About Icelus (twitter: @IcelusTheFallen)


Ludum Dare 37
Ludum Dare 34
Ludum Dare 32
Ludum Dare 26

Icelus's Trophies

Icelus's Archive

[Post-mortem] LD32 Compo: Transmogrify!

Posted by (twitter: @IcelusTheFallen)
Friday, April 24th, 2015 11:11 am



What went right:


  1. Spent one hour Friday night brainstorming, came up with a core game mechanic, and stuck with it. Nailed the mechanic, in my opinion. I had two other cool options, but I discarded them and locked ont the transforming-gun I ended up with. Deciding quickly and letting distractions go so you can move forward is important for jamming.
  2. Spent the bulk of my time (probably >75%) on pure development, gameplay, testing, bug fixing, and tweaking components. I was able to focus pretty well and stay on target with what I wanted.
  3. I accomplished all my critical goals for the game, and some of my secondary ones as well. Could have done better with time allotment for some of the minor areas, but overall I’m happy with my results.
  4. Quickly built out the core mechanic- having a giant gun that turns enemies into objects temporarily that you can use as platforms to advance. By Saturday at lunchtime, I had the player movement, the first enemy moving and killing the player on contact, the gun turning the enemy to an ice cube, and him de-thawing. Focusing on this core element first and testing the hell out of it paid off for me this jam.
  5. Introduced one new mechanic per level for pacing for the first three levels, and built out the gameplay organically. Also just the fact that I managed to build four levels I was happy with, and test it, was a success for me personally.
  6. I used Construct 2 because I am very familiar with it, and it’s stupidly fast for iterating. Rapidly prototyping is probably Construct 2’s shining feature. I was able to compile over a hundred times in two days to test out small changes, bug fixes, and iterate very fast. Rebuilding on a local host takes about one second each time, which is great. I know C# and Java, and I’ve worked with Unity and LibGDX. Could I use those for Ludum Dare? Sure. Do I love coding in Java? You bet. Would I be able to iterate as fast, or spit out the same level of gameplay and features as with Construct 2 in a constrained amount of time? No way. Construct 2 will probably be my prototyping tool for the foreseeable future because I get results with it, and I highly recommend you factor in speed and familiarity with the tool you use for game jams. Unity and LibGDX are great tools, and I will use them for full games most likely. But for me, Construct 2 is super awesome for rapid prototyping.

My entry if you want to play or vote:




What went wrong:


  1. I bit off more than I could chew with the second enemy (raptors). He was extremely buggy, and I spent hours fixing him and making him work properly. After all that, I only had time to include him in one level. Part of it was an aggressive decision to have an enemy that 1) seeked the player, and 2) jumped at an interval that I liked. Next time, I would spend time building out more levels with all the mechanics I could with the snail/ice cubes, and THEN build more enemies. Part of it was also just my inability to code and debug faster with this particular enemy since I was learning about some of his construction on the fly. Now that I have the knowledge of those behaviors under my belt, I can easily implement them next time. However, if I run into another similar problem of “I want to do this AWESOME thing in the game”, I will be more diligent and scope it better. For example, the raptors didn’t need to jump, and he didn’t need acceleration/deceleration. If I cut those, I probably would have saved myself 45+ minutes in coding, debugging, and testing. Time is precious in this situation.
  2. Allocated very little time for art, polish, and sound. Since it was just me, and my strengths are coding, testing, and game design, I focused on those first. I like that decision. However, I only spent about two hours total on art, sound, and polish. I would have loved to have a title screen, a little cutscene at the beginning telling the story, animations for the player and enemies, and music. I could easily have done all of those if I only had three to four more hours to work with. So next time, I will just budget more time for them on the last day.
  3. Difficulty curve was wayyy too steep for level four. New mechanics I added: raptors, enemies jumping from the edge of the screen at you, enemies in cages that can be released by buttons, and having multiple enemy types in a level. This was just too much. For example, next time I move level four to level eight, and have three levels between there where I introduce the raptors, and each of those mechanics individually per level and build up.


Gameplay gif:



Had a ton of fun making this game, and I only spent 21 hours actually developing it from start to finish. Gotten a lot of good feedback so far, I’m considering just turning this into a full-fledged game and building it out. I love the idea of turning lethal monsters into ice cubes and springs and such, and using them to navigate puzzles, Ludum Dare was a blast as usual. Hope you had fun, and enjoy playing the rest of the Ludum Dare games!

[cache: storing page]