Thursday, May 14, 2020

Expressly Forbidden - Fixing the "Forbidden" Mechanic

This article is not WotC safe.
What's more tempting than the forbidden fruit?

It all began with one of our very first glimpses at an unprinted mechanic - an infamous one - from Avacyn Restored "devign". But it's gone so much further, and has taken on its own life...

Look, if you dare, at the forbidden cards contained within this sealed containment cell called an "article". Unlock, if you are willing to pay the price, overwhelming power by "reading more after the jump"...

Slimebind | Limo Imobilizante - carta de Magic: the Gathering - MTG
Slimebind by Mark Behm
At the 2008 World Championships, Richard Garfield played an amusing deck with an unusual mechanic. It had the ability to shuffle cards called "Gunk", which did nothing, into the opponent's deck. These "Gunk" cards would blank future draws. When Mark Rosewater saw this mechanic, he was struck by the novelty of the idea of shuffling cards into your deck from outside the deck. As an offensive weapon, however, Gunk was both pretty subtle and unfun. It's definitely frustrating when you draw a Gunk card instead of an actual card, but is it particularly fun for either player? Is it particularly interesting? Mark Rosewater began exploring the fun core of the concept - shuffling cards into libraries from an outside zone - but in a more fun fashion.
Helvault by Jaime Jones
Around 2010, Avacyn Restored was likely deep in "devign", the period where design would hand off a file to development, but both would have their hands on it briefly so that design could make changes from the feedback given to them by development.

Avacyn Restored featured a bold, exciting, revolutionary mechanic. One that had some history, some cachet, and lots of novelty. That mechanic was not Miracle... at the time. Miracle has its own storied history, and would be made to work by Design and Development teaming up, but when "devign" began, design had been hard at work on a different mechanic.


Cards with the forbidden card would not start in your deck - they would be placed in exile. It was intended to capture the trope of the Helvault, an artifact that sealed away dangerous, evil, demonic creatures and magics from the world of Innistrad, the seal of which would be broken during the story. Forbidden cards would be shuffled into a player's library by cards that were not themselves forbidden, allowing you to draw them at a later point in time.

There's not a lot of detail on the exact execution of Forbidden, unfortunately. Did cards shuffle in specific forbidden cards? Any one from your sideboard? Could they shuffle in multiple that shared a name? However, we know what was the death knell for Forbidden. For one, it meant a LOT of shuffling, which is not an action you want all the time. For two, it had immense development issues. The fact you were shuffling just a couple cards into 30 or 50 card decks (probably) meant you had astronomically low odds of finding your Forbidden cards. The cards had to be worth going through this hoop and waiting to find, so they had to be powerful... but if you immediately drew it, then it was very swingy indeed.

Shuffling outside cards into libraries was just unfeasible. Not only was the action of shuffling too much, but the swinginess of when you'd draw the card made it unfun. Forbidden was set aside, though Mark Rosewater is confident he can make it work someday. 

At the same time, however, I too was confident I could make it work.

Stairs To Infinity by Steven Belledin : ImaginaryPathways
Stairs to Infinity by Steve Belledin
In 2018 - nearly a decade after Gunk - I was idly designing one day and thinking about Forbidden cards, when inspiration suddenly struck me.

Forbidden had two problems: it was a lot of shuffling, and even when you did shuffle, the cards did not make up a significant enough proportion of your deck to be worth shuffling in, because the chances of you drawing one were so small, even if you shuffled in all 4 of a card, or more.

What if the key was not to shuffle at all? What if you found a category of card that covered a significant proportion of the deck, and "stapled" the forbidden cards to these cards, so that when one such card was drawn, it would 'count' as having drawn a forbidden card? When thought of like this, the choice was obvious: land cards. Lands are themselves simple, so you don't get two cards worth of information in a single draw. Lands share relatively similar proportions of decks across formats and archetypes - though they do vary between aggro and control, that variance is something to welcome.

This would eventually lead to the Forbidden mechanic as it can now be seen in Xerex, a set I and my friend Passthechips are working on:

Sharpened edges removed for child safety.
Woah there, what's a "key"? Well, just like the original Forbidden mechanic, these cards need cards in your actual library to help you get them. The cards in your actual library grant "keys":
"I pass to you who'll vanquish me."
As you can see, a key is basically a "ticket" to get one future forbidden card. It means that, for every land card in your deck, you've "stapled" one forbidden card to it... but only for the next land card you draw. If you want more forbidden cards, you need to get more keys.

At first, we did not have "key" technology. We started very clunky - we simply moved cards from the forbidden library to a holding zone, and then when you drew a land, you moved a card from the holding zone to your hand. Introducing two new zones was complicated, but having them go to exile was also not ideal. We tried more radical things as well, such as this:
Because people love Chronostutter!
An "unsealed" card was on that was literally on top of the seal card. You see, when you 'unsealed' a forbidden card, you would move the seal one card downwards through your forbidden library, leaving a forbidden card on top of it. And you could only draw unsealed, available cards... This, you know, worked, but it was much more complicated then we needed.

