Play Better Commander
A hand with three lands and four spells clears the mulligan bar. It's keepable. But keepable and good aren't the same thing, and a lot of games get away from you starting from a hand that was never wrong to keep, just never right for the deck it was dealt to. This article is about hand quality, not the keep-or-ship mechanics; for the London mulligan, the free first mulligan, and the actual math, see how mulligans work.
A good opening hand does something specific your deck actually wants to do, on a turn your deck can actually reach it. It's not a checklist of "enough lands and some spells," it's a match between the seven cards in your hand and the game plan printed on the rest of your deck.
The exception matters as much as the rule: a hand of nothing but lands and ramp can be an excellent keep in a slow control deck or a big-mana deck that wants to be doing its real thing on turn seven or eight. That same hand is a disaster in an aggro deck that needs to be attacking by turn two or three. There's no seven-card hand that's universally good. There's only a hand that's good for the deck you're holding it in.
For the mechanics of when to keep versus ship, the London mulligan procedure, and Commander's free first mulligan, see How Mulligans Work →. This guide picks up after that decision: you've got seven cards you're allowed to keep. Are they actually good?
Land count is the first filter, not the only one. Once a hand clears that bar, ask four more questions before you decide it's good rather than merely acceptable.
Does it interact with your commander at all? Your commander is guaranteed, sitting in the command zone from turn one. A hand that ignores it entirely, when your deck's whole plan runs through casting and protecting that commander, is weaker than it looks even with a fine mana base.
Does it contain your deck's actual early gameplan? Not "does it have spells," but does it have the kind of spell your deck leans on early: ramp for a ramp deck, interaction for a control deck, cheap threats for an aggro or go-wide deck. A hand with four spells that are all the wrong shape for your deck's opening turns isn't doing the job those four cards look like they should.
Can you actually cast this on curve, given these specific lands? Three lands that are all one color when half your hand needs a second color isn't the same as three lands that cover your curve. Check the hand's colors against the spells, not just the count of lands against the count of spells. If the ratio concept behind your deck's land-to-ramp-to-spell balance is unfamiliar, see The Commander Deck Template → for how that ratio is built in the first place.
Does it have any answer to a bad flip of events? A hand that's only good if your next three draws are all lands, or all cheap spells, is fragile. A hand with some flexibility, a card you can play whether or not you hit your next land drop, holds up better across the range of ways the game can actually start.
Four lands, a five-mana artifact, a six-mana finisher, and a card-draw spell that costs four. By land count and spell count alone, this is a comfortable keep.
In an aggro or go-wide deck, though, this hand does nothing until turn four at the earliest, and its first real impact lands on turn five or six. By then the aggro plan has already lost its reason to exist: falling behind on board while your deck's whole identity is being ahead on board.
The same seven cards in a control or ramp deck are fine, even good. Nothing here needed to happen before turn four in that deck anyway, and turn five onward is exactly when its cards start mattering.
The lesson isn't 'this hand is bad.' It's that the hand's plan and the deck's plan have to point the same direction, and here they don't, for one of the two decks.
Three lands, a two-mana mana rock, a three-mana ramp spell, and a seven-mana payoff creature.
This hand does exactly what a ramp deck wants: it accelerates on turns two and three, and by turn five or six it's casting a card that would normally take until turn eight or nine.
It would look terrible on land count alone (only three lands feels thin), but the two ramp pieces functionally act as extra land drops, so the hand's real mana development is closer to five sources by turn three.
This is the case for reading a hand's plan instead of just counting lands: a three-land hand with the right ramp can outperform a five-land hand with none.
Four lands and three specific removal spells, each one built to answer a particular kind of threat (one only hits artifacts, one only hits enchantments, one only hits creatures with flying).
This hand looks action-packed, but it's betting that this specific pod's board is going to produce exactly those threats, and soon. If nobody plays an artifact or a flier for the first five turns, two of these three cards are dead weight.
A control deck can often afford to hold narrow answers and wait, since its whole plan is surviving to a late game. An aggro or combo deck holding the same three cards has spent its whole hand on maybes while doing nothing proactive.
The fix isn't to never keep a hand like this. It's to recognize that its quality depends heavily on what the table is actually doing, not just on what's in your hand.
Three lands, two pieces of equipment, and a cheap creature to hold them, in a voltron deck built around a specific commander.
This looks like the deck's exact plan on paper. The catch: none of these three creatures is the commander, and the deck's actual payoff only works once that commander, or a specific support piece for it, shows up.
