TLDR: Keep a reactive opening hand when its mana works, its interaction is likely to matter on time, and the hand can convert the time it buys into development, cards, pressure, or a stable position. If you are asking should you keep a reactive opening hand in magic, do not count removal spells and call the hand safe. Ask what happens after you cast the first answer.
A hand containing three lands and three answers can be excellent or nearly nonfunctional. The difference is not how reactive it looks. The difference is whether those answers line up with the opponent, whether you can cast them without neglecting your own plan, and whether the remaining cards give you a route to get ahead.
What counts as a reactive opening hand?
“Reactive opening hand” is a strategic description, not a formal rules term. It usually means a hand weighted toward removal, counterspells, discard, protection, sweepers, or other answers rather than threats, acceleration, card selection, or engines.
Reactive does not automatically mean defensive. A tempo deck might counter one spell to preserve an early attacker. A combo deck might hold protection for its decisive turn. A control deck might trade resources until its card advantage takes over. In each case, interaction serves a plan. Trouble begins when the hand contains answers but no clear idea of what those answers are buying.
Use this question as the first diagnostic: after I spend my first answer, what does this hand do next? A strong answer might be “cast a card-selection spell,” “deploy a threat while holding up another interaction spell,” “reach a sweeper,” or “untap into my card-advantage engine.” If the honest answer is “draw something useful,” the hand is relying on hope rather than structure.
The four tests for keeping a reactive opening hand
1. Can you cast the relevant spells on time?
Start with mana, because theoretical interaction is worthless while stranded in hand. Check the number of lands, colors, untapped sources, and the actual turn on which each spell becomes available. A two-mana counterspell does not protect you on turn two if your second source enters tapped or produces the wrong color.
Then check whether holding up interaction conflicts with essential development. Suppose your only early choices are casting a mana rock or leaving two mana open. If you counter something, you may delay your own four-mana play by a full turn. That can still be correct, but the hand should be judged according to that cost—not as though it can develop and interact simultaneously.
Land count alone does not settle this question. When evaluating a lean reactive hand, apply the same color, curve, and missed-land-drop tests used for a two-land opening hand. A hand with fewer lands can work when its early plays help find mana or when its curve remains functional after one missed draw. Expensive answers do not provide that flexibility.
2. Are the answers actually live?
Broad interaction is safer against an unknown opponent. A flexible removal spell or counterspell may have several useful targets, while graveyard hate, artifact removal, or a creature-only answer might do nothing. In an unknown Game 1, discount narrow interaction unless your format makes its target overwhelmingly likely.
Matchup knowledge changes that calculation. After sideboarding, a narrow answer can be among the best cards in your deck because you know which threat or resource it stops. Reid Duke’s mulligan framework emphasizes this contextual value: a card that is weak in one matchup can be a major reason to keep in another. That article predates the London mulligan, so its strategic framework is more useful here than its historical procedural details.
Even a relevant answer must line up on timing. A counterspell counts as an early play only if the opponent is likely to present something worth countering while you can hold up the required mana. Against a deck that develops through cheap permanents before your counter is available, the card may not cover the window that matters.
Also consider whether the expected threat can resist your answer. Target restrictions, ward costs, hexproof, and protection can turn apparently efficient interaction into a mismatch. Knowing how ward, hexproof, and protection affect interaction helps distinguish a live answer from one that merely resembles a live answer.
3. Does the hand convert time into progress?
One-for-one interaction usually preserves the current balance rather than creating a winning position. That is useful when your future draws are stronger, your existing threat is dealing damage, or your engine will eventually produce more cards. It is less useful when both players are drawing toward powerful plays and only the opponent has a developed plan.
Look for at least one conversion mechanism:
- Card selection that finds lands, threats, or the correct second answer.
- A threat that can be deployed before or after interacting.
- An engine that produces recurring cards, mana, or board presence.
- A stabilizer such as a sweeper or resilient blocker that catches up broadly.
- A matchup-specific plan in which stopping one pivotal spell creates a large advantage.
You do not need all five. You do need something more concrete than “my deck has good topdecks.” Every deck has good draws available somewhere in its library. The mulligan decision concerns what this hand can reasonably accomplish with ordinary draws.
4. What does another mulligan cost?
Under the London mulligan, a player taking a mulligan draws a new hand equal to the starting hand size, then puts a card on the bottom of the library for each mulligan taken. The procedure was introduced to improve access to functional hands while preserving a cost for repeated mulligans.
