Math and Magic: The Gathering go together like peanut butter and jelly. It seems like every week the community has a statistics dissertation explaining the odds of having something happen in a given game, given a certain type of deck construction. But there is a format out there that doesn’t have a need for that. It’s entirely deterministic. And it doesn’t even require the players to be in the same room, or looking at the same screen. But let me take a step back.
Back in the halcyon days of internet forums, the early 2000s, someone, somewhere, came up with a format that could be played entirely through private messages and public posts. One which doesn’t need player input, except of course for the initial deck submission. That format became known as Three Card Blind. In Three Card Blind (or 3CB as I’ll refer to it), each player in a given “tournament” will submit a decklist to the tournament organizer, usually just as a private message. The organizer would then seed the tournament, and run through the matches by themselves. They would then be able to determine the winner after going through all the matches.
But wait, you might ask, how in the world are they supposed to figure out how a match will play out. Well, the most defining thing about the format (and exactly as the name implies), each decklist only has three cards. Because of the limited size of the decklists, a match can actually be played more like a series of steps to determine the outcome.
The Rules of the Games
The way in which a game of 3CB plays out is actually fairly straight forward. The tournament organizer runs through as the "player" of each deck in a single match. They then play out the game, acting as both players, and run through two games. Why two? Well, in a normal match of magic, there is a slight variance in determining who will play first on game one. Since there are three games in a normal match of Magic, then one of the two players is going to have a slight advantage by going first twice. But 3CB is more about deterministically finding the winning deck. So instead, there are only two games, one where each deck plays first. This might result in a tie, but in a swiss system this shouldn't matter much for determining an overall winner. There are some other important differences to normal Magic play in 3CB. Each player plays with their hands revealed, so both will have perfect information of what the other is doing. This is important for determining what the "optimal play" would be. Players start the game with all three cards in their hands. Players do draw a card at the start of their turns like normal, however if they draw from an empty library they don't lose the game. For non-deterministic cards, such as which has a coin flip involved, the result of each "random" effect results in whatever effect is the most beneficial to the opponent. This is meant to prevent games from being "unsolvable", while still maintaining some balance to the game.
The tournament organizer runs through "optimal" plays in order to try and make each deck win. Let's say, for example, that player one has a deck of , , and , and that player two has a deck with , , and . Game one player one goes first. They play , . Opponent plays , and plays on the Hierarch. Player one can't do anything next turn, and player two follows up with a , which will slowly chip away for the win. Game to player two. Game two, player two goes first (even though they just won, both players get a shot at going first in one of the games). They play and . Opponent plays . Bolt on the Hierarch again, and player two takes game two. Match goes to player two.
These are some pretty bad decks, but hopefully they help illustrate how a match plays out. There might be some scenarios which would result in each player winning if they go first. This would result in a match result of one to one. A tied match. In whatever seeding system the tournament organizer is running, they point losses, wins, and ties accordingly, and go through rounds until a distinct winning deck is there. Usually wins incur three points, draws one point, and losses zero points. An important note about the deterministic steps is that the tournament organizer should record what steps they believe were optimal, and post them with the tournament results for others to inspect. This results in an entire history of every move in the tournament, one which in theory would be copied exactly if players submitted the exact same decklists with the same seeding.
Bans to Keep Things in Check
The goal of 3CB is of course to beat out your fellow competitors, but there's a spirit of the game that players are supposed to abide by. That is, you're supposed to find interesting ways to beat out your opponent, rather than simply coming up with the most powerful set of three cards that you can think of. In fact, the "ban" list for 3CB coerces the meta to prevent such things. Players often talk about how their decks in other formats get "banned". Well, Three Card Blind actually does ban decks. A whole swath of them in fact. Any deck that can win on turn one is banned. There are too many to list. As more cards get printed there's bound to only be more and more three-card combos that win on turn one, so any deck that would win on turn one is simply banned. This applies to effective turn one wins as well. For example, a deck of , , and does not technically win the game on turn one, but effectively makes a game state that can't be surmounted.
Discard effects also have been determined to largely be too good, so any deck that makes the opponent discard more than one card per turn is banned. Tournament organizers can also make distinct bans depending on how they want their tournament to run. There are regular cards that get banned from tournaments, such as , (largely for the reason stated above), and . The tournament organizer can (and often does) come up with constraints to help keep things interesting as well. Maybe that is just commons, or only Modern legal cards. The sky's the limit on choices!
Theory Crafting a Good Deck
A winning deck largely depends on the rules that the tournament organizer has provided, but I do have some thoughts as to what makes a list good. Generally, the way that decks seem to either lose or tie is by being a single resource shy of the opponent, so I believe that recursion effects are really good in this format. Likewise, cards that are "free" allow you to expend more resources against your opponent's. It gives you the ability to both react and be proactive in certain scenarios. With all that said, here's a deck that I think is sneakily good without being obviously insane
Three Card Blind List by Brian
The two key resources that I found to determine games were pressure from creatures and removal. If you can stick one power on the board, you win the game. Similarly, if you and your opponent both have one power and are racing, then you can sneak ahead just slightly with a ping effect. allows you to have a choice of reusing whichever resource of the two you need more. It also allows you to technically shock something, since you can a two toughness creature during your upkeep, then it to the top of your library before your draw step, then finish the creature off. being a land also gets around counterspells, though I'm not sure how big of an issue that would be for tournaments. And lastly, all three of these cards are uncommon (or at least have had an uncommon printing), so if that's the stipulation you're under I think it's a really solid choice.
If you have some friends that you play with, or keep in touch with online, this is definitely something fun to think about. Try organizing a little "event" yourself. Its something that I've done with people I was drafting with before, as something to have happen in the background during the event!
If you would like more information about this format, I suggest checking out this article from 2005 from MTG Salvation, and the Gamepedia Wiki for the format.
Until next time!