Skip to content

MTG Commander Mulligan Guide: Keep or Ship

Table of Contents

You can build the cleanest 99 in the world and still lose on Turn 0 if your opening hand doesn’t actually play Magic.

In Commander, mulligans are less about finding “the nuts” and more about avoiding the most brutal thing in multiplayer: falling behind and staying behind. Let’s fix that.

TL;DR (Keep This Mental Checklist)

  • Keep hands that can make your first 3 land drops or have 2 lands + ramp and a clear plan.
  • Ship hands that are missing colors, missing mana, or missing a plan (seven cards that do nothing together).
  • In faster pods, prioritize untapped mana + early interaction. In slower pods, prioritize ramp + an engine.
  • Your “good” keep changes a lot based on archetype: midrange wants setup, combo wants access, control wants stability, stompy wants curve.

Commander Mulligans: The Rule in One Breath

Commander uses the London mulligan: every time you mulligan, you shuffle, draw 7, then put a number of cards on the bottom equal to how many times you’ve mulliganed.

In multiplayer (aka normal Commander), your first mulligan doesn’t count, which is why people call it the “free mulligan.” Practically, you get two looks at 7 before you start going down cards.

The Universal Keep-or-Ship Test (Works for Every Deck)

Before you fall in love with a cool spell, ask these three questions:

1) Can this hand function if I draw two blanks?

Imagine your next two draws are basic lands (or totally off-plan). Can you still:

  • cast ramp / setup spells?
  • cast your commander on time (or do something meaningful before then)?
  • not immediately fold to the table’s first scary permanent?

If the answer is “no,” you’re probably looking at a ship.

2) Do I have mana + colors, not just “lands”?

A 3-land hand is not automatically a keep if it’s the wrong colors. Color screw is the silent killer in Commander because you don’t get to “race” your way out of it.

Rule of thumb:

  • 2-color decks: you usually want access to both colors by turn 2–3.
  • 3+ colors: you usually want access to at least 2 colors immediately, and a realistic line to the rest.

3) What is this hand trying to do by Turn 3?

Your plan can be simple:

  • “Ramp into commander”
  • “Develop an engine”
  • “Hold up interaction and hit land drops”
  • “Assemble combo pieces / tutors / draw”

If you can’t say your plan in one sentence, ship it.

Pod Speed Matters (Your Keep Changes With Power Level)

This is the part most people miss: a “keep” is relative.

  • Chill pods: you can keep hands that are a half-turn slow if they snowball (value engines, bigger ramp, setup pieces).
  • High-power pods: you need hands that do something now—cheap ramp, cheap interaction, clean mana.

If you want a dead-simple way to calibrate this, tie your mulligan decisions to your MTG Commander Power Level checklist (the one where you’re tracking fast mana, tutors, free interaction, and combo density).

A Simple Decision Tree (Screenshot This)

START
│
├─ Do you have 2–4 mana sources (lands + rocks/dorks) that cast spells?
│     ├─ NO → SHIP
│     └─ YES
│
├─ Are you missing a key color you need early?
│     ├─ YES → Usually SHIP (unless you have fixing + time)
│     └─ NO
│
├─ Can you do something meaningful by Turn 2–3?
│     ├─ NO → SHIP
│     └─ YES
│
└─ Does the hand match your archetype plan?
      ├─ NO → Often SHIP
      └─ YES → KEEP

The “Good Hands” by Archetype

1) Midrange Value (Ramp + Engine + Interaction)

Midrange is the classic Commander handshake: develop mana, stick an engine, don’t die, take over.

What you’re looking for

  • 2–3 lands + a ramp piece
  • 1 payoff/engine (draw, token maker, advantage permanent)
  • 1 flexible interaction spell (optional in slower pods, important in faster pods)

A clean midrange keep often includes a 2-mana rock:

Arcane Signet
Arcane Signet
Mana Cost: 2
CMC: 2
Uncommon
Type: Artifact
Description:
T: Add one mana of any color in your commander's color identity.
Flavor Text:
The power to draw on the wisdom of your allies at any time, from any place . . . or trade cat pics!

Common midrange trap: “All gas, no ignition.” Seven spicy spells that don’t actually cast on time.

Hand Examples (Midrange Value)

KEEP — Midrange (Sultai-ish)
Command Tower, Breeding Pool, Forest,
Arcane Signet, Cultivate, Rhystic Study, Beast Within
Plan: T2 Signet → T3 Cultivate or hold up removal → engine online.
KEEP — Midrange (Boros-ish)
Sacred Foundry, Clifftop Retreat, Plains,
Talisman of Conviction, Smothering Tithe, Swords to Plowshares, Wheel of Fortune
Plan: Stable mana + ramp + interaction; T4 Tithe can take over.
SHIP — Midrange
3 taplands, 4 spells that cost 4–6 mana
Plan: “Hope nobody does anything for three turns.” (They will.)

2) Combo (Access + Speed + Protection)

Combo mulligans are the most misunderstood because the hand doesn’t need to look “fair”—it needs to be functional.

What you’re looking for

  • Mana that casts your draw/tutors early
  • Either:
    • a compact piece set + protection, or
    • strong card draw + ramp to find pieces fast

In faster pods: you also want at least one piece of protection/interaction (even if it’s just to stop someone else winning first).

