Commander Card Draw: How Many Sources Keep a Deck From Running Out of Gas?

TLDR: Most midrange Commander decks should begin testing with roughly 10 to 14 reliable Commander card draw sources, plus any cheap cantrips or selection the strategy needs. Do not fill those slots with one kind of effect. A practical package includes three to five repeatable engines, three to five immediate refills, and two to four flexible sources that work early or recover a depleted hand. Treat impulse draw, recursion, and commander-based advantage according to how reliably they produce usable resources.

Commander makes sustained access to cards especially important. A deck contains exactly 100 cards including its commander, normally follows singleton construction outside basic lands, and is restricted by the commander's color identity. That combination limits redundancy and makes each color combination solve card flow differently. The official Commander rules cover the format's construction requirements.

The useful question is not simply, “How many cards say draw?” It is, “How many cards can keep this particular deck making meaningful plays?” A cheap cantrip, a five-mana refill, a graveyard recursion effect, and a permanent that draws repeatedly all contribute differently. Counting them as interchangeable is how a deck reaches turn seven with several nominal draw spells and nothing useful to cast.

How many Commander card draw sources should you run?

Use 10 to 14 reliable advantage sources as an opening test range for a conventional midrange deck. That is a deckbuilding heuristic, not a format rule. Start near the lower end when the commander supplies dependable advantage, the curve is low, or many spells replace themselves. Move toward 14 or more when the deck has expensive threats, regularly trades cards with multiple opponents, or cannot rely on its commander staying in play.

Role	Practical starting range	What the role accomplishes
Repeatable engines	3–5	Generate additional resources across multiple turns when their conditions are met.
Burst draw or refills	3–5	Rebuild a hand immediately after deployment, interaction, or a board wipe.
Flexible or early sources	2–4	Smooth opening turns or provide useful cards without demanding a large mana investment.
Recursion or temporary access	0–4 additional or overlapping slots	Recover specific resources or provide conditional access, but may not replace conventional draw.

Do not add every row mechanically. Cards can occupy multiple roles, and a deck with a reliable engine in the command zone needs a different package from one whose commander is expensive and provides no resources. The real goal is enough independent access that losing one permanent or having the commander removed does not shut down the deck.

Your mana also has to support the package. A hand containing two expensive refill spells is not functional card flow if the deck frequently misses its fourth land. If that sounds familiar, check the deck against a broader Commander mana base framework before adding still more draw.

What counts as card draw or card advantage?

Card draw is literal: an effect instructs a player to draw one or more cards. Card advantage is the wider resource concept. It can include drawing, obtaining additional cards in hand, accessing exiled cards, recurring cards from a graveyard, or generating useful objects that replace cards you would otherwise need to spend.

That broader definition is useful for deckbuilding, but the categories should stay visible. Temporary permission to play an exiled card is less flexible than putting it into your hand. Returning one particular permanent from the graveyard is powerful when that permanent matters, but it cannot find a land or answer that was never drawn. If you want a deeper look at the resource tradeoff, see tempo versus card advantage in MTG.

Effect type	What it does	How to count it
True draw	Adds cards from the library to the hand	Count fully when the mana cost and conditions are realistic.
Cantrip	Performs a small job and replaces itself	Count for smoothing, but normally not as a full advantage source.
Selection or filtering	Improves the quality of available cards without necessarily increasing quantity	Count as consistency, not automatically as net advantage.
Impulse draw	Temporarily permits cards exiled from the library to be played	Discount for expiry risk and available mana.
Recursion	Reuses cards that have already reached the graveyard	Count separately unless the deck fills its graveyard reliably.
Commander advantage	Provides draw, access, tokens, tutoring, selection, or recursion from the command zone	Count according to casting cost, trigger reliability, and removal exposure.

Burst draw fixes an empty hand now

Burst draw includes effects that provide several cards at once or otherwise refill a depleted hand. These are strongest after you have deployed several permanents, exchanged interaction, or rebuilt following a board wipe. They also let aggressive and low-curve decks convert fast deployment into another wave of threats.

Judge a refill by its usable output rather than its maximum output. Ask how many cards it provides in an ordinary game, what must be paid or sacrificed, and whether you can use the new cards promptly. A six-mana refill might look impressive but still consume the entire turn. A cheaper conditional effect can be better if the deck satisfies its condition naturally.

A healthy package normally includes at least a few immediate refills even when it has strong engines. Engines can be removed before producing value; a burst spell does its work immediately if it resolves.

Repeatable engines keep the middle turns moving

A repeatable engine generates cards or equivalent resources over several turns. The best engine for a deck is usually not the one with the highest theoretical ceiling. It is the one triggered by actions the deck already wants to take: casting its normal spells, attacking, creating tokens, playing lands, sacrificing permanents, or interacting with opponents.

Evaluate an engine using four questions:

Does it provide value immediately, or must it survive a full turn cycle?
How often will the deck meet its trigger condition without changing its normal play?
Does it require additional mana when it triggers?
Is it still acceptable when drawn late or after the supporting board has been removed?

An engine that triggers twice in a normal game is often better than one promising five cards under perfect conditions. Requiring the right permanent, spare mana, and a successful attack creates three separate failure points. This is a common example of how individually powerful cards become Commander deckbuilding mistakes when their surrounding structure is wrong.

