In Memes Die Too, post memes on social network, watch them spread and acquire fans! Become so viral you'll be offered to join the infamous Memelords and Memeladies Society!
I developped Memes Die Too on my own for Game Off 2022. This big game jam hosted by GitHub lasted one month, and the theme was cliche. Game Off is dedicated to open source software.
Out of 567 submissions, my game was the 4th most rated game. It was ranked #15 / 567 in Innovation, #28 / 567 in Theme interpretation, and #64 / 567 overall. Within the jams voting span (1 month), Memes Die Too was played 364 times.
At some point memes die
Embody a memer on social media. One would simply watch their meme spread until it is considered a dead meme, a cliche meme.
This concept was very original, and I felt able to develop it alone as the base mechanic is very simple. There was some risk though. In the pitch, the player had no other agency that pushing a button and see what happens.
I decided for the second pitch. I wrote a document to better define it:
Solo dev : interests & planning
I choose to join this jam in solo for 2 reasons:
I never worked on a game I pitched myself before
- Owning full creative responsibility
- Go experimental (hard to do so with a team of complete strangers in a jam context)
I already made a few prototypes on my own, but this was definitely next-level
- Get to know better Unity
- Learn basic visual art tools
On a practical side, I had to address a strange planning problem:
So I decided for the following:
for addressing unknown dev time
Minimum viable product. Dev a feature in its simplest form possible. Then test to see if it adresses the creative needs. Only then decide on adding complexity or another feature.
Dont plan features ahead as the game or views on it will have changed at this point
for addressing gameplay uncertainty
- Prioritize gameplay design over everything else until gameplay has reached an MVP
A bottom-up approach to gameplay
The plan
The virality simulation being a very experimental feature, I had no clue on how it would function. So I started by building the simulation, and playing with its parameters to understand it.
The simulation consists in random users sharing the meme with a certain probability. For each successful share, 3 child users will be spawned in the next generation. They will share the meme with the same probability. And the cycle continues.
Users also have a probability to share the meme on FaithBlog. After a certain amount of FaithBlog shares, the meme is considered cliche and the simulation is stopped.
The simulation can also organically stop if no user is sharing the meme in a generation.
In parrallel, I made a list of potential features. These features were all tied to the fantasy of posting memes on social networks.
Features were varied. For instance, I had plans to give users a full profile banner (Description, Banner, PP) or a notification feature where uncle from FB could message the player.
How it turned out
At first, the simulation was extremely volatile. I had to do quite some maths to understand why :
With this maths study, I was able to add stabilizing features in the simulation :
The first generations feel much more random than the following ones. The logic is simple : as theres only a few users, the fact they share or not the meme has proportionnaly massive effects. (When there are a lot of users in a generation, its more of a statistic distribution.)
- I added Bots who always share the meme in the first generation
- I added a Universality statistic: the first x generations can't have 0 sharers (x = 4 to 10)
hidden mechanic
The chances for a user to share the meme is called virality. My mathematic analysis points out that, if the simulation reaches thresholds of virality and number of users, it then grows exponentially. This doesnt create gameplay issues per say, but it creates very bad looking share-trees, completely flat. It also removes some excitement in watching the tree grow.
- I made virality vary according to the number of users in a generation. This limits the horizontal size of the share-tree, so they look better. It also makes the tree grow for longer.
- The variables for this behavior are invisibly affected by a fan task (see Fans below)
Designing these stabilizing systems, I followed 3 important guidelines:
- simplest parameters possible so they can easily be linked to other systems
- most independant possible parameters so interactions with those are predictable
- most robust systems possible no value of the parameter breaks the simulation
With these systems set up, I could choose features from my list and link everything together in a nice gameplay loop. Here are the systems I developped :
Creative reason
I needed a feature to give players agency on what meme to post. A meme editor was out of scope, and linking it to the simu would bring a lot of design issues. My solution was then these cards. They join a meme with simu stats, and choosing among memes gives agency.
Gameplay use
Gameplay-wise, the meme selection screen is a choice. The choice is made interesting by having to build a strategy through statistics comparisons, linked to the fans feature. Memes are also interesting because they can have bad statistics, so the choice is more complex.
Side note : cards UX
At first, memes only show their virality stat. Its the most useful stat for early gameplay. Players can choose to discover other stats later. I made this choice for players to understand the hierarchy of stats, and relieve cognitive load while learning the basics.
Creative reason
I wanted to emulate the feeling of community promised by networks. Players gain new fans each time they post a meme, then can affect them to various tasks linked to simu variables. Supporting this fantasy, players manage fans individually and can see their faces.
Gameplay use
Affecting fans is also a choice. Again, its a tool to build a strategy. However, contrary to memes, the simulation is only positively affected by fans. As fans cannot leave, acquiring and managing fans creates the progression of Memes Die Too. Thus I have a system for global strategy and progression (fans), and a system for small scale tactics (memes).
The macro gameplay loop is very simple. It ensures a clear link between players actions and simulation results.
I got extremely positive comments from players! (https://itch.io/jam/game-off-2022/rate/1815131)
