I was late, read through my emails and saw that GMTKJam reminder. I had things planned during the weekend, didn't assemble a team and had limited time to work on a game. All these things played in my favor in avoiding the worst thing to do in a game jam: overscope. If I was to make a game, it should be done as quickly as possible. I thus opened my trustful pain.net with, in mind, a creepy face line art, white lines on black background, not unlike the one in Teeth Simulator. There was no design, just that idea of a face. As I came to draw the mouth I thought: "Let's make it large and ominous". And then it was too large and I thought: "This is way too large... so large it could be a digestive tube." There was my first design: let's eat food using physics, without control of the jaw. I drew the jaw separate from the head, then food (red meat and green lettuce). And then, I thought "teeth", drew some hastily. One was way too big I thought, and the second idea came in: "Let's level up our teeth". I had no idea what would make teeth level up, I just knew they would! Then I drew pupils of different sizes and kitchenware and that was it. "I am not drawing anything else" I thought. And I didn't draw anything else. I made some minor changes though: made the food white to have it multicolored in game and made the teeth yellow.

As I worked on the game, ideas came and went, each going through a split-second decision-making process about whether they should be implemented now, later, or never. Here are some ideas that were never implemented:

• Giving the game different behaviours depending on what you eat. Bad idea as it would give players real agency in the game and players were most likely to play only once; I preferred having one polished track than two half-done tracks, especially since players would only see one of those and leave the game forever to try one of the other 5000.
• Telling the player how much they eat. Again, too much agency. And I didn't want to reward the act of consuming; I wanted to make it digusting, yet addicting. Teeth Simulator has absolutely no "game goal" and I wouldn't fake one.
• Changing the kitchenware each "wave". I had the idea of how I would have done it and this would have brought great diversity without compromising the main message. It might even have reinforced it. Apart from the bowl, there would have been a wine glass, a plate and a skull.
• Teeth adjustment. Just slightly pivoting the teeth in their sockets left and right, with unsettling noises playing, while the mouse button is held.
• Particles when eating. As this would have been neat, I put feedback efforts into sound design. The food splitting into smaller parts was good enough.
• Stuck food would prevent respiration and lead to asphyxiation. Also, food going through the digestive tube would be "eaten". Both those ideas were considered, but discarded as what the game represented to me became clear.

About the sentence generation.. It is an idea so weird I don't even remember when or why I had it. It just kind of imposed itself in the game. I quickly created a small domain-specific language to cram all those words in. Its implementation was done on the last day.

As I created Teeth Simulator, the underlying message became clearer and clearer to me, helping in further design decisions, leading to the last implemented features of all: the unstable pupil that expands on a chewing spree. The underlying message isn't clever nor powerful. It is just a truth I had to express. Something I wanted to tell everyone without them knowing it: We are consuming destructively in a way that doesn't even feed us. We buy random stuff to possess, and as we buy, we vote for the destruction of what makes the Earth survivable to us. You noticed how when food goes through the digestive tube, it just falls? How there is no reason given to you to eat whatsoever except to increase your ability to consume? And how your teeth eventually fall anyway?  As said right on the game's page, we consume to to have an ephemeral and futile sense of agency in this decaying world.

We don't have to.

Anyway, that "message" was more of an inspiration to drive design decisions than anything else.