Proposal: storing booster generation instructions on the set info

In mtgjson, most sets have the boosterV3 property, which is an array of rarities. Let’s see, for example: Shards of Alara

"boosterV3": [
  [
    "rare",
    "mythic rare"
  ],
  "uncommon",
  "uncommon",
  "uncommon",
  "common",
  "common",
  "common",
  "common",
  "common",
  "common",
  "common",
  "common",
  "common",
  "common",
  "land",
  "marketing"
]

This does not reflect at all how boosters should be generated. I think better format for that. I looked around, for example, mtggen uses this format to generate packs: https://github.com/copperdogma/mtgen/blob/master/mtgen/wwwroot/ala/packs.json. That may be too unterse, but is readable.

They way I see it, it would be better to create a gen property in set, and that should hold the values needed to create a booster. That is a proposal on how to do it.

"booster" : [
  1,
  [
    1,
    "mythic rare",
    7,
    "rare"
  ],
  3,
  "uncommon",
  9,
  "common",
  1,
  [
    1,
    "foil",
    5,
    "common"
  ],
  "land",
  1,
  "marketing"  
]

and, also, the foil distribution, since this is a post Time Spiral set:

"foilDistribution" : {
  "land": 8,
  "common": 88,
  "uncommon": 24,
  "rare": 7,
  "mythic rare": 1
}

And maybe even some additional description, pointing to the source of the info.

“description” : “That’s the set that debuted the mythic rare rarity https://magic.wizards.com/en/articles/archive/making-magic/year-living-changerously-2008-06-02”

This gen property could be further extended to collect info about Tournament Packs, wal-mart boosters, and so on. Also, some cards “rarity” value does not reflect at all how they collate into the pack. Island is a “common”, on that set, that value does not reflect on how it is collated on a booster pack. So, Island, in that set, should have a “genRarity” : “land” property. That property should be optional, used only for cards that have a genRarity different from the displayed rarity. The land genRarity alone should help in issues like Modern Horizons having a snow land for each pack, or Guilds of Ravnica having a guildgate on each pack.

I know that it is a ton of work, but, that’s not an edge case, or a special rule. That’s the way usual boosters work since Alara.

Afterwards, it can be used to tackle the madness that is the Alliances rarity system https://mtg.gamepedia.com/Talk:Alliances It would also help with cards that have special rarities, like masterpieces, zendikar expeditions, power 9 in vintage masters, timeshifted cards, and so on.

---

Related:

• https://github.com/mtgjson/mtgjson/issues/225
• https://github.com/mtgjson/mtgjson/issues/265#issuecomment-471282586
• https://github.com/mtgjson/mtgjson/issues/35#issuecomment-423776108

[taw]

Sets like Fallen Empires, Homelands, Alliances and Chronicles, these sets have multiple cards in a rarity on a sheet, and often have card variants that further complicate things.

Those sets are actually 100% known, Wizards back then provided official information about that, you can check them out.

The commons are actually split onto two sheets where there are actually a handful of cards that printed more frequently. Which ones? We have no idea.

Wizards has certain rules about color variance in a pack, and rumor has it number of creatures and non-creature spells.

So what actually happens is that collation is done by hand. It includes color, estimated power level etc.

For a few sets we have full print sheets, but mostly we don’t. It’s just a limitation of what we have, we don’t make up fake data here.

The commons are actually split onto two sheets where there are actually a handful of cards that printed more frequently. Which ones? We have no idea.

There’s crazy amount of information here, if anyone really cared.

Your “weird sets” example is starting to become the norm. Wizards can now set rules on how boosters are made, like a Planeswalker in every pack. (It’s unknown enough that MaRo has guessed himself how it works.)

It’s still just a few sets, and it’s not clear if it’s really going to become common. It’s not even that we can’t figure it out, just that those special sets aren’t a priority.

And we can always ask for this information.

MTGJSON is a resource that is delivering factual information, not guesses.

We do a lot of guessing already. Like what’s with our rarity field really? Or releaseDate? Or format legality (Chinese standard where). It’s all a spectrum.

[ancestral]

Here’s my take:

I don’t like guessing.

MTGJSON is a resource with factual information.

If you want to build a booster generator, don’t put it in this project.

If the data can’t be sourced reliably, then it shouldn’t be captured.