Until then, this hand is two pieces of equipment with nothing worth equipping. It's a fine hand if that draw comes on time, and a hand that's doing nothing at all if it doesn't.
This is the trap version of 'has a plan': a plan that's only real once you draw into the one card that completes it is a bet, not a plan. Weigh it as a bet, not as a sure thing.
Keeping a hand that "has lands" but no actual plan. Three or four lands with nothing that matches your deck's early gameplan clears the mulligan bar and still isn't good. Land count is a floor, not a quality score.
Judging a hand by its most exciting card. One powerful spell doesn't make a hand good if the other six cards can't get you there, or don't do anything on the way. Judge the whole shape of the hand, not its highlight.
Keeping a hand built for a different matchup than the one you're actually in. A hand full of narrow removal is only as good as the threats it happens to answer. Read the table you're actually sitting at, not a hypothetical one.
Forgetting what your commander already guarantees you. Your commander is available from the command zone regardless of your draw. A hand that ignores that, or one you're judging as if the command zone were empty, is being scored against the wrong baseline.
Treating "has enough lands and some spells" as the finish line. That's the mulligan decision, already made. Hand quality is a separate, harder question about whether those specific cards do your deck's specific job.
There's no hand quality bar that applies the same way to every deck, because "good" is defined by what your deck needs to be doing and when.
Aggro and go-wide decks want action by turn two or three. See the aggro archetype guide → for what that plan looks like in practice. If a hand's first real play is turn four or later, it's already behind the clock that deck is trying to set for everyone else.
Control and value decks can keep slower, land-heavy hands, because their job is surviving to a power turn that might not arrive until turn seven or later. See the control archetype guide →. A hand that does nothing proactive for four turns is often exactly what a control deck wants, as long as it can answer whatever the table throws at it in the meantime.
Combo decks want one of two things from a hand: a fast route to their pieces, or enough interaction to protect a slower draw toward them. See the combo archetype guide →. A combo hand with neither speed nor protection is a hand hoping nobody disrupts it, which isn't a plan against a real table.
Voltron decks want their key equipment or aura package, or at minimum a way to protect a single creature, not just any threat on board. See the voltron archetype guide →. A voltron hand full of unrelated creatures without the payoff package or the protection to keep one alive is a hand that looks busy and does very little.
A slow, patient keep is a known risk in a pod where someone else might explode out of the gate. A control hand that's perfectly correct on paper can still lose to an opponent who draws a fast, uncontested start before your deck's plan comes online. That's a real cost of playing the long game, not a reason to abandon it, just something worth weighing when you already know your pod skews fast.
It does something your deck's early gameplan actually calls for, not just "a spell you can cast."
It lines up with your commander, or at least doesn't ignore that your commander is already guaranteed.
The colors in your hand's lands actually cast the spells in your hand, not just cover the mana cost on paper.
It has some flexibility if your next few draws don't cooperate, not a plan that only works on a perfect draw.
Its first real play lands on a turn your deck can actually afford to wait until.
You're judging the hand's whole shape, not just its single best card.
Keepable is a floor set by the mulligan math: enough lands, not an obvious disaster, worth more than a fresh seven. Good is a much higher bar. A good hand does something specific your deck wants to do, on a turn your deck can actually reach. Plenty of hands clear the keepable bar and then do nothing useful for six turns. See our guide on how mulligans work for the keep-or-ship mechanics; this article is about what happens after you've decided to keep.
No, and this is a hand that new players keep more often than you'd expect, because 'three lands' sounds safe. Three lands with nothing else is not a plan, it's a coin flip on what you draw next. It's keepable in the loosest sense (you won't get mana screwed) but it isn't good until you know what those lands are going to cast.
No. A single powerful card doesn't make a hand good if the rest of the hand can't support getting to it or doing anything before it lands. Judge the hand's overall shape: what it does on turn two, what it does on turn four, and what it does if you miss a land drop. A hand that's built around one exciting card and nothing else is a common trap.
Functionally, yes, for planning purposes. Your commander is always sitting in the command zone regardless of what you drew, which is part of why hands with lands and ramp but no other action can still be good hands in the right deck. See how mulligans work for more on why this changes Commander math compared to other formats.
Yes, constantly. A hand of four lands, a mana rock, and two expensive payoffs is a strong keep in a ramp or control deck and a weak one in an aggro deck that needed to be attacking by turn three. Hand quality is always relative to what your deck is trying to do, not an absolute property of the seven cards themselves.
TurnZero HQ tracks results per deck across a season, so you can see whether your keeps are actually matching your deck's plan, instead of guessing from memory.
Create Free Account