Postmortem Report

Project: Chill Out

Genre: Platformer

Timeframe: 23th October to 29th October (with a patch on the 30th)

Game Design: Azmio

Artist: curricle

Programmers: Robin Betka, Googol Planck

Sound/Composer: Raphael Batot, nickmontopoli

The project was launched for 172th weekly jam with the weekly topic of catapult.

We got a team of 6 together fairly quick via messaging people who looked for a team in discord (mostly done by Azmio) and after the theme was announced on Friday we were brainstorming our ideas.

The biggest challenge we had in this process was that time zone wise we were all across the globe, from the US, Europe and to Asia, which made this process more time consuming than usual. We trimmed down our ideas to 2 but we wanted to let the entire team vote, to not leave people behind. We ended up implementing the platform idea because we thought it would be the best to implement and most fun to play.

We decided to implement this from scratch without any template at all and we started working on it on Sunday.

We choose Unity as the technology as it is the most productive and most easy to use engine and also perfect for the game we were building. Our sound artists decided to use a middle ware called Wwise to learn more about creating sounds with it.

Our ideas were ambitions and we knew that if we can pull it off it would be a massive game for a game jam. We decided to use Unity’s' tilemap engine to implement the game and we used dummy assets everywhere to get the game logic going. We started working on the coding of our main character, which was a very challenging task until the end. With jumping, dashing, gliding and wall jumping there were lots of tasks to do. 

Designing these we experimented a lot to get the right feel, we tried things ourself but also googled ideas. Especially when it came to bug fixing where we had issues that the wall jump detection was not working well for example, we had to get very creative in the solution, and learned that faking things often make the better solution. If you only allow a jump or wall jump if an actual collision occurs, the timing to hit them is very critical. So we ended up designing trigger colliders that were larger than the physical player collider, and checked manually with Physics2D functions if we are allowed to jump or wall jump. This greatly improved the timing of these elements which made the game much fairer to play. 

In general, we used trigger extending the actual physical collider a lot, to give the player a greater chance of doing things. There were small design changes during the course of developments such as the way the catapult could be placed. Overall, we were very surprised how fast the development process was on this game, considering we only had 4 full days of actual work on it. There were small technical things that did not make it and we had small problems creating art because we had to recruit a new artist in the middle of the project, but overall, the result is very impressive. 

Code that was written on the later half was a bit rushed and not necessarily clean, as we got the tunnel vision to bring it over the finish line and not focusing on optimal tasks. The sounds created with Wwise are very impressive and dynamic and it is a very good middle ware to use with Unity and all the art turned out extremely well as well. The team thinks the level design turned out very well as well, with some little fixes being done with patches (can’t have them too easy :)) When recruiting a team in the future it might be helpful to ensure some kind of time zone consistency at least, as working with 3 completely different time zones was sometimes a challenge and we did not realize we had this issue until we started.

Overall, we got a very good playable product which only needs some polishing in the inputs, as it sometimes does not feel 100% fair and consistent but considering this was a weekly game jam game it is fine. Unity was the right and only choice for this because I heavily doubt, we could have done it in this time with any other technology. We learned lots of sound creating, adjusting our dummy assets to the real assets and lots of things about collision management, because to make a fair platformer you need to help out the player without them even realizing.