I remember how excited you were when we announced a contest crafted by the community itself, at the end of Wondev Woman in July. I must admit, we were excited too! We knew a project of this size wouldn’t be without hurdles, but we couldn’t wait to see what game the creators would come up with. This is a remarkable and huge challenge.

The Fantastic Four

No need to really present Magus, reCurse, pb4 and Agade anymore. You regularly have the chance to read their postmortems in the forum.

Why did we chose them?

At the beginning, we wanted to welcome any community member interested in contributing, but we finally decided to pre pick the four creators for three reasons.

First, we knew them well enough to trust them with such a project. As a first of its kind, the community contest already included a bunch of unknowns.

That’s why we also chose people who were good –or should I say, excellent– at our contests. This way, we ensured the game would be interesting and deep enough.

Finally, we knew they would very much enjoy creating a game. Given the time they spent on creating tools for CodinGame or debating contests in the chat/forum, we were pretty sure they would go the extra mile and produce a great game.

From Zero to Mean Max

Below is our first roadmap; it turned out to be quite accurate.

July-August: We started smoothly; everyone had holidays, anyway. The first objective for us was to show Magus, reCurse, pb4 and Agade how we built our contests and what parts they would need to code. At the same time, they had to start brainstorming and come up with ideas for a game.

September: Time to choose a game and create a first prototype! After that, we could assess it and decide to go for it (and eventually announce a start date) or start another idea. That’s also when the Fantastic Four had to ponder the theme and title of the contest. The chosen theme here is obviously MadMax, although we came surprisingly close to choosing Baywatch.

October: We thought one full month would be enough to complete the development of the game. More on that below.

November: We initially planned a release date for mid-November, more to complete the roadmap than anything. We agreed we didn’t want to pressure the team in any way. They all have jobs and can’t dedicate all their free time to this project. In the end, we’ve managed to more or less keep to the initial schedule, so we’re pretty happy!

Building a Contest

You might wonder what happens on the inside.

I’m sure you can understand how difficult it is to work with others on the same code base. For this reason, we all decided that only one developer will actually code the referee (the main part). Magus took the lead in the developments, since he wrote the prototype that was chosen. For simplification purposes, we kept the integration of graphical assets and a bunch of annoying stuff to code on the CodinGame side. We’ll work soon on creating the needed tools so that anyone could easily create a fully functional game.

But creating a contest is not just about coding (in Java by the way).

First, there are a lot of discussions on our Slack. We never talked about the contest face to face; we just discussed on Slack. As of today, we are currently reaching the 12k messages mark. And we’re all fond of 20-lines block messages (which count as one).
Before choosing the game, we actually discussed 16 other ideas for games. Some in more detail than others. The final decision took several days and votes; it was very close.

Then, as you might guess, we’re testing the game. We have a private CodinGame arena, and we all create different AIs to test the rules and the viability of the game in all different leagues. Now, even if they intend to test the Legend league, the created AIs are much more simple and some might even become the bosses of the Wood leagues. In order to achieve the best experience, you have to discover all possible strategies and act on the rules to prevent annoying strategies from working against you.

Finally, there is a bit of writing. As you know, the statement is an important part of the game. If well written, it helps a lot to enter the game and start having fun. Needless to say, the referee will be available on GitHub, as usual.

Week Minus 2

The contest starts on Friday the 17th and will last 10 days. The whole contest will be very similar to a classic CodinGame contest. There will be prizes (not decided yet) for the three top players, and also CodinGame T-Shirts for the top 20 players. And, of course, a new contest also comes with new CodingHubs! As usual, we are maintaining all the necessary information in a Trello board. Participating in a contest is fun; joining a CodingHub just makes it more fun. Try it!

We’ll most probably hold a streaming session on Monday the 20th on Twitch to help everyone get started; stay tuned.

Edit 06/01/20: the Mean Max game has since been released on the platform, check it out.

The Dawn of an Era

We’re looking to create more community contests, and to hold them more often.

After Mean Max, we will discuss what went well and what could be improved with the team of creators. This will help us to create a better experience for future creators. For example, we will probably reduce the teams to two contributors only.

If you’re interested in contributing to such a challenging project, just send me an email at [email protected]!

Unfortunately, we can’t tell you more right now about the game; we can still show you the sketches for the artwork below. WIP!


We can’t wait to see you on the 17th — what a lovely day!