What Is Stax in MTG?

What "stax" actually means

A stax deck attacks what opponents are allowed to do, rather than how much life they have. Every card on this page takes away a resource: mana, untap steps, cards, permanents, abilities, or the attack itself.

"Stax" is player slang, not a term in the Comprehensive Rules. It comes from a Vintage deck built around Smokestack, Tangle Wire and Crucible of Worlds, and over time the name detached from that specific list and became the label for the whole idea: win by shrinking the game rather than by racing inside it. You'll hear the same strategy called prison, resource denial, or "hatebears" when it's built on cheap creatures, and those are dialects of one archetype, not separate ones.

The distinction worth holding onto is stax versus control. A control deck answers threats after they appear, one at a time, with counterspells and removal. It is reactive and it spends a card per threat. A stax deck changes the rules up front with permanents, so the threats never get cast. That's why stax cards are almost always permanents and almost always symmetrical, and it's why a stax deck's opening turns look like it's doing nothing: the payoff is everything that doesn't happen afterwards. Read Control: The Commander Archetype → for the other half of that pair.

Breaking parity: the one idea that makes stax work

Almost every stax card is symmetrical. A symmetrical lock helps nobody, so the entire skill of the archetype is being the player who minds least.

Sphere of Resistance taxes you the same {1} it taxes everyone else. Smokestack makes you sacrifice too. Winter Orb stops your lands untapping as well. None of that is a drawback if your deck was built so the constraint costs you less than it costs the other three players, and there are five reliable ways to arrange that:

1. Be cheaper. A tax is a percentage, not a number. A {1} surcharge on a deck whose average card costs two mana is a 50% tax; on a deck averaging four it's 25%. Stax lists run low curves for the same reason a good mana curve → matters generally, just more so.

2. Own fewer permanents. Smokestack, Pox and every edict effect scale with what you have on the battlefield. A deck holding four permanents outlasts three decks holding twelve, so "do less" is literally the plan.

3. Use mana the lock doesn't touch. Winter Orb's text says lands, so mana rocks untap normally underneath it. A stax deck that ramps on artifacts keeps most of its mana while the table keeps one land, which is exactly why the pairing has survived for thirty years. See Mana Rocks: The Complete Guide →.

4. Untap outside the untap step. Seedborn Muse untaps your permanents during every other player's untap step; Derevi, Empyrial Tactician untaps a permanent whenever a creature deals combat damage to a player. Both make untap denial fully one-sided, and both are why those two cards read as innocuous and play as oppressive.

5. Empty your hand on purpose. Ensnaring Bridge and Necrogen Mists cost nothing to a deck that has already deployed everything it drew. The stax deck's hand is empty by design, so the cards that punish a full hand are one-sided in its favor.

The eleven types of stax

Every stax card takes away a specific resource, and which resource it is determines who it hurts, how it gets answered, and how the table reacts to it. Here's the whole taxonomy in one table, then each category in turn with the cards that define it.

No real deck picks one row. A functioning stax list combines three or four categories so the pieces cover each other's gaps: a tax effect slows the table down, an untap effect stops them catching up, and a spell restriction stops the one card that would have answered both.

The gentlest category and the most widely played. Nothing is forbidden; it just costs more mana than the table can spare.

Taxation is where most stax decks start, because a cost increase is a soft effect that scales into a hard one. One Sphere of Resistance is an inconvenience. A Sphere plus a Thalia plus a Grand Arbiter is a table where a four-mana card costs seven, and at four players nobody has seven. The category also includes the format's cheapest disruption. Thorn of Amethyst, Vryn Wingmare and Lodestone Golem all do variations of the same job.

Sphere of Resistance ↗

Every spell everyone casts costs {1} more. Two mana, perfectly symmetrical, and devastatingly one-sided in practice: a deck averaging two mana per card pays a 50% surcharge, a five-color goodstuff pile pays 20%.

Thalia, Guardian of Thraben ↗

The same {1} tax on noncreature spells only, stapled to a 2/1 first striker. Creature decks barely notice her; the artifact ramp and interaction decks around them do.

