Considering redoing a Legends theme, would a special sheet of legends be the way to go?

There were many problems with the way Legends were handled in Kamigawa. One major problem was they took up most of the rare spots, while a few showed up at uncommon and none at common. It has been claimed that you can't have a block theme if it doesn't show up at common. Furthermore, it doesn't leave much room for other rares.

Can a special sheet of legends fix this issue? Following the Innistrad DFC and Dragon's Maze lands model (not Time Spiral Timeshifted), this special sheet can contain all rarities through proper distribution. Furthermore, a legend is guaranteed in each pack without neglecting other nonlegendary rares, albeit some show up more commonly than others. Thus the theme would be "complete" because it shows up at common.

For that matter, what would a common Legend look like? Probably vanillas like Isamaru and the ones from the Legends set.

If there are a lot of them.. how are they legendary?

"Oh great, ANOTHER ..." really isn't the feeling you want from a legend. The game-play of having the legend super-type is, after all, entirely downside.

Oh, stupid me. The obvious solution is to have the legend sheet be really really large, so that although you are guaranteed a legend (making the theme common) any given legend is still rare.

Another problem with Kamigawa block was that, although the theme wasn't at common, it was also trying to be themed around a mechanic inherently about uniqueness. Kamigawa block had a lot of "forgettable" legends. Loads of rare kami that blend into one another. It diminished what it means to be a legend because the legends couldn't all be exciting and eye-opening.

This would still be a problem under this proposal.

Mmm, that's true. There was a certain "Huh, that's a legend?" quality to it. Not helped by "I'd much rather run some random uncommon instead".

just because they get their own sheet, doesnt mean there are a lot of them. innistrad had only 36 dfc and dark ascension 22. if these were same numbersr for legends, it's still more than the average set, but not overwhelming.

you still only get 1 common legend per pack (not discussing foils). you may get duplicate for some legends, but not that often. more likely you'll get different legends, so you get to see them duke it out.

i did say this followed the innistrad model instead of time spiral. looking back, this may be the better model. all legends are evenly distributed, no common legends. so the chances of seeing a particular legends is even less frequent. and you pack more unique legends per sheet and per set.

If there aren't a lot of them then at one per booster, you'll be seeing some of them quite a lot. Suppose there are 22; then you'll be seeing any given legend with about the frequency of any given uncommon. 36 may be about the sweet spot so they're present in every booster but not overwhelming... but that's still 36 legendary creatures that you need to make distinct not just from each other but also from your other rare cards.

It's challenging. But, hey, that's why we make custom cardsets, right? For the challenge of it? So yeah, go for it. I think it'll be tough to make it work, but not impossible.

I'm also amused that these will show up as basic ;)

thats not how probability and rarity works. you can have 36 unique cards and have them show up at different rates. hence common, uncommon, and rare DFCs. otoh time spiral used uniform distribution for 121 cards. then each card is equally rare, with 1/121 chance to get any given card. basically they had same rarity as mythics do now.

so first you figure out which distribution model to use. from there you find out how many unique cards to make and at what rarity.

part of the appeal with openig innistrad packs was a higher chance to open 2 rares and/or mythics (foil makes it 3) in a single pack.

Uh? Yes, I know about Innistrad's double-rare packs (and Planar Chaos's ones too), but I was responding to a comment in which I thought you indicated a preference for a uniform distribution, a flat rarity scheme: You wrote " all legends are evenly distributed, no common legends."

well for uniform distribution, you just have to decide which rarity to follow and there's the number of unique cards you need. rares are about 1/60 in big set, so 60 unique cards. uncommons 1/20, so 20 uniques; commons 1/10. we can forget mythics' rate of 1 in 120.

also my numbers above are off because i didn't search correctly. there were only 20 and 13 DFCs in Innistrad and Dark Ascension, respectively.

they don't have to replace basic lands in the pack. it could be one of the common slots.

uniform distribution has the benefit of lower chance of getting the same card twice. however, is that a good thing? why can't two players pick their own copy of the same legend, or the same player picking multiples of the same legend (e.g. to get a better chance of drawing it in a game)?

also, it would be more interesting to design legends of varying rarities. simple ones like Isamaru can be common, and then there are crazy, cool ones that deserve to be rare or mythic.

Part of the problem with Legendary is that, after a few cute tricks dealing with how they get sacrificed, there really isn't much to work with, design-wise. It's just a toggle switch, in the same way that you can write any word after the type em dash, and call it a theme. "All Legendary creatures get +2/+2" seems like a cool ability. But there's only so many times you can keep going to that well. "Target Legendary creature gets first strike." "Whenever a Legendary creature does X." etc., etc. Do that enough times, and you end up with a boring set. Or, in other words, you end up with Kamigawa.

That's not to say one can't get innovative with the Legendary design. It's just important to see the missteps for what they are. Personally, I think you could go rather far with a keyword mechanic that triggered off Legendary. That way, you can keep reinforcing how important Legendary creatures are, without constantly reinventing the Legendary bonus (and without constantly boring players with repetition of the word 'Legendary', since that will be in the reminder text.)

Instead of looking to Innistrad, I'd look to Coldsnap, and the use of Snow-Covered Lands. Over there, they proved that a special sheet can be used to have a real effect on the game... though, any player could play an off-color Snow land, and Snow lands are hard to get rid of, so maybe they aren't the great example I want them to be. I think the Snow theme worked perfectly for Coldsnap, but that required one special sheet worth of cards that any player could use, and a smattering of other permanents that produced snow in the regular line-up. In order to make a set that's 'about' Legendary permanents, you'd probably need a special sheet, a number of rares, plus numerous uncommons that dipped into theme. Otherwise, you might be able to insure that each player can get 3 Legendaries, on average, in their draft pool... but you can't guarantee those Legendaries will be castable...

This is relevant since the new collation technique used for Dominaria basically achieves something really similar to a special sheet. I'd say collation is a new answer to what otherwise would have used a special rarity/sheet.

No common legendaries though, so that's out of the way. As an additional though I actually think separate legendary sheet would have been welcome - even if the current solution is great as well.

The collation method for Dominaria is more complex, and that means higher chances of errors. Sirgog noticed that on MTGO, which tries to emulate the paper collation method, that Dominaria packs are missing the foils for legendary cards. He suspects the error might be due to the complexity of the collation.

Special sheets have been done several times before, so experience and simplicity makes it less prone to errors. Moreover, any errors are easier to test, detect, and rectify.

I personally am a fan of special sheets and when I create any kind of special tricks like this I go for simulated special sheets as well.

As you say: Less prone to errors, and easier to realise to begin with.