How to Mulligan in Commander: When to Keep, When to Ship

The most important decision in Commander happens before you play your first land: do you keep this hand, or send it back?

Most players either keep everything (and spend 5 turns doing nothing) or ship perfectly good hands because they "don't see their combo piece." Neither approach wins games.

Here's how to actually evaluate Commander opening hands.

The Bare Minimum: What Every Hand Needs

Before we get into nuance, let's establish the floor. A keepable Commander hand must have:

1. Enough lands to function through turn 3-4

For most decks, that's 3-4 lands in a 7-card hand. With 2 lands, you need early ramp to make up the difference.

The math: you draw 10 cards by turn 3 (7 starting + 3 draws). With 36 lands in your deck, you'll draw ~3.6 lands on average. So 2 lands + ramp is fine. 2 lands + no ramp is gambling.

2. Something to do turns 1-3

A hand of 4 lands and 3 six-drops is unkeepable. You'll sit there playing lands while everyone else develops their board.

Turn-1 plays:

• Sol Ring, Arcane Signet, Llanowar Elves
• Swords to Plowshares (if you expect early threats)
• 1-drop enablers like Birds of Paradise or Esper Sentinel

Turn-2 plays:

• 2-mana rocks
• Removal for early threats
• Card selection like Ponder or Preordain

Turn-3 plays:

• Your commander (if you're on a commander-centric strategy)
• 3-mana rocks or ramp spells
• Card draw engines

3. A path forward

"I can cast these cards, but then what?" is a bad sign. Your hand should either:

• Deploy threats that win the game (aggro/midrange)
• Set up card advantage engines (control/value)
• Assemble combo pieces (combo)

If your hand is all setup with no payoff (or all payoff with no setup), that's a mulligan.

Mulliganing by Archetype

Different strategies have different hand requirements.

Aggro/Voltron

What you need:

• Your commander on curve (turn 2-3)
• Equipment/auras OR protection
• Enough lands to keep deploying threats

Ship if:

• You can't cast your commander until turn 5+
• All protection, no threats
• All threats, no protection (you'll get blown out by one removal spell)

Example keep (Voltron):

3 lands, Sol Ring, Swiftfoot Boots, All That Glitters, Sram, Senior Edificer

You can play Sram on turn 2, equip boots turn 3, start drawing cards. Even if Sram gets removed, you have enough mana to recast and protect.

Example mulligan:

2 lands, Sword of Fire and Ice, Sword of Feast and Famine, Colossus Hammer, Argentum Armor, All That Glitters, Blackblade Reforged

Zero ways to cast your commander. All payoffs, no mana. Ship it.

Control/Value

What you need:

• Lands to hit land drops
• Early interaction OR card draw
• A win condition (doesn't have to be in your opener)

Ship if:

• No early plays at all
• All answers, no card advantage (you'll run out of gas)
• Mana-screwed (control NEEDS to hit land drops)

Example keep:

4 lands, Rhystic Study, Counterspell, Swords to Plowshares

You have interaction for early threats, a card draw engine to bury opponents in value, and enough lands to operate. You don't need your win condition yet — control finds it later.

Example mulligan:

2 lands, Cyclonic Rift, Time Warp, Expropriate, Jin-Gitaxias, Core Augur, Consecrated Sphinx, Omniscience

This is "the dream" — if you could cast these cards, you'd win. But you can't. You're dead before you hit 5 mana. Ship.

Combo

What you need:

• Mana to execute your combo (or ramp to get there)
• Tutors OR combo pieces
• Protection for the combo turn

Ship if:

• No way to find your combo
• Combo pieces but no mana to cast them
• No protection (you'll get countered)

Example keep (Consultation combo):

3 lands, Sol Ring, Mystical Tutor, Demonic Consultation, Thassa's Oracle

You have the combo, mana to execute it turn 3-4, and a tutor to find Pact of Negation or hold up Force of Will if you draw one. This wins fast.

Example mulligan:

3 lands, Thassa's Oracle, Jace, Wielder of Mysteries, Laboratory Maniac, Leveler

All payoffs, zero enablers. You can't combo with this hand — you need Demonic Consultation or Tainted Pact, and you have no way to find them. Ship.

Midrange/Goodstuff

What you need:

• Curve (lands + plays for turns 1-4)
• Card advantage OR threats
• Some interaction

Ship if:

• All ramp, no payoffs
• All payoffs, no ramp
• No interaction against fast decks

Example keep:

4 lands, Llanowar Elves, Cultivate, Rhystic Study, Beast Within

You ramp turn 1, ramp again turn 3, deploy Rhystic turn 4, and have removal for threats. Clean curve, card advantage, interaction.

Example mulligan:

2 lands, Avenger of Zendikar, Craterhoof Behemoth, Finale of Devastation, Triumph of the Hordes, Overwhelming Stampede, Genesis Wave

This hand costs approximately 45 mana. Ship.

