Mulliganing Strategy: When to Keep or Redraw
The opening hand is one of Magic: The Gathering's most consequential decisions, and it happens before a single card hits the battlefield. Mulliganing — the act of shuffling a hand back and drawing fewer cards — is a structured rule set that gives players a controlled escape from unplayable starts. Understanding when to use it, and when to live with an imperfect hand, separates players who win despite bad draws from those who lose because of them.
Definition and scope
A mulligan is the formal decision to reject an opening hand and draw a new one, governed by the London Mulligan rule that Wizards of the Coast adopted universally in 2020. Under the London Mulligan, a player who chooses to mulligan shuffles their hand back into the library and draws 7 cards again — then, once all mulligans are complete, puts a number of cards on the bottom of the library equal to the number of times they mulliganed. A player who kept at 5 cards effectively saw 7 but chose 5; a player who kept at 4 chose 4 from a fresh 7.
This structure replaced earlier systems that drew one fewer card per mulligan immediately, making the London version significantly more forgiving. It is format-agnostic: the same rule applies in Standard, Modern, Commander, and every sanctioned Constructed format. Commander uses a modified "free mulligan" (the first mulligan is taken to 7, not penalized) due to the singleton nature of 100-card decks.
How it works
The practical sequence runs as follows:
The bottoming decision is where substantial skill lives. A player keeping at 5 looks at 7 cards and sends 2 to the bottom — essentially filtering out the worst 2 cards in that 7-card draw. Correctly identifying which 2 cards are expendable requires knowing the deck's functional redundancies and which cards are dead draws in the expected early game.
Common scenarios
Land-light hands. A 1-land hand with high curve cards is almost always a mulligan. A 2-land hand with a 3-drop, a 2-drop, and interaction is frequently keepable, depending on whether the deck runs 24 or more lands and whether both lands are untapped. Aggro decks running 20 lands can justify keeping 2-land hands more readily than control decks that need 4 lands to operate.
Flood and screw on the same hand. Paradoxically, a 5-land hand is often worse than a 1-land hand — at least 1-land hands have action. Five lands and 2 spells means drawing more lands is actively harmful. Most competitive players mulligan heavy land hands without hesitation.
Format-specific considerations. In Vintage and Legacy, free spells like Force of Will or Daze mean a 6-card hand with 2 lands, a cantrip, and Force of Will can be stronger than a clunky 7. In Commander, the free first mulligan means aggressive mulliganing for a playable hand costs nothing on the first shuffle.
The "functional" one-lander. Certain decks — storm combo decks being the clearest example — can keep 1-land hands that include a fetch land and a cantrip, because the cantrip finds land 2 often enough to justify the risk. This is deck-dependent, not a general principle.
Decision boundaries
The clearest framework available in competitive play comes down to 3 questions, applied sequentially:
- Does the hand have a functional mana base for the first 3 turns? A hand that cannot cast anything until turn 4 almost never justifies a keep at 7.
- Does the hand have a plan? A hand full of situational cards with no coherent threat-or-answer structure is a hand without a purpose. Card advantage and tempo principles apply here: a hand that answers things but cannot develop a board, or develops a board but cannot answer threats, is fragile.
- Does the hand improve on a mulligan to 6? This is the hardest question. A clunky 7 might be better than the realistic average 6, because the bottoming effect only partially compensates for the card lost. A skilled player does the math on what a 6-card hand from this deck typically looks like.
The contrast between aggro and control strategies is pronounced here. An aggro deck like a mono-red burn strategy wants 2-3 lands, 4 cheap threats or spells, and doesn't care if the hand is redundant — redundancy is fine when every card does the same job. A control deck like an Azorius draw-go shell needs land, early interaction, and at least one late-game payoff represented somewhere in the hand or on the next few draws. Keeping a 7-card control hand with 5 lands and 2 counterspells is not keeping a "good hand" — it is keeping a slow hand that will run out of cards.
The foundational resource for the rules themselves is the Magic: The Gathering Comprehensive Rules, maintained by Wizards of the Coast, which defines the mulligan procedure in section 103.4. Players new to the broader game structure can orient themselves through Magic: The Gathering Authority, where rules interactions and format distinctions are treated in depth. For hands involving complex board states or sequencing decisions, reading the board state becomes as relevant to the mulligan calculus as the hand itself.