In 2011, three friends searched through a giant stack of chaff to make a silly single-use draft experience. But much like the Eiffel Tower (built in 1889 as a gimmick for the World's Fair), I recognized the potential of a curated "bad card cube" and I've been maintaining it ever since.

The Sucky Cube has a very specific design philosophy intended to keep games fair and fun. Here are the rules I set for myself:

1. The ideal reaction when looking at a given card in this cube should be "Haha, wow, that seems bad." You're not going to see Squire here, since it's obviously a bad card. It's funny, but it's also boring and no one's excited to play it. We want cards that are bad, but also unique and interesting, like Homarid and Mages' Contest!
2. Hidden synergies. Ideally, cards should be bad because they lack support, not because they're objectively bad. Hunter Sliver can be powerful in a sliver deck, but here, it feels like a great accomplishment when you can line it up with Sliversmith! In fact, a main theme here is tribal cards with very little tribal support. This makes changelings and cards like Artificial Evolution potentially quite powerful!
3. Build-around cards. While there aren't true archetypes beyond the hidden synergies, certain cards are designed to be built around, like Last Stand and Tyvar, Jubilant Brawler. There are also several mini-packages (often only 3-5 cards), like Mercenary tribal, Multicolor synergy, and Learn/Lessons.
4. Consistent power level. There are lots of bad cards out there in the 1-3 mana range that are completely unplayable, but bad cards that cost 4 and up are often bad for their cost. We need to have cheap cards that aren't completely outclassed later in the game. Note there are some bombs that are a little "pushed", like Icatian Town and The Raven's Warning.
5. Creatures also scale in a fair way. 3/3s for 5 will do a lot of work in this environment, but you can also get 2/1s for 2 just like Garfield intended! The basic rule is that a creature's mana cost is 1 less than the sum of its power and toughness.
6. Removal shouldn't be too versatile. A creature should cost about as much to kill as it does to summon, and unconditional removal should cost around 6. Cheaper kill spells are allowed if it's a two-for-one at your expense.
7. Avoid effects that can end the game without warning. In real Magic, "creatures can't block this turn" is like a 3-mana effect, but this effect has no place here at almost any cost.
8. Avoid cards that can single-handedly take over the game if left unchecked. The Hive was actually cut for this reason; it's just too good in the long game. Exception: if it's a creature with 1 toughness (e.g., Vitaspore Thallid), it's probably okay, as there are a lot of 1-damage removal spells. In fact, every color has access to them! (Thanks, Psychic Purge and Hornet Sting.)
9. Milling effects and lifegain must be very rare! Games can go long, so milling would be too good (especially in 2HG!) and lifegain rarely does anything except prolong the game.
10. Tutoring effects, recursion, and copying can be fair even around the usual rate. What are you gonna use it on, a good card?