If you want to win more games, learn how to Mulligan in Commander without getting cute. Most “bad luck” Commander games start the same way: you kept a hand that technically does something, then you spent the next six turns playing catch-up while three other players did actual Magic.
I’m not saying you should mulligan to five every game like you’re hunting Sol Ring. I’m saying you should stop keeping hands that rely on one perfect topdeck, one unanswered mana rock, or one “surely I’ll draw a land” prayer. Those are greedy keeps. They feel brave for about 30 seconds, and then they quietly lose.
This guide is about practical keeps, when to ship it, and why greedy hands lose more Commander games than misplays ever will.
Quick table check
Before we talk hands, do this quick mental check:
What turn does my deck want to start interacting with the table?
How fast is this pod?
Does my commander fix problems, or does it need help to function?
If you don’t know the pod speed, it’s worth having a fast Rule 0 chat. Commander is way easier when you know whether you’re playing “battlecruiser until turn 8” or “someone might win on turn 4.”
How the Mulligan in Commander actually works
Commander uses the London mulligan. You draw seven, decide if you’re keeping, and if you mulligan you shuffle up and draw seven again. When you finally keep, you put a number of cards on the bottom of your library equal to how many mulligans you took.
In multiplayer Commander, your first mulligan is effectively free. You still shuffle and redraw, but you don’t have to put a card on the bottom for that first redo. After that, each mulligan costs you one card on the bottom when you keep.
Also, in multiplayer you draw on your first turn even if you’re going first. That matters because a six-card keep often turns into a seven-card hand pretty fast, while a “keepable” seven that can’t hit land drops stays miserable all game.
So yes, mulliganing is a resource. But it’s a resource you should actually use.
The keep test: mana, early play, and a plan
When you mulligan in Commander, you’re buying yourself a functional start, not hunting a perfect seven.
A keepable opening hand in Commander usually has three things:
1) Mana that casts spells on time
You want enough mana sources to play Magic, not just “a land and vibes.” For most casual pods, a solid baseline is:
2 to 4 lands, with at least two colors you need
a path to the third land drop by turn 3 (draw, ramp, or more lands)
not all tap lands, unless your deck is truly slow
If you keep hands that stumble on mana, you’re not just behind one player. You’re behind three players. That’s why mana mistakes feel so brutal in Commander.
If your mana base regularly hands you awkward openers, fix the base first. This is the boring part, but it wins games: The Essential Guide to Lands in Magic: The Gathering.
2) Something to do in turns 1 to 3
This does not have to be explosive. It just has to exist.
Early plays look like:
a two-mana ramp spell or mana rock
a cheap draw spell or setup piece
a piece of interaction you can actually cast
If your first meaningful play is turn 4 and it isn’t a board wipe, you’re basically asking the table to snowball you.
3) A plan that doesn’t require a miracle
You don’t need “the perfect hand.” You need a hand that turns into a game.
A good question is: if my next two draws are medium, do i still function? If the answer is “no, but if i topdeck exactly a Forest…” then you already know.
And remember, your commander is always available. Most decks don’t need to mulligan for a specific engine card. They need mana plus a reasonable curve so the commander actually matters when it hits.
Practical hands to keep
Here are some real-world examples. I’m describing them in categories so you can map them onto your deck, not copy a fake seven.
Keep 1: three lands, ramp, and a card that matters
3 lands with your colors
1 two-mana ramp piece
1 cheap draw or selection spell
1 interaction spell or early blocker
1 mid-game payoff
This is the boring keep that wins. It hits land drops, develops, and isn’t dead if someone removes one permanent.
Keep 2: two lands plus guaranteed third mana
2 lands
1 two-mana rock or ramp spell that you can cast
at least one castable spell on turn 2 or 3
This is fine even without a third land in hand, because you’ve got a real path to three mana. The key is “castable.” A hand with a ramp spell that needs double green when your only green source is your third land is not a path. It’s a wish.
Keep 3: two lands plus cheap card draw
If your deck runs a lot of cheap draw, cantrips, or looting, two lands can be enough.
What you want is:
2 lands that cast your early spells
a cheap draw spell you can cast on turn 1 or 2
a curve that starts at 2 or 3 mana, not 5
You’re basically saying, “I’ll spend early turns smoothing, then play a normal game.” That’s a real plan.
Keep 4: the “one land plus Sol Ring” hand with a safety net
Sometimes it’s correct to keep a one-land hand with fast mana. But only when it’s not a house of cards.
The safety net looks like:
the one land produces a needed color
you have another cheap mana source or draw spell that digs
your hand is not all four- and five-drops
If your entire hand depends on Sol Ring living, you’re not keeping a hand. You’re placing a bet. And the table loves cashing those bets.
If you’re still tuning your ramp package, it helps to know what “normal” looks like: The Best Mana Rocks in Commander Format MTG.