That means a weak seven should not receive extra credit merely because it technically casts spells. Seeing a fresh seven and selecting the least useful card to bottom gives you meaningful control over a six-card hand. Conversely, once you are already choosing whether to go to five, a merely functional reactive hand deserves more respect. Mulligan standards should tighten as the cost rises, an important principle even though older strategic discussions used a different mulligan procedure.
Two examples: the functional keep and the trap
A reactive hand worth keeping
Consider a seven-card control hand with three lands in the necessary colors, a cheap removal spell, flexible interaction, a card-selection spell, and a four-mana stabilizer. This hand passes all four tests. It can cast its spells, answer different card types, use card selection when no answer is required, and work toward a play that can stabilize the board.
Its strength is not that it has two answers. Its strength is that every early branch remains productive. If the opponent presents a threat, the hand interacts. If the opponent does not, it improves its draw. If the first exchange is unfavorable, the stabilizer offers a broader recovery plan.
A reactive hand that should usually go back
Now consider three lands, two creature-removal spells, a narrow artifact answer, and an expensive finisher. Against an unknown opponent, this is often a mulligan. The mana may be functional, but the interaction could miss entirely, and the finisher does not repair the empty turns before it becomes castable.
The same hand could become keepable against a known creature deck that relies on a key artifact, especially if the lands cast every spell cleanly. That is why fixed rules such as “keep any hand with three lands and two removal spells” fail. Card role is contextual.
Adjust the decision for your deck and matchup
Control decks can keep reactive hands more often because trading resources is part of their route to the late game. Even then, an interaction-only hand can fail if its answers are too narrow or it lacks the mana and card flow needed to continue exchanging.
Midrange decks usually want a balance. Interaction is valuable when it clears the way for a threat or prevents the opponent from going over the top. A hand that answers three things but applies no pressure may give a ramp, combo, or control opponent enough time to draw through those exchanges.
Aggro, combo, and other linear decks should be more suspicious of answer-heavy hands. Their interaction usually exists to protect or enable the proactive plan, not replace it. A protection spell without a threat, or disruption without a clock or combo development, may not be a functional hand.
Play versus draw also changes the timing test. On the play, you may deploy a threat or mana source before the opponent can establish their position. On the draw, you see an extra card sooner, but the opponent gets the first development step. Do not treat that extra draw as compensation if your interaction is already one turn too slow.
Commander requires a higher development standard
In Commander, a hand containing lands and several one-for-one answers can look responsible while quietly putting you behind. You have three opponents producing resources, and trading one card for one opponent’s card does not address the other two players. A good Commander keep normally wants development—ramp, card advantage, a useful early permanent, or commander access that supplies one of those functions—alongside flexible interaction.
Multiplayer mulligan rules also make the first decision more forgiving: a player’s first mulligan does not increase the number of cards put on the bottom. Multiplayer players also do not skip their first draw step. In an ordinary Commander game, that makes it easier to reject a seven that has answers but no development.
Do not abuse that freedom by searching for perfection. A hand with workable mana, an early development play, and flexible interaction is usually doing its job. The purpose of the first mulligan is to improve functionality, not guarantee the strongest possible start.
The fast keep-or-mulligan checklist
Before keeping a reactive hand, answer these questions in order:
- Can I cast my early spells with the lands and colors already present?
- Will at least one answer be live during the turn that matters?
- Is the interaction broad enough for an unknown opponent, or specifically strong in this known matchup?
- What productive play can I make if the opponent gives me nothing worth answering?
- After the first exchange, how does this hand develop, gain cards, apply pressure, or stabilize?
- Does holding up interaction delay my own plan more than the answer is worth?
- How costly is another mulligan at my current hand size?
- In multiplayer, does this hand develop enough resources to compete with several opponents?
Make the hand explain its next turn
The practical answer is to keep a reactive opening hand only when the interaction forms part of a complete sequence. You should be able to describe what you can cast, what you are trying to stop, and how the resulting time advances your plan.
On your next mulligan decision, do not ask, “Do I have enough answers?” Ask, “What does this hand accomplish after the first answer?” If the sequence leads to development, pressure, cards, or stabilization, keep it. If it leads only to waiting for a topdeck, take the mulligan while the cost is still reasonable.