Commander Mulligans: What Makes a Keepable Opening Hand?

A keepable Commander opening hand should do something useful during the first few turns without depending on perfect draw steps. In most decks, that means functional colored mana, at least one early play, and a route toward either more mana or more cards. Faster tables also demand relevant interaction or meaningful acceleration.

The important word is functional. Three lands do not automatically make a hand good. A mana rock does not fix every color problem. Seven individually powerful cards can still form a miserable opening hand if they cannot work together. Evaluate the hand as a short plan, not as a collection of cards you are happy to see.

The London mulligan rewards disciplined decisions

Multiplayer Commander uses the London mulligan. You draw seven cards after each mulligan, then put a number of cards on the bottom of your library equal to the mulligans taken. Multiplayer also gives you one free mulligan, so your first mulligan still leaves you with seven cards. A second mulligan means keeping six, while a third means keeping five. The free mulligan does not automatically apply to two-player Commander. The official Commander rules and Wizards’ explanation of the London mulligan cover the underlying procedure.

That free seven-card redraw should make you less willing to keep a hand that merely might work. If the first seven has bad colors, no early play, or an implausible curve, ship it. You are not losing a card yet. Later mulligans require a harder decision because a functional six is often better than a speculative seven, but repeatedly searching for a dream hand eventually costs too many resources.

Multiplayer Commander also lets every player draw during their first draw step, including the player taking the first turn. That extra card provides some help, but it is not permission to keep a hand requiring two or three consecutive lucky draws.

Start with mana you can actually use

Count usable sources, not just lands. Ask what colors the hand can produce on each of the first three turns and whether those colors cast the spells you intend to play. A Forest, a colorless utility land, and an off-color dual may technically be three lands while still failing to cast a two-mana ramp spell that requires green.

Two or three lands is a common starting point, but deck construction changes the calculation. A low-curve deck with cheap selection may comfortably keep two. A deck packed with five- and six-mana spells may need three lands plus acceleration. If questionable hands appear constantly, the problem may be the deck rather than your mulligan decisions; revisit your Commander mana base and fixing instead of expecting opening draws to repair it.

Treat one-land hands with suspicion. One land plus a cheap accelerant can look explosive, but the plan may collapse if the accelerant does not provide the needed color or if the second land never arrives. Keep such a hand only when the deck has enough cheap selection, redundant acceleration, and suitable draw steps to make the risk deliberate rather than hopeful.

Look for a real early play

Your opening hand should normally affect your development by turn two or three. That could be ramp, card selection, a setup permanent, a cheap creature central to the deck, protection, or interaction. The exact card type matters less than whether it advances your game plan on time.

A hand containing two lands and five spells costing five or more has cards you may eventually want, but no plan for reaching them. It relies on drawing lands while the rest of the table develops. Likewise, four lands and three expensive payoffs may be stable without being productive. You are unlikely to be mana-screwed, but you may spend several turns doing nothing meaningful.

Mana curve matters because cheaper spells are available earlier and can connect one turn to the next. Wizards’ guidance on building a mana curve makes the broader point that lower-cost cards support more consistent sequencing. In an opening hand, a cheap spell only counts as an early play if your lands can cast it when it matters.

Demand a path to continued resources

A good hand does not need ramp and card draw and interaction all at once. It does need a credible next stage. Three lands, a two-mana ramp spell, and a useful four-drop provide a sequence. Three lands, an inexpensive draw engine, and a defensive spell provide another.

Be careful with hands that empty quickly but cannot reload. Cheap creatures may produce an impressive first two turns, yet leave you topdecking after a board wipe. At the opposite extreme, a grip full of draw spells can fail because none can be cast before turn four. Resource generation must arrive at the right time.

Do not let the commander excuse a weak hand

Your commander is always available from the command zone, but that does not make every hand consistent. A commander can supply cards, mana, or selection only after you can cast it. If it is removed and cast again from the command zone, each previous command-zone cast adds two generic mana to its cost.

Ask whether the hand functions if the commander arrives a turn late or gets answered immediately. A hand full of auras for an expensive Voltron commander, sacrifice payoffs without an enabler, or tribal rewards without early creatures may technically support the deck’s theme while doing almost nothing independently.

Commander-dependent hands are not always wrong. A cheap commander designed to unlock the deck can justify more reliance than a six-mana engine. The warning sign is a hand whose cards have no useful floor until one specific permanent resolves and survives.

Adjust interaction to the table’s speed

At a relaxed table, three lands, setup, and a medium-speed value engine can be excellent. At an optimized table, the same hand may be far too slow if it cannot accelerate or stop an early combo or snowballing engine. Current Commander Brackets are a pregame-conversation tool rather than a rigid matchmaking system, but their increasing speed expectations reinforce the practical point: a keep is contextual.

Do not overcorrect by keeping seven reactive cards. A three-land hand with several narrow answers and no draw, ramp, or proactive development may stop one opponent once, then fall behind everyone else. Interaction is part of a plan, not a substitute for one. Prioritize answers that are cheap enough to cast while developing and broad enough to hit likely threats.

Opening hands that look better than they play

Opening hand pattern	Why it is dangerous	Likely decision
Three lands, but none cast the early green ramp spell	The land count is acceptable while the colors break the intended sequence	Mulligan unless another early line is available
Two lands and five expensive spells	The hand needs several favorable draw steps before it participates	Mulligan in most decks
Four lands, colorless rocks, and colored spells	Generic acceleration may not provide the colors the hand needs	Keep only if the mana sequence genuinely works
One land, one accelerant, and expensive payoffs	A missed land drop or color mismatch strands most of the hand	Usually mulligan
Three lands and several narrow answers	The hand can trade once but has no development or card flow	Keep only when those answers match a fast table
Support cards for an expensive commander	The hand may do nothing if the commander is delayed or removed	Usually mulligan without an independent line

Use a four-question keep test

Before keeping, run through four questions:

Can this hand produce the colors its early spells require?
What am I doing during the first three turns?
How does the hand reach more mana, more cards, or its central engine?
Does this plan match the speed and likely threats of the table?

The goal is not to find seven perfect cards. It is to keep a hand with functional mana, purposeful sequencing, and enough resilience to survive an ordinary disruption or missed draw. Commander gives you a free first mulligan in multiplayer. Use it to start with a plan rather than a prayer.