Keep 5: four lands with action
People hate four-land hands because it “feels flooded.” But in Commander, four lands plus action is often great, especially if your commander costs four or five.
What makes this a keep:
at least one early play
at least one piece of interaction or draw
the lands aren’t all colorless or tapped
Four lands means you get to play your commander on time. That’s not a punishment.
When to ship it
Here are the hands that look tempting and quietly throw games.
The one-land prayer
If you have one land and no cheap draw, no extra mana source, and no way to function if you miss your second land drop, ship it. Even if the spells are sweet. Especially if the spells are sweet.
Commander punishes stumbles because you can’t bully the table with tempo. You’re one person.
Two lands, wrong colors
Two lands is not “keepable” if you can’t cast your early spells. This is the classic trap in three- to five-color decks. You see two lands and a rock and think you’re fine, then realize your rock doesn’t fix the missing color and your hand is full of double-pipped spells.
If you can’t cast anything relevant until turn 4, you’re not keeping. You’re spectating.
All gas, no way to start the engine
A hand full of bombs feels great, but it usually means your deck is built to lose slowly.
If your opener is:
2 lands
5 spells that cost 5+
That’s a mulligan. You’re hoping to rip land, land, ramp, and you still might not affect the board before someone snowballs.
The tapland traffic jam
Three lands is usually fine. Three lands that all enter tapped is a different story.
If your first two turns are “land tapped, go” and “land tapped, go” you just gave the table free time. In some pods that’s survivable. In many pods, that’s a death sentence.
The fragile fast-mana keep
If your keep is built around one artifact and the table has even a basic level of interaction, you’re asking for a bad time.
Yes, artifact removal exists in Commander. It exists because of hands like this.
Why greedy keeps lose in Commander
A greedy keep is any hand that only works if the game goes exactly your way. And Commander is a format built to stop that from happening.
Here’s what greedy keeps usually rely on:
They rely on one card staying alive
Fast mana hands can be fine. Fast mana hands that collapse to a single Nature’s Claim are not fine.
When your opening line is “if nobody interacts with me, i’ll be ahead,” you’re describing a plan that every opponent at the table wants to break.
They rely on drawing the right half of the deck
In a 100-card deck, you don’t get to assume your next draw fixes you. You have to earn that assumption with draw spells, tutors, or redundancy.
If your hand needs a third land and a second color source and an early play, that’s not one topdeck. That’s three.
They treat Commander like a duel
In a duel, you can sometimes keep a sketchy hand and race. In a pod, if you stumble, you’re not racing one player. You’re letting three players take uncontested turns. That’s a different math problem.
And this is why the “greedy keep” feels worse in Commander than any other format. You don’t just fall behind. You get left behind.
Adjust your mulligans to your deck and your pod
Mulligan decisions are not universal. A few quick tweaks help a lot:
Low-curve decks can keep two lands more often, especially with draw or a rock.
Big-mana decks should mulligan harder for ramp plus lands. If your deck needs six mana to matter, you can’t keep hands that start doing things on turn 5.
Decks with lots of cheap interaction can keep slightly riskier mana if the interaction buys time.
Decks that rely on a specific commander ability should prioritize hands that cast the commander on curve and protect it.
And if your pod is higher power, you should be more honest about what “keepable” means. A slow but stable hand can be fine at casual tables. At higher power tables, a slow hand is just a polite way to lose.
Commander mulligan FAQ
Do you get a free mulligan in Commander?
In multiplayer Commander, yes, your first mulligan is free in the sense that it doesn’t cost you a card when you keep. After that, each mulligan effectively costs one card because you bottom that many.
Do you draw on turn 1 in Commander?
In multiplayer Commander, yes, even the first player draws on their first turn. That makes six-card keeps a little less scary, but it doesn’t make bad mana magically good.
Should i mulligan for Sol Ring?
Not as a default. If your deck is functional without it, keep functional hands and let Sol Ring be a nice bonus. Mulliganing purely to find fast mana is how you end up keeping fragile openers that fold to one removal spell.
How many lands should i keep?
Most of the time, two to four. Two lands is fine when you have a real plan to hit the third mana source and you can cast your early spells. Four lands is fine when you have action and your commander costs four or five.
What’s the biggest mulligan mistake in Commander?
Keeping a hand that can’t play real Magic until turn 4 because “it has good cards.” Commander is full of good cards. The game rewards the player who gets to cast them.
Final thoughts
Learning to Mulligan in Commander is basically learning to respect the first three turns. You don’t need perfection. You need mana that works, an early play, and a plan that survives normal interaction.
If you’re unsure, ask yourself one last question: if i keep this, am i making the game easier for me, or am i just hoping everyone else stumbles too?
Hope is not a strategy. Shuffle it back and take the free redraw.