It was Passthechips who ultimately came up with keys. They are wonderfully resonant and work both simply and effectively. In playtesting, we've found that the suspense of wanting to draw a land is similar to that of Landfall, but exponentially greater. It serves as a natural catch-up mechanism for those who are flooded. And with other mechanics and cards in the set, we can accommodate for the fact that lands are drawn less frequently as the game goes on.

Contraption - MTG Wiki
At last, Steamflogger Boss has purpose. Even if it didn't really make sense.
Finally, how did we address the forbidden library itself? One of the problems with Contraptions, in retrospect, was that you were greatly incentivized to draft not only as few of them as possible, but sometimes, literally just one of them. It was far better to tutor up a single Contraption consistently, in some cases, then it was to have two mediocre Contraptions and one good one to get 'value' from each of the CRANKs. 

How would we make sure that players were incentivized to draft Forbidden cards, rather than treat keys as a tutoring mechanism? We experimented with keys being able to dig further into your Forbidden deck, a la Sleight of Hand or Anticipate, but it always over-complicated what should have been a punchy mechanic. A minimum deck size would make it so that your drafts could just completely whiff if you missed the mark by the skin of your teeth, so that was out. However... it was close to the idea. We wanted players to be drafting towards a certain benchmark, but we didn't want them to be punished all-or-nothing if we missed it.

It was my friend Tommy Occhipinti who came up with the simple, novel solution... by looking to how actual Magic does it. What do you in an actual Magic game if you want to play, but you don't have enough cards? For example, if you draft Yorion, Sky Nomad, how do you fill the blank space in your deck if you don't quite reach 36 playables? The answer was simple: fill it up with basic lands.

And you thought Wastes were bad.
Okay, not the basic lands you're thinking of. That was the original suggestion, but I felt there was a kind of 'sanctity' to the forbidden library only containing forbidden cards. In addition, if keys read "when you draw a land, draw another land", that's actually not that bad. We wanted to really make you feel like you were being punished for not reaching the minimum number of cards in your forbidden library, which is why we went with Depths. That might ultimately be the wrong call! We're still playtesting Depths now, to figure them out. But it's an elegant solution, though one only designed for Draft currently.

One of the most interesting things about lands in the forbidden library is that they 'soften' the whiff. If you have excess keys - rewarding dedication to the archetype - you can immediately spin the wheel and draw another forbidden card. Since the deck comes pre-stacked to put Depths on top, this means that if you want to dig through the deck, you need to draft a good balance of both cards that grant keys as well as forbidden cards. However, whichever way your proportion falls unbalanced, it works out. Too many keys, too little forbidden cards? You'll dig through the Depths faster. Too many forbidden cards, too little keys? All your keys will hit gas.

If you thought normal librarians were scary...
This may be our current iteration of Forbidden, but who knows if it will end up our final one? There's a lot of playtesting left to be done, and beyond even that, we have to figure out exactly how the pack collocation works. Currently, we have one guaranteed colorless forbidden spell in each pack replacing the basic land, and mix colored spells among the other cards as normal, for an as-fan approaching 1. (Unfortunately, Planesculptors does not support more advanced 'slot' customization for drafts.) So far this has worked well, but we'll be testing adjustments, and making haste to add uncommons and rares so we can test how it works out with different proportions at different rarities. While most set playtests linger in the commons for a while, the unique demands of Forbidden need to be solved a lot more holistically!

So who knows - maybe next year, I'll write another article like this one, and we'll see what's changed. But I can confidently say, given our playtesting so far, and the fun we've had... this mechanic does work. It might need improvement, it might not work in Constructed, but in our format? It works!

In a regal age ran I.
I hope this article was fun and informative. I encourage you to check out our set channel at the Custom Magic Discord, where you can join playtest drafts, as well as provide suggestions, feedback, and commentary on the set. 

I've just recently gotten back into the 'game' of custom Magic, so you can look forward to more articles here as well in the near future! Perhaps next time we'll discuss how we go about designing Forbidden cards, or how we conducted playtesting them. Or maybe I'll just show off some cool designs. What would you like? Let us know in the comments!


  1. Oh cool! Yeah, I love forbidden, it's really flavourful and exciting.

    I keep thinking of other ideas that stem from that seed, like "look at the bottom N cards of target player's library, shuffle one of them into the top N cards of their". But that's just always still going to be fiddly and unfun.

    I love the flavour of the forbidden lands, but it still sounds fiddly trying to draft many forbidden cards (unless like 1/3 of the set is forbidden? I guess maybe it is). I was expecting you to say you just got the choice of which card to unlock when the land came. Or, each card is unlocked by a particular type of land so you need a mountain, not just a land.

    1. Those are possible areas to explore, yeah!

      As-fan is a big issue. Currently, with an as-fan of 2 per pack, you're guaranteed to see 6 in your draft at the very least. So you have to work a little harder than the minimum to get a full forbidden deck... and to make it a synergistic and good one, a little harder than that.

      It's something we're working on very actively. Maybe the minimum size goes down. Maybe we add more keys in the set. Maybe we use Cycling instead of Surveil to make draws happen more often, but vice versa, preserve the suspense of whether you get a land or not. Etc...

  2. Would love to know more about the playtesting process, and also about the flavor and/or game experience you're hoping players get out of your implementation of the mechanic!