Cantrips improve consistency but rarely refill a hand

Wizards uses “cantrip” as slang for a card that draws a card as a secondary effect rather than as a formal printed keyword. A cantrip generally replaces the card spent to cast it. That improves consistency because the spell can perform a narrow job without permanently reducing hand size, but one card spent for one card drawn normally produces no net card advantage.

Cantrips are particularly useful in decks that care about casting spells, filling the graveyard, triggering permanents, or finding specific setup pieces. They are less convincing when included only to inflate the draw count. If a deck has eight cantrips and two genuine refills, it may move smoothly through its first turns and still run out of gas later.

Selection follows the same principle. Drawing and discarding, discarding and then drawing, or looking at several cards and keeping one can improve card quality without increasing total resources. That is valuable, but it should not occupy every slot intended to rebuild a hand.

Impulse draw depends on mana and timing

Red commonly receives temporary exile-based access known as impulsive draw. The cards generally need to be played within a stated window rather than remaining available indefinitely. This can create genuine card advantage, but only when the deck has enough mana and legal plays to use the exiled cards before permission expires.

Pay close attention to whether an effect says “play” or “cast.” Permission to play an exiled card can include playing a land, while permission to cast applies only to spells. Neither wording automatically grants an additional land play. Normal timing restrictions and the usual limit on land plays continue to apply unless the effect explicitly changes them. Unused cards remain in exile after the permission ends.

Impulse draw therefore improves as the curve falls. Exiling two cards with one mana left is much riskier when the deck is full of five- and six-mana spells. Treat short-duration access as less than a full draw source when the list routinely lacks the mana to exploit it.

Recursion is useful, but it does a different job

Recursion can return a card to hand, permit it to be cast from the graveyard, replay lands, or repeatedly restore a permanent type. It acts like card advantage when it turns a previously spent resource into another useful play. Graveyard decks may derive a substantial part of their staying power this way.

It still should not automatically replace draw. Recursion depends on the graveyard containing an eligible target, often narrows which card types are available, and becomes weaker against graveyard disruption. It also cannot solve every empty-hand problem. A recursion spell that needs a creature in the graveyard does little when the missing resource is a land or removal spell.

Use recursion as a complementary package unless the entire deck is designed to treat its graveyard as an extension of the hand. Even then, include ways to function when that zone is disrupted.

How much should commander-provided advantage count?

A commander may draw cards directly or provide selection, temporary access, recursion, tokens, or tutoring. That availability is meaningful, but it is not the same as having an unremovable engine. Each time a commander is recast from the command zone, it costs an additional two generic mana for each previous cast from there.

Count commander-based advantage conservatively. A cheap commander that produces value on entering the battlefield is more dependable than an expensive commander that must untap, attack, and deal combat damage. As a practical adjustment, a highly reliable advantage commander may let you trim two or three independent sources from the starting package. An expensive or fragile commander should barely reduce the count at all.

Match the package to the deck's curve and plan

Card flow should be distributed across the turns when the deck needs it. Early sources find lands and setup pieces. Midgame engines convert the deck's normal actions into sustained resources. Late refills recover after the first plan has been answered.

Spellslinger decks should favor cheap cantrips and engines triggered by casting spells, while retaining a few substantial refills.
Creature and token decks want engines that reward creatures entering, attacking, or dealing damage rather than generic effects they struggle to enable.
Graveyard decks can lean harder on recursion, but still need conventional access when the graveyard is empty or disrupted.
Lands decks should prefer effects integrated with land plays, land retrieval, and additional land access.
Aggressive decks need low-cost sources and post-deployment refills; slow engines that consume an entire turn can break their pressure.
Low-color and colorless decks may have fewer unrestricted draw options, so artifacts, commander abilities, utility lands, and synergistic engines carry more weight. Wizards' color-pie guidance confirms that colors receive card access in different forms.

Test the package instead of trusting the number

Begin with your estimated 10 to 14 reliable sources, then goldfish several opening hands and record what happens. Can the deck make land drops? Does it have something productive to do after deploying its first wave? When an engine appears, can the deck trigger it immediately? After a simulated board wipe, how many cards remain available?

Test without assuming the commander survives. If the list stops functioning whenever its commander costs four extra mana, add independent draw or reduce the curve. When tuning an expensive list before committing to cards, Commander playtest cards can make repeated games and structural changes easier.

Before using any specific card, confirm its current Oracle text and rulings through Gatherer, which Wizards identifies as its official card database. Check the current Wizards banned-and-restricted list for Commander legality as well.

The practical takeaway

Start most midrange Commander decks with 10 to 14 genuine advantage sources, then add cantrips and selection for consistency rather than counting them as full refills. Mix immediate draw with repeatable engines, discount temporary access that your curve cannot reliably use, and treat recursion and commander advantage according to their failure points.

Your next step is simple: label every resource card in the deck as smoothing, burst draw, engine, impulse access, recursion, or commander-dependent advantage. If one category is doing nearly all the work, rebalance it. A deck does not stay fueled because it reached a magic number; it stays fueled because its sources work at different stages of the game and remain useful when the first plan fails.