Draft & Limited
Eight players, three booster packs each, one card taken at a time and the rest passed on. When the packs are empty you build a 40-card deck from the roughly 42 cards you picked, with 17 lands and as many basics as you want. The pick order most players learn first is BREAD: Bombs, Removal, Evasion, Aggro, Duds.
Draft is a Limited format, which means you build your deck out of cards opened at the table rather than from a collection you already own. Nobody arrives with a deck. Everyone starts with the same three unopened packs, and the deck you play is assembled over about forty minutes of picking cards one at a time and passing the rest to your neighbor.
That single rule change is why draft feels different from every constructed format. There are no four-ofs, no perfect mana bases and no expensive staples, so a new player and a fifteen-year veteran sit down with the same raw material. It is also the cheapest way to play competitive Magic: you pay for three packs, and you keep every card you pick.
A sanctioned draft seats eight players, and eight is what a booster box and a store event are built around. Smaller pods work fine for a kitchen table, they just make each pack come back to you sooner.
Everyone opens their first pack at the same time. With current Play Boosters that is 14 cards you can draft, plus a token or art card that is not part of the draft.
Look at the pack, take one card, and put it face down in front of you. Nobody shows anybody anything.
Pass the remaining 13 cards to the player on your left, and pick up the 13 cards arriving from your right.
Keep going until every card in every first pack has been taken. You now have 14 cards.
Open pack two. This time pass to the right, so you are receiving from the player on your left instead.
Open pack three and pass to the left again. You finish with roughly 42 cards.
Build a deck of at least 40 cards from that pool. Add as many basic lands as you want, for free, from the store's supply.
Play three rounds of matches. Everything you did not put in the deck is your sideboard, and you can swap cards in between games.
The left, right, left rotation is not decoration. It means the players feeding you cards in pack two are the ones you were feeding in pack one, so the information you gathered in the first pack applies to a different set of neighbors in the second. That is the whole reason signals are worth paying attention to, and it is covered below.
Eight players is the traditional number and it is the hardest part of drafting to arrange. In June 2025 Wizards made that an official problem with an official answer: Pick-Two Draft, a format built for exactly four players that works with any Magic booster packs.
The arithmetic is the clever part. Four players taking two cards at a time consume a pack at the same rate eight players taking one card do, so three packs each still leaves everybody with about 42 cards. The pool is the same size as an eight-player draft, which means the decks are the same quality rather than a thinner version of them. Passing still goes left, right, left across the three packs.
Four players also means two rounds rather than three: the two winners play each other and the two losers play each other. Wizards' own playtesting put the whole thing, draft and deckbuilding and matches, at about 90 minutes, which is what makes it a weeknight format rather than an evening-consuming one.
Everything else on this page applies unchanged. The pick priority is the same, the 40-card deck is the same, and signals still exist, though they are louder: with four seats a pack comes back to you twice as often, so a color that is open stays visibly open.
BREAD is the pick-order mnemonic almost everyone learns Limited with, and it is genuinely the fastest way to stop making bad picks. It stands for Bombs, Removal, Evasion, Aggro, Duds, in that order of priority. When two cards in a pack look about equal, the one further up this list is the pick.
A card that wins the game on its own if nobody answers it. In Limited that is usually an expensive rare creature or planeswalker that takes over a board within a turn or two of landing. Take it first, almost regardless of color, because the games where you cast it are games you mostly win.
A cheap, unconditional way to kill a creature. Removal is worth more in Limited than in any constructed format, because everyone has bombs and nobody has four copies of an answer. A two-mana instant that kills anything beats a slightly bigger creature every single time.
Creatures the opponent cannot block: fliers, menace, trample, unblockable, anything that keeps attacking once the ground has gummed up. Limited boards stall, and the game is usually decided by whoever can still connect after both players have stopped being able to attack profitably.
Efficient ground creatures and the tricks that let them attack well. This is the bulk of the deck rather than the exciting part: enough bodies at enough different costs to have a play on turns two through five. A 3/3 for three is unglamorous and it is what wins matches.
Situational cards, expensive cards that do not win the game, and anything in a color you are not playing. Take them last. The useful half of this letter is that it gives you permission to pass a rare: a card you will not cast is a dud no matter what the rarity symbol says.
It is a beginner's tool and it is showing its age, so it is worth knowing the three places experienced drafters ignore it. There is no letter for card advantage, so a card that draws two and replaces itself gets filed under Duds by a strict reading. It ranks a mediocre flier over an excellent two-drop, which is backwards in most modern sets, because Evasion sits above Aggro. And it says nothing about your curve or the archetype you are actually in.
The practical version: use BREAD for the first half of pack one, when you have no deck yet and any good card might make it. From about the middle of pack two, the curve and the archetype take over, and a card that fills the two-mana hole in your deck beats a stronger card that makes your curve worse. Draft the deck, not the cards.
Every pack you receive has already been picked from. What is left in it is a message about what the players upstream of you are not taking, and that message is more reliable than anything you can infer from your own cards.
The clearest version: a card that is obviously one of the two or three best in its pack arrives on your sixth or seventh pick. Two or three players saw it and passed it. That color is open, and moving into it means the packs will keep rewarding you for the rest of the round. The opposite signal matters just as much, and is easier to miss: if you are in green and green has dried up completely by the middle of pack one, the player to your right is in green too, and you are fighting over a pool that only supports one of you.
