Welcome back to Making Magic with chezkevin, my series where I talk about design decisions for a recent Magic: The Gathering Commander reskin project.
Readers familiar with Magic: The Gathering may realize that this is named after Mark Rosewater's blog about Magic design. My apologies to him😉.
Today we're talking about just that: design, and how this project took root and bore fruit (the root is my brain and the fruit is cardboard).
The secret fun part about this was looking at all the Metroid fanart that exists out there -- and there's a whole lot. I really enjoyed looking at a piece of art and mapping it to a Magic card in my head. So instead of keeping it there, I put it in a Google Spreadsheet. The spreadsheet had numerous columns, notes on the Metroid analogue, links to the art, links to the Scryfall page for that card. Scryfall was invaluable as a database for existing Magic: The Gathering cards: I used card design webpages and I had to get the right frames, pinlines, crops, rules text and more. I was making last-minute tweaks even after I sent my cards off to print; there are just so many details.
I used Imgur to host art images, then applied them to MTGCardBuilder at first. You can find the MTGCardBuilder archive here. However, once I finished my initial 91-card decklist, I realized that I couldn't actually send them to print due to a copyright which MTGCardBuilder adds to every card. I looked to the mpcproxies subreddit which specifically calls out this issue, and recommends CardConjurer.app instead, so that is where I went.
I also used Paint.NET, a free image editing application (poor man's Photoshop), to edit any images as needed. Sometimes, it was faster to edit an image from Paint.NET using existing resources rather than pulling up another design on CardConjurer.app.
Finally, I used ChatGPT for numerous purposes: image upscaling, image editing (ie, removing the HUD from video game screenshots), flavor research, and QA passes for things such as transparent seams on cards. ChatGPT was a assistive tool, and all the creative decisions were mine. For example, ChatGPT would suggest 10 different flavors based on existing first-party lore alongside its own generations, and I would choose or iterate on them.
It was particularly helpful for keeping the spreadsheet updated, because that was a lot of hassle that I did not need. A small summary of my tools:
- Deck Decision Ledger Google Spreadsheet
- CardConjurer.app
- Paint.Net
- Imgur.com
- ChatGPT.com
- Worth mentioning: Pinterest/Deviantart for fanart
The spreadsheet evolved as I needed to reconcile 'cool cards that I found art for' with 'cards that actually belong in the deck I'm building.' Eventually I split out into a 'decklist' tab and a 'sideboard' tab where I could keep track of the collector #'s, and separate those two categories out while still designing both. You could say the process itself, mutated as I worked on it.
Special Treatments
I reserved treatment prints for the following, in general:
- Pixel art
- Vertical art
- Exceptional art
That's where the card design can get fussy, but it can also give some very pleasing aesthetics. So it's often worth the hassle. I found that the biggest hassle for me was the Nickname frame. Some Nickname frames have complete opacity on CardConjurer while others only have partial. In general I restrained myself from re-naming cards for the sake of it. There's a certain classiness to it when you can do that. Of course, other times specifically name proper nouns so those have to be. I'll just close with a small gallery of cards I loved designing. I'll also list all of my assets below.
Chezkevin Assets (All Google Drive)
- Deck Primer
- Mainboard
- Art Cards & Tokens
- Sideboard (Includes Sideboard Treatment prints)
- Mainboard Treatment prints
- Extras (just off-color lands and cards I forgot to print)
- Deck Decision Ledger
Read the full series:
No comments:
Post a Comment