Introduction and special thanks
This was my first LD, and I came here fully expecting to faceplant from over-ambition or poor time management. But I submitted my puzzle game, ‘Additive’, polished and complete with hours to spare. It’s the first complete game I’ve ever done, and it’s gotten more positive feedback than I had dared to dream, and the process of making it shoulder-to-shoulder with all you other fine people (especially on IRC) was awesome.
Before I start the post-mortem, I want to thank ‘Cake’ from IRC for giving me a quick pep talk around six hours into the compo, when I was feeling particularly unprepared. I also want to thank all of the IRCers who complained that my first release candidate was too cryptic, and especially ‘Tau’, who played through the entire thing and deliberately found ways to break it. My game would be half of what it is without such playtesting. On with the post-mortem!
Part 1: The pre-feedback post-mortem
This is actually the second post-mortem I’ve written for Additive. The first one, written immediately after submission and before voting opened, can be found here. Its salient points are:
What went great
- I knew my strengths and weaknesses at the start of the competition, and this guided me towards reasonable ideas and goals.
- Making a puzzle game was a great decision because the very nature of a puzzle game works against feature creep.
- I used a notebook to organise myself and work through problems.
- I set some time aside to brainstorm ideas.
- I developed a quick way to make levels, which was more efficient and accessible than building the game board in Unity’s editor.
- I avoided crashes by limiting sugar and caffeine intake.
What went poorly
- It was very difficult to design puzzles.
- The game is not colorblind-friendly because I had a hard time letting go of the aesthetic I had developed. This is not the kind of developer I want to be!
- I didn’t have enough time to deepen the game’s mechanics.
Part 2: The post-feedback post-mortem
More things that went great
- I decided on a feature lock after the first day. By the end of the first day I had a working game. The second day was dedicated solely to polish and level creation: no gameplay changes allowed.
- I did early testing. Sticking the game up on my Facebook delivered exactly zero constructive criticism. Posting the game on IRC for fellow devs to play got me immediate blow-by-blow feedback, and my time budgeting on Day 2 allowed me to work on every single one of the issues that were raised.
- I put a lot of effort into visual player feedback. I changed the buttons on the main menu at the last minute to make them look more buttony and clickable. There’s a nice marker to indicate the selected block. The marker and the selected block pulse with color. Blocks animate towards their new positions instead of just teleporting over.
- I spent even more time on aural feedback. When you click on a block, it makes a sound. Deselecting a block makes a different sound. You get a different sound again when you try to make an invalid move. When blocks combine, the sound they make depends on the outcome. Sound makes a game feel alive and reactive. Skimp on graphics before you skimp on sound.
- I think I picked a fairly cohesive style. The game is an exercise in minimalism (like I said in Part 1, I knew I was bad at art). Its presentation was informed by the effect of parenting a spotlight to the camera and tilting the camera 45° towards a plane. I felt that an understated look deserved an understated and elegant sound, so I used single piano notes in GarageBand iOS for sound effects. The game would have been a dissonant mess if I had used SFXr.
- Having the tutorial levels was a good idea. It gave me the opportunity to dress the game up with prose, and it also communicated the essence of the game efficiently and enjoyably. It became even better when I added explicit instructions on suggestion from the IRC testers.
The only other thing I can think of that went poorly
- The black squares imbalanced the game. It was intentional that you’d be able to walk the black squares around in the last level, gobbling everything up with wild abandon. I didn’t know that this was possible in the other levels, and the game’s difficulty suffered as a result. Some of the levels have black squares adjacent to each other because I was actively trying to avoid this exploit. If you poke through my timelapse and my source code, you’ll see that I actually did have other ideas for color combinations and win conditions (some of them were actually suggested by players in the game’s comments), but I knew that the game was already pretty good with black squares as they are, and I had no time to change the mechanic and playtest it to my satisfaction.
In all, I’m incredibly proud of what I made. It turned out far better than I expected, and the experience is invaluable. I am already feeling the itch for some rapid prototyping, so you know I’m going to be back in December. Thanks for reading, and don’t forget to play Additive if you haven’t already!
A surprise for you, dearest reader!
I’ve uploaded pictures of the dev notebook entries I made during LD24 to Additive‘s page on my website. It starts the day before LD, and ends with the list of things I wanted to address in this post-mortem.