Pack two is where signals pay off, because the direction reverses. Whatever you learned about your left neighbor in pack one is exactly what you need in pack two, when they start feeding you. A player who stays flexible through the first half of pack one and commits in the second half will end up in the open color far more often than one who decided everything on pick one.
The deckbuilding half of a draft is where most games are actually won or lost, and it is the part beginners rush. The numbers below are the Limited defaults, and they are defaults for a reason: decks that deviate from them usually lose to decks that do not.
Forty is a minimum, not a suggestion. Every card you add past 40 dilutes the pool you draw your bombs and removal from, and the cards you would add are by definition the worst ones you drafted. If you are hesitating between two cards for the last slot, take the cheaper one.
Basic lands are supplied by the store and do not come out of your pool, so the seventeen lands cost you nothing and you never draft them. If you are splashing a third color, the fixing has to be real: one dual land is not enough to reliably cast a double-costed card in your splash color.
Five errors account for most bad draft decks, and every one of them is a decision rather than a lack of knowledge.
Deciding your colors from your first pick and never changing. The first pack tells you what everyone else is not taking, and that information is worth more than one card.
Taking a card because it is rare. Rarity is a print-run decision, not a power level, and a common removal spell beats a mythic you cannot cast.
Drafting a five-color pile because every pack had something good in it. A deck that cannot cast its spells on curve loses to a boring two-color deck every time.
Playing 43 cards because you could not bear to cut three. Every card past 40 makes your bombs and removal statistically harder to find.
Counting a card you will side in as a maindeck card. Build the deck you want to draw from, not the deck that shows off your pool.
And one habit worth having instead: before you shuffle up for round one, count your two-drops. If there are fewer than five, cut the most expensive card in your deck for the best two-mana card in your sideboard. It is the single highest-value thirty seconds in Limited.
Count your lands, then count them again after any swap. 17 is the number.
Sort your pool by color before you build. You cannot see a two-color deck in an unsorted pile of 42 cards.
Keep the whole pool at the table. Your sideboard is every card you did not play, and matchups change.
A draft is three rounds against three different opponents with a deck that exists for one evening, which makes it oddly hard to learn from. The deck is gone by next week, so the thing worth remembering is not the list but the pattern: which color pairs you keep ending up in, whether your 2-1 records come from good decks or good draws, and how often the aggressive deck beat the durdly one at your table.
TurnZero HQ tracks Draft alongside Commander and Standard, so a draft night produces the same record every other night does: who played what, who won, and what the run looked like across all three rounds. Over a few months that answers questions no single evening can, starting with the useful one, which is whether the archetype you always take is actually the one you win with.
Eight players sit in a circle and each open the same booster pack. Everyone picks one card, keeps it face down, and passes the rest of the pack to the left. You keep picking one card from each pack handed to you until all the packs are empty, then repeat with a second pack passing right and a third passing left again. You end up with roughly 42 cards, from which you build a deck of at least 40 cards, adding as many basic lands as you like.
With current Play Boosters you draft 14 cards from each of three packs, so about 42 cards in total. Older Draft Boosters gave 45, which is why a lot of older advice quotes that number. You then cut that pool down to a 40-card deck, so you are playing a little over half of what you drafted.
Pick-Two Draft is an official Limited format Wizards introduced in June 2025 for exactly four players. Everyone opens three booster packs as usual, but takes two cards per pick instead of one, passing left, right then left. Four players taking two at a time empty a pack at the same rate eight players taking one do, so each player still ends with about 42 cards and builds a 40-card deck. Four players means two rounds, and the whole event fits in roughly 90 minutes.
Yes, and since 2025 there is an official format for it. Pick-Two Draft is designed for four players and produces the same 42-card pool an eight-player draft does, because everyone takes two cards per pick rather than one. It works with any Magic booster packs. Before that the usual workaround was to hand out extra packs, which works but is not what the packs were balanced for.
BREAD is the standard pick-order mnemonic for Limited: Bombs, Removal, Evasion, Aggro, Duds. Bombs are cards that win the game alone, removal kills a threat, evasion is creatures that cannot be blocked, aggro is the efficient creatures that make up the bulk of the deck, and duds are the cards you take last. When two cards look equally good, take the one higher up the list.
As a starting point, yes, and it will immediately stop a new drafter making bad picks. It has three known gaps: there is no letter for card advantage, it ranks a mediocre flier above an excellent two-drop because Evasion sits above Aggro, and it says nothing about your curve or your archetype. Use it for the first half of pack one, then let the deck you are building decide the rest.
17 lands in a 40-card deck, which works out to the same ratio as 24 lands in a 60-card constructed deck. Drop to 16 only if your deck is genuinely cheap and tops out at four mana, and go to 18 if you have several cards costing six or more. Basic lands are free and unlimited, so you never have to draft them.
Two, almost always. A two-color deck casts its spells on the turn it wants to, which is what wins Limited games. A third color is worth it only when you have real fixing (dual lands, a mana creature, an artifact that filters) or when the third color is one or two genuinely game-winning cards you are splashing for.
The cards being passed to you tell you what the players upstream are not taking. If a strong red card comes around on the seventh pick of the first pack, the two or three players to your right are probably not in red, which makes red an open lane for you. Reading signals matters most in the middle of pack one and again at the start of pack two, when the direction of passing reverses and you start receiving from the other side of the table.
In paper, yes. Every card you pick is yours to keep after the event, which is part of why drafting is a reasonable way to build a collection. On MTG Arena the cards are added to your digital collection instead.
Record every round of a draft night and see which archetypes you actually win with.
Create Free Account