This is the post-mortem of my fourth video game Loot Monkey, a 2D, flip-screen platformer. I am an amateur and part-time game designer and have been designing games for almost three years. (I work full-time and so the game was completed in the evenings and weekends.)
I had thought about making Loot Monkey for some time and had a few notes before starting but the concept was clear before I started. Even so, beginnings are delicate times and during those first few days I never knew if it would all come together. (Spoiler: It did.)
Loot Monkey started as both a mission to make an 8-bit style platformer with modern graphics (i.e. rip off Jet Set Willy but creating my own visual style) as well as a distraction from the frustration of various shit going on around me (don't ask).
The scope of the project was deliberately small in order to prevent the toxic feature creep and asset bloat of my previous game (although that was a 2D top-down space game I was determined to avoid languishing in the sluggish months of coding hell again). Setting clear goals and restrictions from the start really helped and meant that I could plough through the process.
For example I knew I wanted to use a pixel-reading level loader for a block based game that would make making the levels (rooms) a lot faster and easier (because I am really bad at level design). I was also dead set on having photographic backgrounds for the levels as well as having levels (rooms) that were differently shaped (not rectangular). I also wanted to keep the sound and music simple with a 1980s feel, and have voice acting for the part of the Evil Overlord if cost permitted.
In terms of timescale I had no idea what to expect. Having read that Matthew Smith created Jet Set Willy in about 8 months (full-time) I hoped to speed that up with a much faster level creation process, perhaps getting it down to 4 months with sustained effort. I over-estimated as you can see...
Timeline - Milestones
(timings are approx)
Day 0 - Loot Monkey project initiated.
Week 1 - Level maker completed. Monkey animation and movement completed.
Photoshop/Illustrator (other image programs are available but they are crap)
Audacity (great audio tool and free)
Garageband (great for loop based music and free)
SXFR (sound effects galore and free)
What Went Well
Starting strong always feels like the best start and with Loot Monkey this meant having a cohesive, fun idea with a clear goal of what to aim for. It felt like starting with a densely-packed well-defined nugget of something from which I extrapolated a platform game - I could see what Loot Monkey needed to be and did it. I remain surprised how easily everything fell into place and I think starting with a strong idea was the key. I also think that exercising restraint was as important so the project didn't increase in scope or complexity and lag as a result.
This was by the far the most satisfying part of the project and just as well because it honestly took about 80% of the overall time to complete the 70+ levels. Having nailed the level creation and testing process very early on with the aim of making it fool-proof and quick without sacrificing on the fun element of actually creating Dougie's world meant that I had a shameless amount of joy while making the game. And all without ever need to mess with any of the code when creating levels (it was near enough flipping between Photoshop and Unity).
CXFR continues to provide a wealth of sounds, more so when used with Audacity. Garageband was surprisingly feature-rich and easy to use despite never touching it before Loot Monkey. Moment to moment gameplay really lifted when sound and music were added. Extra sparkle was purchased from a very talented voice actor who took my script and instructions and created the gravily Evil Overlord I wanted.
I can draw really well, honest. But I can't draw anthropomorphic vector graphics that well and I really can't animate for shit and knew this long before Loot Monkey was started. So I searched the internet for vector-based animated monkey and managed to find and buy the right artwork. (I wanted vector-based artwork so I could scale Dougie as needed). I then altered the supplied files which included animating the tail, re-shaping Dougie's face and re-colouring him to be less poo-brown and more huggy-fuzzy orange-brown. Some tweaks in Unity were needed to get Dougie's animation just right and suddenly he lived and was dodging chomping monsters!
Pace of project
I loved the pace of Loot Monkey as the process had been broken down into simple elements (creating levels) which meant I was able to pick it up whenever it suited me (most evenings) and make progress in small but distinct increments. Creating Loot Monkey part-time in the evenings and at weekends meant I had to be reasonably disciplined, but not too structured in managing my time and I relished that freedom. There was no crunch, no stress, no guilt, no bullshit artificial deadlines to hit, no hassle, no development hell, no frustration. Since I managed to get most of the coding out of the way early on, Loot Monkey was very much a creative project as opposed to a technical one. And even the coding parts never felt like slogs because technically it was very simple.
Following advice from others as well as prior experience I wanted a test mode in Loot Monkey and at the same time decided to add other game modes. These were initially an over-long list (I sometimes get carried away as much as the next games designer). I quickly trimmed them to the core 5 I felt the game needed and stopped before I ruined things.
What Went OK
With help from platform tutorials (from Kyle Pulver [LINK]) about how to code jump assist and platform edge assist I was extremely happy with how Dougie moved in his world. In conjunction with Unity's physics engine and some necessary designer's tricks I liked Dougie's movement and responsiveness. However I am still not happy with a couple of sticky issues that are possible and feel that by today's standards he can seem too simple and not actiony enough.
I love the enemies and how they contrast with the rest of the game's environment but I think I needed more variety and had some issues with edge detection that still feel like a plaster rather than a proper correction (most people won't notice them but I know they are there). Also there were no conveyor belts like there are in Jet Set Willy and nothing like keys or switches, nor projectile-firing enemies, although I think these could have diluted the concept. But then there is always the possibility of adding them in a sequel.
What Went Badly
I am old-fashioned in the sense that I build it first and then talk about it, because I feel I'm lying if I talk about something that isn't ready yet and could still change before release. (I design and code flexibility into my games as I need to have room to change my mind. It's also why I love using Unity.) Marketing for Loot Monkey was practically nothing until public release. On the upside the creative process never got pissed on and I got to really enjoy myself for almost two months. I also hired a talented video man to do my trailer, screenshots and GIFs. But it was too little, too late. I hate marketing and it went badly.
I definitely want to repeat this experience when making all my future video games because to enjoy making video games it's up to me to make the process enjoyable. For Loot Monkey I did and it was. *BIG GRIN*
Note to self
Be disciplined, enjoy yourself, and marketing, marketing, marketing.