Introducing Archonitos

I wanted to create something to give away at Keyforge Celebration 2022 because I think giving weird things to strangers is fun. I figured Keyforge fans are interested in procedural generation, so I came up with the idea of characters based on the Archon shapes in Keyforge, but I wanted them to be silly and cartoonish. Hence the name Archonitos (little archons).

Each Archonito card is unique. I wrote a simple python script to make a set of random choices. The character image has five traits: shape, color, texture, eyes, and mouth. Each trait has ten options, for a total of 10^5 possible character images. Each of these options is stored in a GIMP layer and I can turn the visibility on or off for each selected trait to make the combined image.

Then the script generates a name (based on a little custom script I wrote) and assigns a random number between 0-100 for each of the three stats: Charm, Strange, and Fingers. Then each Archonito has a team name that is drawn from lists of science fiction and fantasy words for a science fantasy mashup (to be on-brand for Simon Finchley), a temperament drawn from a list of emotions and traits, and two card associations, one from a list of cards from popular trading card games and board games, and one from a list of tarot cards. So the total number of possible Archonito cards is quite large.

I wanted the cards to look like they might be from a game from an alternate-timeline 90’s, but I have not designed a game to go with the cards. At least, not yet.

I made the cards tarot-sized and gave each card a tarot correspondence so that, if I ended up with extra, the cards could potentially be used for patchwork tarot trades or for some tie-in with Tarot of the Trunk.

The first batch has fifty cards in it. I may do a second batch if there is interest, but the cards would be for sale to cover printing costs. I would also change a few things in the generation options so that each batch is unique. If I had to print a very large number, I would try to figure out how to more fully automate the GIMP-layer image generation process, which was manageable for such a small batch of images, but would quickly become unwieldy for a large print run.