The Making of Dewdrop

I thought I'd start writing things down so I'd get better at it in the long run. Game development is something I want to take seriously, and being able to put my thoughts out is a skill I should practice. Rough draft so far.

The Game Plan

The jam is 9 days long, so we scheduled the plan to roughly

Day 1: Brainstorm

Day 2-3: Core Mechanics

Day 4-6: Content

Day 7-9: Polish

Brainstorm

The theme is Rain or Shrine. We try to ensure that we relate the theme to the core gameplay mechanic, rather than the setting. After the brainstorming session with a few dozen ideas, we decided on just focusing on the Rain part. Shrine as a topic felt too thematic for us. For rain, we had two ideas.

1. A tile-based puzzle game, where you control a sun over nature. Evaporate the ocean, control the wind currents, rain on burning forests. The main drawback of this design was that it didn't sound as fun, and we didn't have a goal for the player. Could still be a cool concept if developed more though.

2. You move and manage a few rain clouds and put fire out as they appear. This seemed more chaotic and fun, but we didn't know where to go from there. The lack of depth in mechanic made it shallow.

We wanted something that was both fun and mechanically in-depth, so we went from idea 1 and made it based around a raindrop. Naturally it turned into a platformer and we started organizing a bunch of ideas and interactions. The water would change states between ice, steam and water, and have different movements.

We created a table of game objects and interactions, and stopped when we thought we had enough for a full game.

Development

With the plan now set, everyone could start working on the game. On just day 2, we got most of the core mechanics done. Now it all comes down to level design and polish. 

Player Movement

The game would've been drastically worse without playtesting. From the recent GMTK video, we decided to start playtesting from the moment we had a single level. It went horribly. Player control felt tedious and unintuitive, there were many exploits and bugs, and most importantly it wasn't fun. We had to find the root issues.

After a few redesigns of player controls for the rest of the day, we finally got a workable setup.

Water - Click to aim, release to jump, power is distance from player.

Steam/Ice - Click and hold to move left and right.

Level Design

After having a simple level, I decided to make an end goal for the game. I figured we'd build up levels to lead to a complex level. This ended up never making into the final game, but we did learn from playtesting it.

This level was designed after exploring different mechanics the player could do. This included mechanics that emerged from a simple mechanic and combinations of mechanics. The result of playtest from the team was that it felt too confusing, and it didnt feel good discovering them. It was more of a "Boo, I didn't know you can do that" rather than "Ah ha, I got it".

I believe that this was an issue due to

- not having any buildup to the level,

- a poor cohesive idea for a singular level

This was not good sign. We decided to keep it simple and focus on the platforming aspects of the game. Each level should introduce one new thing, and build on top of it over several challenges.

At this stage, we also started cutting out interactions. Ice was also completely removed at this point. We decided on just making 5 levels, and polish them as much as we could.

From here we made levels 3, 4 and 5.

Level 1 focuses on water movement.

Level 2 introduces moss.

Level 3 introduces steam.

Level 4 introduces fan.

Level 5 combined everything from levels 1 through 4. I also wanted to introduce emerging behaviours that weren't introduced before.

From playtesting, we tweaked the levels. Namely

- Protect players from themselves.

- Make the intended paths clearer.

- Make impossible paths obviously impossible from first glance.

- Balance the difficulty.

The biggest hurdle was level 5. Some players found it perfectly balanced, while others found it too hard. It was a long stretch of challenges that required consistency and planning.

Checkpoints? This was a controversial topic among players. This was meant to be the final level.

In the end, we could see the satisfaction of overcoming the final level, despite the frustration. Most playtesters said to keep it as is, so we stuck with it. That said, there were still a lot of tweaks made.

- A signboard in the beginning to point to the exit. If players new they were going somewhere, it might help them persevere.

- Lower difficulty. Less precise jumps. Less points of death.

- Smaller player hitbox. This ended up requiring us to retweak every level too.

There were a lot of discussions and debates, and not everything went smoothly, but the game turned out better than we could've hoped for. Watching people try the game always made me anxious if they'll like it or not.

By now its the end of day 7, with two days to complete a the polish we can afford to. It's always a rush at the end :')

Final Thoughts

I want to thank Raccudon for the impressive droplet animations and world building from our vaguest descriptions of what we wanted.

KiraDragon for the beautiful ending piece to tie the narrative.

Damyk for being brutally honest and realistic on gameplay, time schedule, and communication.

ZeeHatcher for being a god as usual.

Playtesters dragged in by DamyK. 

Not sure where this project would go from here. There are certainly many mechanics left untouched, and the control scheme would fit a mobile game. Regardless, it was a fun ride.