With about a day left till voting ends, seems like as good of a time as any to do my LDJam 52 post-mortem.

My game for LDJam 52 was called "Soulful Harvest." Contrary to the name you don't actually harvest souls, but you prevent ghosts from stealing the crops that you DO harvest.

What I believe was the most important experience I got from making Soulful Harvest was having to make significant changes to the implementation of my game from how I conceived it. I read that this is true of storytelling as well, perhaps it is extendable to any creative work: you can't be afraid to get rid of what isn't good.

My initial concept required a LOT more button presses to play. In order to plant on farmland, you needed a seed. Getting a seed required a button press on the seed shack, and planting the seed required a button press on the empty farmland. Lost farmland (grass) couldn't be replanted at all! A little bit of playtesting quickly revealed that this was tedious and frustrating.

What broke through the frustration was distilling the player's actions down to the essentials. Fundamentally the gameplay was about positioning yourself to defend as large an area as possible. Needing to consciously plant every lost seed was just a burden and distracted from what I wanted the player to focus on. I fixed this by automating all crop planting. Planting on empty farmland happened just by walking over it and no longer required a seed. Lost farmland could be reclaimed by walking over it with a seed, which was automatically grabbed by walking near the seed shack. By doing this I believe I distilled the player's interaction with the game down to just what makes it fun.

On the last day, I realized that there was a lot going on in the game, which the automation I did could obscure. I knew that I would hate having to read a few pages of explanation if I were playing my game, so I felt a tutorial was in order. I based the design of the tutorial off of some tips I learned from the developer's commentary of Portal.

1. The player should be forced to use new concepts to advance.
2. New concepts should be introduced one at a time.

Doing this ensures that new players have to learn new concepts rather than skipping them, and don't feel overwhelmed by their inclusion. I believe including this tutorial in my game was a huge benefit (and a few people in the comments pointed out that they liked it), and I plan to do similar things for all of my future games.

If there's anything I wish I would have done more of, it's user testing. Something I've become very aware of is that as the creator of my games, I have a unique inside perspective to how they work. Things that feel obvious to me may be entirely cryptic to someone else. My boyfriend was able to playtest the game for a bit while I was making it, which was extremely helpful in getting a little outside perspective on how my game felt to play. However, he's just one person and can't catch everything everyone will see in it. Two common complaints that I got were that the play area was too big, and the dash felt useless. Both are criticisms I agree with, but I wasn't sure how big of a deal they would have been until I released it and got feedback from strangers.

All in all I'm very happy with how this game jam went. I really enjoy the community that game jams provide, and it's very satisfying to have made something at the end of it. I aim to return for LDJam 53! Below are some pictures from the development of my game.

First page of brainstorming for my game. None of the enemy designs ended up getting into the final product.

The design of the player and deciding how they would control. I also spent some time figuring out some specific upgrades the player would have. Originally the double jump was going to be merged with the dash.

The first sprite of the player, which I used for about a day. I never removed it from the game files because I forgot.

An early build of the game which was used for playtesting. The background was nonexistent and the tileset was all single color. While it can't be seen in the screenshot, the player had far fewer animations. The double jump was still merged with the dash.