The London Mulligan: Use It

Commander uses the London mulligan rule: you draw 7, then put cards on the bottom equal to the number of mulligans you've taken.

This is MUCH better than the old "draw one fewer" rule. It means:

Mulligan to 6 is often correct.

You're looking for functional hands, not perfect hands. A 6-card hand with lands, ramp, and a plan beats a 7-card hand that does nothing.

Mulligan to 5 if you're in a high-power game.

At high-power tables, the game is often decided turns 4-6. A durdly 6-card hand loses to a focused 5-card hand that executes its plan.

Mulligan to 4 is rare but sometimes right.

If your strategy REQUIRES specific cards (combo), and your 5 doesn't have them, go to 4. You'd rather have 4 cards that win than 5 cards that lose slowly.

Common Mulligan Mistakes

Mistake 1: Keeping "Almost" Hands

"I have 2 lands and Farseek. If I draw a land, I'm fine."

No. You're gambling. The odds you draw a third land by turn 3 aren't great, and if you miss, you're dead in the water. Ship and find a real hand.

Mistake 2: Shipping Because You "Don't See the Win"

Your opening 7 doesn't need your combo pieces or your finisher. It needs to get you to turn 4-5 alive and functional. You'll draw 15+ cards in a Commander game — you'll find your win condition.

Exception: dedicated combo decks at high-power tables. Those need to see their combo or tutors.

Mistake 3: Keeping 1-Landers with "Good Cards"

Sol Ring, Mana Crypt, Arcane Signet, Llanowar Elves, and 3 lands — oh wait, only 1 land.

This is a trap. Fast mana is great, but you still need lands. A 1-land hand with 4 ramp pieces is a mulligan 95% of the time.

The only exception: 1 land + multiple

or ramp pieces that let you chain into more lands (e.g., Mana Crypt → Exploration → land from hand → Sakura-Tribe Elder). Even then, it's risky.

Mistake 4: Valuing Card Quantity Over Quality

"I'll keep this 7 because mulliganing to 6 feels bad."

Wrong mindset. A 7-card hand that does nothing loses. A 5-card hand that executes your strategy wins. Stop counting cards and start evaluating function.

The Mulligan Decision Tree

Use this flowchart:

Step 1: Do I have 2+ lands (or 1 land + multiple 0-1 cost ramp)?

• No → Mulligan
• Yes → Go to Step 2

Step 2: Can I make a play turns 1-3?

• No → Mulligan
• Yes → Go to Step 3

Step 3: Do I have a path to winning (or finding a win condition)?

• No → Mulligan (unless you're control and have card draw)
• Yes → Go to Step 4

Step 4: Does this hand match my deck's strategy?

• Aggro: Can I deploy threats early?
• Control: Do I have interaction + card draw?
• Combo: Do I have combo pieces or tutors?
• Midrange: Do I have a curve?

If yes, keep. If no, mulligan.

Power Level Adjustments

At casual tables (games ending turn 10+), you can keep greedier hands. You have time to draw out of awkward situations.

At high-power tables (games ending turn 5-7), mulligan aggressively. Every turn counts. A hand that "might work" loses to a hand that executes a plan.

At cEDH tables (games ending turn 3-5), mulligan to your combo or interaction. There's no time for "value." You're either winning or stopping wins.

Frequently Asked Questions

Should I keep a 1-land hand with Sol Ring and Mana Crypt?

Almost never. You need to draw lands to function, and even with fast mana, you're one land away from doing nothing. Mulligan unless your deck is VERY low to the ground (mostly 0-2 cost spells).

Should I mulligan a slow hand at a casual table?

Depends on what "slow" means. If you're doing nothing turns 1-3 but have a clear plan for turns 4-6, keep it. If you're doing nothing until turn 7+, ship — even at casual tables, that's too slow.

When should I mulligan to 5?

When your 6-card hand fails the decision tree. Don't keep a bad 6 out of fear. A functional 5 beats a broken 6 every time.

What if everyone else keeps 7 and I'm on a mulligan to 5?

So what? They kept bad 7s. You kept a good 5. Card quantity doesn't win games. Functional hands do.

Conclusion

Mulliganing well separates good players from great ones. Stop keeping bad 7s because "it might work." Start shipping aggressively and keeping hands that actually execute your strategy.

The rules:

• 2+ lands (or 1 land + serious ramp)
• Plays for turns 1-3
• A path to your win condition
• Matches your deck's archetype

Keep hands that meet these criteria. Ship hands that don't. Mulligan to 6 (or even 5) without fear.

Your win rate will thank you.

Want to track your mulligan decisions and see what works? Build your deck on GrimDeck and use our hand simulator to goldfish opening hands and practice mulligan decisions.