Kitty Maestro is a single screen game and this should mean level design is pretty straight forward, right? Cats are easy going and all that...
Kitty Maestro is the third game I've released on to the Switch. My aim from the start was to create a weird music arcade game that featured short riffs on a treble clef that the player - a cat composer - has to collect and then "perform" in the Meowditorium.
Development started in February 2019 and took a single screen, fixed grid approach to combine what I love about Nintendo's Game & Watch series with a few nods to other 8-bit games like Mikie and 16-bit games like WizKid.
Technically this was the most fuss-free game I've made including the procedural music generation and my first implementation of AI pathfinding (based on A*). The combination of the art generation and level design was, let's say, a learning experience for this game.
Timeline milestones (simplified)
February/March 2019 - music generation code and refinement
April/May - game grid, AI pathfinding, city map, player/enemy movements
June/July - art style, demo level generation and testing
August - sounds, meows, testing
September - testing and LotCheck submission (demo, passed quickly)
September to January 2020 - levels, art for levels, testing
January 2020 - LotCheck submission (full game, passed quickly)
February 2020 - Kitty Maestro released on Nintendo Switch.
Illustrator (a lot)
Parallels (because I'm on Mac)
What went well
Coding and process
Technically this was the smoothest process I've experienced when making a game. Of course there were bugs and the odd road block but these were quickly over come and didn't interrupt the flow.
The overall process was clear and clean too, including the LotCheck process. I feel with Kitty Maestro I was starting to find my own method and magic instead of being overly influenced by anything I was told or had read. Although...
Cats and Playing
I had wanted to make a video game that was based on Nintendo Game & Watch games having read the English translation of The Unofficial Game & Watch Collector's Guide. It would combine cats as well as much needed experience in level design. This influenced the art style a lot and working with literally about 8 fixed shades of grey was fruitful and liberating.
I had also read Homo Ludens and consciously incorporated one of the lessons from that book, namely, playing is an abstraction from reality in order to safely test something. The thing I wanted to test, although never formally stated at any time, was to do with drawing inspiration from the sounds of a big city to compose music. The twist being the city is a city of cats and the composition is broken into small riffs that meow!
I loved coming up with a cat city (Petropolis) and all the cat-pun based locations for it. Conceptually I loved working on this game and feel that are still not enough cat video games in the world!
The theme of music comes across and I pleased with getting music notes that are accurate (pitch, tempo, rhythm, key) and in particular the game grid mechanics operating on a fixed tempo where everything that moves is literally on a timed grid. I aimed to create a music game that had no music, only the polyrhythmic sounds of the player on the game grid, combined with the meowing riff completing as the player collected notes. I also wanted a game without a scoring system because it just didn't need one.
I finally practiced what I've been preaching for years and made a free demo. I put together the demo very easily and quickly and I believe had more sales AFTER the initial release because of the free demo. We need more free demos! They are pro-consumer and for me they show a confidence in your own game.
What went ok
A lot of the visual elements of Kitty Maestro work well. But equally there are some that are unclear. The most obvious of these is the enemies - the various objects and things that the player must avoid. As static art I think they work but as "animated" art that conveys an intended movement direction some are just confusing.
For the enemies that stay pretty confined within a space that don't overlap with other enemies that isn't much of a problem. Enemies that move around a lot, especially when there are more than 3 on screen, and the player cannot work out what's moving and where. Also there is no distinction between enemies that fire projectiles and those that don't... you only know when it happens which can be a jarring experience.
I cannot work out the answer to the conundrum of providing clear feedback to the player and keeping a consistent art style that clearly shows which grids squares are good/bad and avoiding making levels that all seem the same and avoiding having blocky art. Kitty Maestro has some good levels not helped with unclear art. Perhaps not using colour anywhere was a mistake.
What went badly
There isn't one. It's a How To Play screen (Manual button) that no one reads. I understand the just dive in approach to playing because I'm guilty of it too (although I will read quick explanations but definitely avoid games that make me trudge though a remedial class style tutorial). For Kitty Maestro it seems that it needed a tutorial, or that I made a game not easy to grasp. Or both.
Level/Art Design Process
Everything I said about the overall process being smooth didn't apply to the level design and corresponding art design that the level required. Here was my process:
choose next location you would expect to find in a large city
make a list of objects, buildings, animals, sounds with that location
draft a grid showing player, music notes, enemies, obstacles
construct grid in Unity
do art and sounds
test and iterate
repeat from the top for each location in Petropolis
Sound easy right?
This was hard, frustrating, awkward, with lots of mental blocks. Just as soon as I thought I had a flow with a one or two levels it come to a screeching halt. Lessons from one level couldn't really be carried forward to another unless I wanted to repeat myself. In the worst ways it felt like I was literally designing 55 Game & Watch games and I was naive... I found out too late that this type of level design is one of the hardest. Thanks Oliver :-) Knowing I have to "find the level" is very different to actually finding it. Damn damn damn damn damn.
Perhaps a custom tool would have helped. But I couldn't see how since the process wasn't inefficient or breaking me out of a creative flow. The problem was the tough nature of designing 55 such levels and doing it day after day after day...