Survive Another Night Postmortem
I’m James, an indie game dev (@cbGameDev). I work on larger scale projects with my friends, in a company that we built from the ground up and I make more compact games on my own in my spare time. I took part in the December 2020 Epic MegaJam and I wanted to write up a postmortem as I thought it could be interesting to others and useful for myself to reflect on the experience.
SURVIVE ANOTHER NIGHT
The game I made for the Epic MegaJam 2020 is free to play on my itch.io here!
I’d wanted to do a game jam this year but hadn't made time for one yet, so when I saw the Epic jam was coming up I thought I’d get involved. It was a 7-day jam and the theme ended up being: “It’s been a long time, but we’re not done yet.”
THE GAMES CONCEPT
I’m a big fan of zombie stuff so I decided to make a game about a zombie apocalypse. You play a character trying to protect their family for one last night before help arrives. The initial design was inspired by a mix of Original COD zombies and Overcooked style gameplay and the film The Road. I wanted it to be a top-down game with two main stages: prepare and fight. I also wanted to make it multiplayer (more on that later) and thought adding procedural elements could add replayability. Before I started any actual deving I wrote down the core of my idea into a single sentence which I wanted to guide the development direction over the jam.
I’LL MAKE IT MULTIPLAYER…
As this was a slightly longer jam I thought I should have a go at making the game multiplayer. Not any multiplayer mind you, cross PC non LAN multiplayer... that was a mistake.
I burnt a bunch of time following networking tutorials and trying to get stuff working only to realise that it wasn't going to work/be practical. It made me laugh when I realised that I’d spent the first day and a half making a game that wasn't actually a game. I hadn't made any gameplay mechanics. Instead, over that time I’d managed to get multiple players joining into a lobby then joining into a match but the further I was getting the more problems it was looking like I was going to have with ports and tunnelling. In the end, I decided a game jam probably wasn’t the right time to try and work all this out so I decided to cut scope and keep it a single-player game instead.
In hindsight, if I had spent that beginning period making the actual game, I probably would have found more free time to add polish/juice and even to try a simple local multiplayer (Spawning a second local player who could control using a gamepad) instead of the more elaborate one I tried for.
PLANNING & TRACKING
For planning, I used good old pen and paper, Google Sheets and Trello. They each served a different purpose. I didn’t use any version control, I just made sure to back things up every day.
Pen and Paper: Mainly used for quickly capturing ideas, sketching designs etc. I really like to have an analogue way to plan.
Google Sheets: Always super useful, I used it for a few different things. Firstly making an overarching plan for how I wanted the week to go on a day to day basis. Secondly for any more permanent notes (aka non-development tasks) that I wanted to keep e.g. jam deadline, premade assets I’d used etc. Lastly to roughly track my progress, by writing quick, short-hand notes about how each day went. This was predominantly for writing this postmortem and looking back at how the jam went as a whole.
Trello: For task organisation I normally use HackNplan but I thought it would be interesting to try Trello for a change. As I was making the game by myself I wrote very loose and shorthand task titles. My board started off nice and clean and got progressively more messy toward the end as I kept adding more lists. I’m 50/50 on Trello. It did the job (although the website went down for me on Monday) but I’m not a super fan of its layout. I think it’s probably okay for short projects but in the future I think I’ll stick with HackNplan or try something new again.
HOW THE GAME PROGRESSED
Saturday was my only full free day as we were busy in the office. I split making my game into two chunks per day: before work and after work. Interestingly this gave me time to passively think about my games progress while at work. I tried to cook builds off often, cooking one at the end of each session of work (e.g. in the morning/in the evening). To make enough time to make my game I ended up getting up early, going to bed late and drinking a bit too much coffee.
KICK-OFF FRIDAY:
I didn’t do anything on kick-off Friday. I knew the jam had started but I needed to finish off some work and thought I would take a first look at the theme Saturday morning when I was feeling fresh.
SATURDAY:
I sat down Saturday morning with some coffee, looked up the theme and started scribbling ideas down until I got to my zombie idea. I started a project from the Top Down Template, made a Trello board and started following a tutorial series for setting up a multiplayer framework. By the end of the day, I had a main menu leading to a lobby which then led into a map. I’d made good progress and learnt a lot but it had taken me the whole day and I technically spent the day not actually making the game.
SUNDAY:
Morning: I finished off the multiplayer tutorials I was following then had a think and realised that it was all getting a bit out of hand. Here is when I decided to cut scope and just focus on making a single-player game instead.
Evening: I went to work for a bit and when I came home I started work on the single-player version. I fleshed out the basic game loop (pre-mode, build mode, fight mode and win/lose states) and got creating a bunch of the base mechanics: family spawning, AI spawning, Ai movement, AI attacking, HUD menus, escape actor, simple barricade you build/interact with and player movement. I felt like I made so much progress in a short period of time.
MONDAY:
My original Monday plan had been to get all the mechanics fleshed out. I didn't manage to achieve it but I did make a bunch of progress. However, even though I had my game loop in it wasn't feeling like a game yet, more like a bunch of random game mechanics.
Morning: Before work, I got lootables and pickups working, made a resource system and started adding more UI to the HUD.
Evening: After work, I spent more time working on the barricades adding a new window type. I also worked more on the UI, made it so the player could attack the AI, made the AI attack doors and family members and also got family members taking damage. You can see the Ai shoot off when they die, which still makes me laugh when I see it now!
TUESDAY:
Originally I had planned to get most of the art done today; environment art, making the characters, rigging/animating them as well as adding simple post effects. However, things didn't go to plan.
Morning: I struggled in the morning and felt like I wasn't making much progress. I think I was tired from the previous days and work. I tried to get a pipeline for voxel art but it was taking too long (see the art section) so I switched to level design instead. Then I started having a bunch of problems getting the level to feel good (see the game design section). In the end, I made a small amount of art progress and had a messy level design to come back to later.
Evening: After work I felt more energized and decided to leave the art for now, instead focusing on the level design and mechanics. I worked on getting the family to follow you and stop when you tell them to. Turning a generator on “turned the power on”, activating lights in the level. I added billboard dots on interactable objects to make it easier to see what you could interact with. I also worked on zombie waves so they now spawned in rounds, increasing in numbers as well as spawning from a larger range of spawners as time passed. The last thing I did for the day was adding a new character mesh and a few animations for the player. By the end of the day things started to feel like they were coming together a bit, a big change from how I had felt in the morning.
WEDNESDAY:
At this point, I had greatly deviated from my original overarching plan. Because I was behind instead of working on Ai, level design bugs and juice, I decided to try and finish off the remaining gameplay mechanics.
Morning: I got melee weapons and weapon pickups working, made a new type of ai that was already in the world, who stands still until you get close then attacks. I also made the crafting table require power before you could use it. I felt like I didn't get as much done as I wanted to before I headed off to work.
Evening: After work I got character meshes and animations in for the Ai and the family members, which I felt like it helped a lot in making it feel like a game. I also worked on UI and made progress on the crafting system, making it so you could craft armour for yourself and your family.
