Postmortem of Floodtown

Hello everyone!
I decided to join Ludum Dare 42, my first one, 4 hours before it began. I was on impulse, but I don’t think I regret that decision. My game received rather positive comments and I think many players enjoyed it.
But now, after 2 days of rest, I think it’s time to analyze what went good and what went bad. So here we go:
Good things:
- Gameplay loop was enjoyable. To maximize earnings, player must constantly readjust positions of their buildings, which requires constant attention.
- Overclock system allows player to breathe for a short moment, only to make game faster and more alive. Combined with changing global bonus it makes game more alive, and not only „place and forget”.
- Many players found graphic design to be rather good, but most of all clear and consistent.
- Game is relatively bug-free (only two bugs that I know of, with one already fixed for next release). Floodtown fits „Running out of space” theme very well
Okayish things:
- Floodtown does not use any popular game engine, but rather is written in very simple Java, using mainly Graphics2D library. It allowed me to do whatever I want, as long as it’s simple. But that also means that game will not run as fast as C++ code and I had to do a lot of optimalizations on my own.
Bad things:
- Performance of Floodtown was suboptimal on low-end devices (but the game didn’t really require high framerates).
- Some people who didn’t get how the game was supposed to be played, could be lost relatively easily.
- I’m not good at creating sound effects, so some of them (waves sound) were bad and distracting.
What’s next?
I’ve already started rewritting most of the internal code, as I want to fix it first. Then, I plan on adding at least settings menu and tutorial to make game easier to understand.
Also, * thanks to all of you, for playing Floodtown. * I think it’s my best game to date.
If you haven’t played it yet, you can do it here: https://ldjam.com/events/ludum-dare/42/floodtown