Trinisphere ↗

While it's untapped, every spell that would cost less than three mana costs three instead. It doesn't tax big spells at all. It deletes the cheap interaction and free spells that would answer the rest of your lock.

Grand Arbiter Augustin IV ↗

Opponents' spells cost {1} more, your blue and white spells cost {1} less. A two-mana swing from the command zone on turn four, available every single game.

Because taxes are symmetrical, they're the category most dependent on the curve discipline described above. A stax deck running Sphere of Resistance and a five-mana commander has taxed itself into the same corner as everyone else.

Mana isn't a pile you own, it's a thing you get back once per turn. Take away the untap step and mana becomes a one-time resource.

This is the most recognizable form of stax and the one people mean when they picture the archetype. It's also the most brutal, because it doesn't take anything away. It just stops giving it back. Everything the table taps stays tapped, so the game slows to whatever pace the denial permits, which is usually about one spell every two turns. See The Phases of a Turn → for why the untap step is the step with no priority and no way to respond.

Winter Orb ↗

Nobody untaps more than one land per untap step. The archetype's signature card, and note the exact wording: it says lands, so mana rocks untap normally. That gap is the whole plan.

Static Orb ↗

Two permanents of any kind, which closes the mana-rock loophole Winter Orb leaves open. Slower and more expensive, and it hits you almost as hard.

Stasis ↗

Nobody untaps anything, ever. The most absolute untap denial ever printed, and it sacrifices itself unless you pay {U} every upkeep, so it's a countdown you have to solve, not a permanent you cast.

Hokori, Dust Drinker ↗

Lands don't untap; each player untaps exactly one at the start of their upkeep. Winter Orb on a creature body, in the command zone, recastable after every wipe.

Tangle Wire belongs to this family too, from the other direction: instead of preventing untapping, it forces each player to tap permanents down on their own upkeep for four turns and then fades away. That expiry is the point: it buys the stax deck the turns it needs and expires before it becomes the stax deck's problem.

A turn of delay on every permanent an opponent plays, charged one card at a time and almost never contested.

These read as the mildest cards on the page and do far more than they look like they do. A land entering tapped costs a full turn of mana. A creature entering tapped can't block this turn cycle and can't be a surprise blocker at all, which turns off haste, ambush blockers and half the removal-dodging tricks in the format at the same time. In a deck that wants to attack, they're stax pieces that also enable the clock.

Thalia, Heretic Cathar ↗

Opponents' creatures and nonbasic lands enter tapped. Every land they play is a turn slower, every creature is summoning sick and can't block, and she has a 3/2 first-strike body to attack into the mess.

Kismet ↗

Artifacts, creatures, and lands your opponents control enter tapped. The widest version of the effect, and unlike a creature it doesn't die to a Swords to Plowshares.

Blind Obedience ↗

One mana for the artifact and creature half. Cheap enough to land on turn one, which is when it costs opponents the most, and extort turns spare mana into a slow clock.

Authority of the Consuls ↗

One white mana: opponents' creatures enter tapped and you gain a life each time. It reads as a minor annoyance and quietly blanks every haste creature and every surprise blocker at the table.

Root Maze, Frozen Aether and Loxodon Gatekeeper cover the same ground with different splits between artifacts, creatures and lands. Pick by what your pod's decks actually rely on. Against a heavy nonbasic manabase, the land half is worth more than the creature half.

The most effective category and the most socially fraught. Everything else on this page taxes mana; this attacks the mana base itself.

Commander manabases are greedy (four and five colors, dozens of nonbasics, a fetchland package) because the format's other conventions assume nobody is attacking them. That assumption is exactly what these cards punish. Blood Moon and Back to Basics don't destroy anything; they simply make a carefully-tuned manabase stop producing the colors it was built for. Read The Land & Mana Base Guide → for why a Commander manabase is so exposed to this in the first place.

Blood Moon ↗

All nonbasic lands are Mountains. Commander is a format built almost entirely on nonbasics, so this frequently ends a game before it starts. Three players simply stop casting spells.

Back to Basics ↗

Nonbasic lands don't untap. Softer than Blood Moon in that opponents keep their colors, harder in that they only get one use out of each land.

