This year some friends and me managed to finish our Devtober game, Soultower, a game about a lonely soul that tries to ascend a sinister tower to get a chance to live one more. In essence it is a short roguelike where the players can only attack by possessing enemies and using their abilities and mobility options.

Soultower Development

Soultower  was started for last years Devtober and we didn't make it in time, I personally felt really burned due to trying to make a game while already working full time on making games, in conclusion,  more than 11 hours of gamedev a day...too much game development for a month. 

This is how Soultower looked the first day:

And this is how it looked at the end of the first devtober, in this stage Ftalocianina and Alex started adding his work in art and music respectively, something that greatly improved the game but also meant failing the Devtober was even harder as more people work was being "lost":

(I was so tired and burned that I couldn't even integrate some of the great enemy designs made by Ftalocianina, and I kept my ugly enemies sprites...)

After finishing the devtober without releasing the game I decided to take a break from it. Over that year "taking a break" from the game I found myself going back to it to try different things: thinking about different enemies, thinking about progression systems, thinking of different genres for the game, etc

One of those ideas was to make the game a kind of metroidvania with puzzles... It required too much time for such a small game. These are some of the designs from that idea...

All that thinking made me realise we should at least try to finish the game in the next Devtober, specially to make everyone's work appear into an actual game. For that reason, we started this year's Devtober looking forward to make a good and complete game out of what had from the previous year. This postmorten is about this month's journey .

What went right

• Possession is a really fun mechanic. Getting from one enemy to other possessing them and using their abilities is really fun, it makes the game really agile and the different enemy abilities make you change how you play every few seconds.
• Game allows players to play in different styles. One of the things I usually do not like about the games I design is that they are created to be played in a specific way, making players learn how the game "must" be played. This time I decided to make a flexible game where players could play fast and risky or slowly and steadily.
• Learned a lot about Gamemaker 2. Although I have made some prototypes in Gamemaker Studio 1.4 I wanted to force myself to learn Gamemaker 2 to see what it can offer to make games. One of the things I love about Gamemaker is how fast you can get something working in comparison to Unity.
• Took my time to experiment. Usually when I'm making games I put too much stress on me to keep going without experimenting too much (first because I feel like I'm making everyone else in the team lose time and because I feel I'm not being productive). This time I took my time (at least at the beginning) to be open to relax and experiment with some mechanics without feeling anxiety.

What went wrong

• Not having procedural generation based on "chunks" made level design took a lot of time. Due to Gamemaker 2 current autotiling not working in execution time I decided to made levels handmade making me lose a lot of time specially in the last days where I had to check several times to ensure they were fun.
• Lack of decision in the gameplay.  Maybe by trying to make the game playable in different ways I made it too easy to play, game feels too fast with almost no limitations that slow your gameflow in a good way. (Tried somethings in the last days but all of them felt too artificial and forced). This also limited me in adding mechanics that add more depth to the gameplay or that limit the flying abilities of the player.
• Dull Narrative. Although initially we had several characters in the game and the narrative was about discovering what brought them to the limbo at some point we realised telling that kind of narrative in a roguelike was too difficult, players could read the first dialogs many times but the last ones would be found rarely and later in the game, so maybe then they would not remember the beginning of the story for that character..) For that reason in the last days of this month we swapped to a more simple narrative with jokes about your past lives. This dialogs are good and fun but didn't allow us to learn more things about narrative.
• Gamemaker 2 complicated teamwork. Most of my experience is based on Unity and Unreal. With Unity I'm really used to let artists in the team have their own installation of Unity so they can integrate art or music in the game allowing them to iterate and see how the game looks without requiring me to see it. With Gamemaker I couldn't find an easy way to do this, which made me become a bottleneck for their work and making me having less time for code and design.

Conclusions 

Overall we feel really happy with the game we came up and it has made us learn a lot about game making once more. Here is a list of the things I will try to keep or start making next time:

• Work on more procedural stuff from the beginning (If the game allows/needs it)
• Keep trying riskier things at the beginning
• Focus on engines I know better so teamwork is easier
• Have more confidence and decision in design decisions, defining a more clear and controlled gameplay while allowing different playstyles.
• Take it easy, do not feel overwelmed by work and possibilities so easily
• Focus on scope

Thanks for reading my Postmortem! I would like to thank Ftalocianina, Alex and Sergio the amazing people that made Soultower possible. But I would also like to thank Ellian for organising the event and creating the wholesome Devtober community!

This is how Soultower looks like thanks to all this people ^^ Make sure to check it out!