Hand Examples (Combo)

KEEP — Combo (Dimir-ish)
Underground River, Watery Grave, Island,
Talisman of Dominance, Ponder, Demonic Tutor, Swan Song
Plan: Fix mana → sculpt → tutor when safe → protect the turn you go for it.
KEEP — Combo (Izzet-ish)
Spirebluff Canal, Steam Vents, Island,
Sol Ring, Brainstorm, Dockside Extortionist, Mystical Tutor
Plan: Explode mana early, tutor the key spell, keep sculpting.

After that list, here’s the one rock that basically screams “keepable hand” in a lot of decks:

Sol Ring
Sol Ring
Mana Cost: 1
CMC: 1
Uncommon
Type: Artifact
Description:
T: Add CC.
Flavor Text:
Antonio's accommodated every request, asked no questions, and stayed open late. Truly a light in the darkness.
SHIP — Combo
One-land hand with no cantrips or cheap ramp
Plan: “Draw lands.” (Cool. You won’t.)

3) Control (Land Drops + Card Flow + Early Answers)

Control hands should feel… boring. That’s good. You’re buying time, not auditioning for a highlight reel.

What you’re looking for

  • 3 lands (or 2 lands + a rock)
  • at least one early interaction piece
  • some way to avoid running out of cards (draw spell / draw engine)

Control trap: “All interaction, no development.” If you stop one thing but miss land drops, you still lose.

Hand Examples (Control)

KEEP — Control (Azorius-ish)
Hallowed Fountain, Island, Plains,
Arcane Signet, Counterspell, Swords to Plowshares, Fact or Fiction
Plan: Hit land drops, interact early, refill midgame.
KEEP — Control (Esper-ish)
Godless Shrine, Watery Grave, Command Tower,
Talisman of Dominance, Thoughtseize, Anguished Unmaking, Mystic Remora
Plan: Stable mana + early disruption + a cheap engine.
SHIP — Control
2 lands, 5 cards that cost 4+
Plan: “I promise I’m a control deck… later.”

4) Stompy / Beatdown (Curve + Pressure + One Draw Engine)

Stompy mulligans are secretly easy: you need a curve and a way to not gas out.

What you’re looking for

  • 3 lands (or 2 lands + ramp)
  • at least one early creature (or setup piece)
  • one “keep the cards flowing” effect (draw off creatures, draw on damage, etc.)

Stompy trap: keeping a hand that starts on turn 4. If you’re not applying pressure, you’re just a worse midrange deck.

Hand Examples (Stompy)

KEEP — Stompy (Gruul-ish)
Stomping Ground, Forest, Mountain,
Llanowar Elves, Rampant Growth, Beast Whisperer, Return of the Wildspeaker
Plan: Accelerate → deploy threats → refill hand when you dump it.
KEEP — Stompy (Mono-Green-ish)
Forest, Forest, Forest,
Sol Ring, Cultivate, Garruk’s Uprising, Avenger of Zendikar
Plan: Fast mana turns big threats on early; Uprising keeps you from stalling out.
SHIP — Stompy
2 lands, no ramp, all 5–7 drops
Plan: “Topdeck ramp.” (You’re not that lucky.)

The Three Most Common “Looks Fine, Loses Games” Hands

1) The “Wrong Colors” Keep

You kept 3 lands… but you can’t cast your ramp, draw, or interaction on time. You’re functionally on a mulligan already.

2) The “All Ramp, No Payoff” Keep

Ramp is great, but if your hand is just mana and nothing that converts it into advantage, you’ll ramp into… a shrug.

3) The “All Answers, No Engine” Keep

You can interact, but you can’t progress. In Commander, trading 1-for-1 forever is a trap unless you’re also drawing extra cards.

Mulligan Etiquette (Quick Rule 0 Script)

If you’re playing with strangers or a new pod, this avoids drama:

  • “We doing normal London mulligans with the free multiplayer mulligan?”
  • “Are we allowing one extra ‘friendly’ mull if someone has zero/one land?”

That’s it. Quick, clean, everyone’s on the same page.

FAQs

How many lands should I keep in Commander?

Most decks are happiest with 2–4 mana sources in the opener (lands + ramp). Three lands is the comfy default, but 2 lands + a 2-mana rock is often just as good.

Should I keep a one-land hand?

Usually no—unless you have multiple cheap cantrips (or very reliable early ramp) and your colors are perfect. One-land keeps feel heroic… right up until you do nothing for three turns.

Does Commander really have a “free mulligan”?

In multiplayer games, the rules make the first mulligan not count toward how many cards you put on the bottom—so you effectively get two looks at 7 before going down a card.

When should I mulligan aggressively?

When your pod is fast, your commander is expensive, or your deck is built around early setup (ramp/engine/tutors). Slow hands get punished harder the stronger the table is.

Should I mulligan for ramp every game?

Not blindly. Mulligan for a functional plan, not a specific card. If your hand already curves and develops, you don’t need to chase ramp.

What do I put back on the bottom after I keep?

Keep the pieces that make your hand work (mana, colors, early plays), then bottom the most situational or most expensive card(s). If you’re unsure, bottom the card you’d be least happy to topdeck on turn 6.