Armageddon ↗

Destroy all lands. The bluntest resource denial in Magic, and only a real card if you can rebuild faster than the table. Cast it with a board already ahead, or don't cast it.

Null Rod ↗

Activated abilities of artifacts can't be activated. It doesn't destroy a single mana rock, it switches all of them off, which means artifact recursion can't undo it either.

The category also includes repeatable land destruction: Strip Mine or Wasteland recurred every turn with Crucible of Worlds is a soft lock all by itself, and the artifact-mana shutoffs Null Rod, Stony Silence and Collector Ouphe. Contamination goes furthest of all, restricting every player to one black mana a turn.

This is the one category to raise before the game starts. Wizards' bracket guidelines ask players to keep mass land denial out of brackets 1 through 3 entirely. It's only unrestricted at Optimized and cEDH. That isn't a ban, it's a shared expectation, and breaking it without a heads-up is the fastest way to end a Commander night early. See The Commander Bracket System →.

Rather than pricing spells, this category caps them: one per turn, none from the graveyard, or none at a chosen mana value.

These are the hardest counters to combo in the format. A storm turn, a cascade chain, or an infinite loop that needs to cast three cards doesn't get slower under Rule of Law. It stops existing. That makes this category the reason stax and combo are natural enemies at high power, and the reason cEDH stax lists are built around it.

Rule of Law ↗

One spell per player per turn. Storm and combo decks stop functioning outright, and every value deck's big turn becomes a small one.

Archon of Emeria ↗

The same one-spell restriction on a 2/3 flier, plus opponents' nonbasic lands entering tapped. Two categories of stax on one three-mana card, with a clock attached.

Drannith Magistrate ↗

Opponents can't cast spells from anywhere except their hands. No commanders, no flashback, no cascade, no casting off the top: a two-drop that turns off the one card every opponent is guaranteed to have.

Chalice of the Void ↗

Counters every spell of one chosen mana value, all game, for both sides. Set at one it erases the format's cheap interaction; set at two it erases most of its ramp.

Arcane Laboratory, Eidolon of Rhetoric and Deafening Silence are near-copies of Rule of Law across other card types and colors. Teferi, Time Raveler restricts opponents to sorcery speed, which shuts off instant-speed interaction and flash combos. And the surgical versions (Sanctum Prelate, Gaddock Teeg, Meddling Mage, Nevermore) name one thing and forbid it, which is what you want when the table has one card that beats you.

Everything above stops opponents deploying. This category takes back what they already deployed, every single turn.

Sacrifice effects sidestep protection: no hexproof, ward, indestructible or regeneration stops "sacrifice a permanent," because the controller is the one doing it. That makes them the answer to the boards that removal can't touch, and repeatable sacrifice is the closest thing the archetype has to an inevitability engine.

Smokestack ↗

The card the archetype is named after. Each upkeep you may add a soot counter, and each player sacrifices that many permanents on their own upkeep. It escalates until only the player with the fewest permanents is left with a board.

Braids, Cabal Minion ↗

Every opponent sacrifices a permanent on their upkeep: free, one-sided, and repeating from the command zone. She is banned in Commander, and she is the clearest illustration of why a repeatable one-sided edict is a different card in a four-player game.

Pox ↗

Everyone loses a third of their life, discards a third of their hand, sacrifices a third of their creatures and a third of their lands. One card that hits four resources at once, and it rewards being the player who already had the least.

Sheoldred, Whispering One ↗

Every other player sacrifices a creature each upkeep while you reanimate one. Denial and rebuilding on the same permanent, which is the shape most successful stax cards take.

Smokestack rewards the low permanent count described earlier, and it's the clearest example of stax as a resource race: you build a board of four cheap permanents, they build a board of twelve, and then the counter goes up every turn until only one player still has anything. Pair it with a token generator and the parity breaks entirely. You sacrifice a token every upkeep and they sacrifice real cards.

Mana denial slows what opponents can cast. Card denial changes what they have to cast in the first place.

This category splits in two. Discard effects empty hands that are already full; draw restriction stops them refilling. The second half matters more in Commander than anywhere else, because the format's whole engine is card advantage. A deck denied its draw step engines is playing off the top of its library with 40 life it can't use.

