was going to be mousing like that for long. Standing in such an awkward way, coupled with the tightness I was experiencing from sitting for so long for days on end, I ended up with some very, very annoying back pain.

The Friday before the compo, I went to a massage therapist since I figured it was just tightness that needed to be massaged away. After the compo, I went to a chiropractor because it clearly wasn’t getting better and in fact felt worse.

But during the compo, I was taking a lot of breaks. I could sit or stand for a period of time, but when the pain started getting distracting, I’d go lie down on my back for a time before I could start working again.

I lost a lot of productivity to that pain, and I’m still recovering from whatever happened to cause it.

2. Being Uncool

   My entry didn’t get a rating. It seems that in the time since my last LD, there was a change that if you don’t have a high enough cool rating, you don’t get a listed rating. As I was a bit busy and couldn’t dedicate the time to reviewing other games post-compo, I didn’t get many reviews done. Reviews translate into coolness. With over 1,000 entries, people aren’t expected to review everyone’s games, but there is an expected minimum you should rate. I rated a few immediately after the competition, but between not being able to play Unity-based games and having other priorities, I didn’t get back to it. My only listed rating is Coolness, and I was ranked #986 with a rating of 22% cool.

   There were quite a few 100% Coolness ratings, and those people are awesome. Or they just have a lot of time. Either way, I’m glad they exist.

3. Doing a poor job of including the theme

   Evolution is always a contender for the finals of theme selection, and I was caught off-guard when it actually won.

   I could have tried to design something involving random genetic changes in entities and seeing which one adapts better to changing circumstances, but it sounded too obvious and also too open-ended. I wanted to try to keep my entry simple and straight-forward since I wanted to submit an entry to the compo instead of the Jam.

   While I had some design notes that I called “evolution,” such as tweaking variables to create new enemy behaviors and types, it really wasn’t evolving so much as creating variations, the kind of thing you’d see in any shooter and most games in general.

   Perhaps I should have thrown a few more ideas at the wall before settling on this side-scrolling shooter. My wife suggested the idea of an “evolving door” which has the benefit of being pun-tastic, but I couldn’t find a good way to incorporate it into my existing design. Judging by the variety of entries, I could have been a bit more creative.

What I Learned

1. Iterate like you mean it. This is where Agile software development experience really comes into play. If you can create the simplest bit of value, and then build on it, you’re going to ship. If instead you build up scaffolding code to prepare to provide some unknown, vague value, you’re probably going to get mired in delays.

   In previous competitions, I’ve found that I didn’t introduce basic interactivity until a bunch of other things were ready, and I’ve suffered as a result. By getting something player-controllable right away, I was able to not only get a game around it much more quickly, but I was also able to make small changes to the controls until it felt right.

2. Having good tools and knowing how to use them is great for productivity. To iterate quickly, you need to be able to produce functionality quickly.

   I used to try to create decent art, but between not having much practice and not being familiar with GIMP, I would spend way too much time on art, and in the end, it still looked really bad. I wasn’t getting the return on investment.

   Now, my art skills are still not production-quality, but they are passable, and I am able to create decent programmer art in minutes when it used to take me hours.

   Also, experience counts in general. I write way better code today than I did just a few years ago, and I do so much faster. I terrify myself when I look back on earlier Ludum Dare compos and read through my code.

3. I need to take care of myself. Up until the compo, I had been doing yoga and taking regular walks. However, poor posture in front of the computer and sitting for way too many hours at a time in those postures did me in. It’s over a month later, and I’m just now feeling fine enough to start stretching and taking walks again. Maybe I need to seriously investigate the standing desk option and even look into a much better office chair.

Summary

If I could do LD#24 over again, what would I do differently? I’d spend more time upfront trying to create a design better suited for the theme that is also simple enough for me to make. I’d make sure my list of tasks was prioritized so that at all times I was working on implementing something that served the core design. And I’d make sure that I had set aside time after the compo and Jam to rate other games. People worked hard on their entries, and with over a thousand of them submitted, it’s unfortunately easy to get buried. I think the coolness rating does a great job of making things fair, and the name is perfect. I want to be cooler next time.

Being able to get a game submitted is still a great feeling, though. In 48 hours, I created a game where there once wasn’t one. Next time, I should hopefully be healthier and able to focus more on game development, and my next entry should be not only minimally complete, but actually enjoyable to play. I’m still aiming for getting #1 in the Overall category, and while it feels like I’m a long way away from challenging other entrants for that position, I’m definitely way closer than I was when I was participating in LD#11 four years ago.

How did your LD#24 go?