Necrogen Mists ↗

Every player discards a card on their upkeep. Symmetrical, but a stax deck is trying to empty its hand onto the battlefield anyway, so by turn six it's discarding blanks and everyone else is discarding gas.

Oppression ↗

Casting a spell makes you discard a card. It taxes card advantage rather than mana, which is the resource that decks capable of playing through mana denial usually still have.

Sire of Insanity ↗

Every player discards their hand at each end step. Brutal, indiscriminate, and the reason it shows up in decks that operate from the battlefield and the graveyard instead of from hand.

Narset, Parter of Veils ↗

Opponents can't draw more than one card a turn. It doesn't stop them casting the draw spell. It just makes Rhystic Study and every wheel effect into a cantrip.

Spirit of the Labyrinth and Notion Thief cover the same draw-denial ground as Narset, and Hullbreacher did it hard enough to get banned outright. It turned every opponent's card draw into treasure for its controller. Painful Quandary sits on the boundary with the punisher category below, since it lets the spell resolve and charges for it afterwards.

A category of one-card answers. Each of these blanks a mechanic rather than a card, which is why they hit several decks at the table at once.

Commander decks are built on abilities far more than on spells: mana dorks, sacrifice outlets, equip costs, enter-the-battlefield value, tap abilities on commanders. Shutting off a whole class of ability is the highest card-for-card return in the archetype, and it's often not even read as stax when it happens. A Torpor Orb just looks like a weird artifact until you realize four decks stopped working. See ETB vs. Triggered vs. Activated Abilities → for exactly which of these cards catches which.

Cursed Totem ↗

No creature's activated abilities can be activated. Two mana that turns off mana dorks, sacrifice outlets, equip-free commanders, and a large share of the format's creature combos at once.

Linvala, Keeper of Silence ↗

The same effect, but only for opponents. Strictly better for you and strictly harder to cast, and she flies, so she also finishes games.

Torpor Orb ↗

Creatures entering the battlefield don't cause abilities to trigger. Commander is built on enter-the-battlefield value, and this one card removes it from every deck at the table.

Suppression Field ↗

Activated abilities cost {2} more unless they're mana abilities. A tax rather than a shutdown, which is what makes it playable in a deck that has activated abilities of its own.

Elesh Norn, Mother of Machines is the modern top end here. It doubles your enter-the-battlefield triggers and turns off everyone else's, so it's a lock piece and an engine on one card. Hushbringer stops both enter and die triggers on a two-mana flier, and Pithing Needle names a single problem permanent when you only need one thing switched off.

Cards in Commander come from more than the hand. This category closes the other doors.

Graveyard hate is the most-played stax effect in the format by a wide margin, and most people playing it don't think of themselves as stax players at all. Search denial is rarer and more pointed: tutors are how the rest of the table finds the answer to your lock, so a Stranglehold is often the piece that converts a strong board into an unbeatable one.

Rest in Peace ↗

Exiles all graveyards and replaces every future trip to one with exile. The single most efficient answer to recursion, reanimation, flashback, delve, and every graveyard combo in the format.

Grafdigger's Cage ↗

Nothing is cast from graveyards or libraries, and no creature card enters the battlefield from either. One colorless mana that answers reanimation and cascade-style cheating without touching your own graveyard.

Stranglehold ↗

Opponents can't search their libraries, and extra turns get skipped. Tutors are how the rest of the table finds an answer to your lock; this is how you make sure they don't.

Opposition Agent ↗

You take control of an opponent while they search, and get to keep what they find. Flash makes it a punish rather than a warning, and it turns the table's best tutors into your card advantage.

Leyline of the Void and Ashiok, Dream Render handle graveyards from other angles, Mindlock Orb stops all library searching for everyone, and Aven Mindcensor cuts a tutor down to the top four cards at instant speed, which is the version you cast in response to the tutor rather than before it. Drannith Magistrate, from the spell-restriction category above, is the widest zone-denial card in the format.

If you're building around the graveyard yourself, this category is what your deck has to survive. See Graveyard Recursion →.

The defensive subset. These don't limit what opponents can do, only who they can do it to.

Pillowfort is stax pointed inward. Ghostly Prison doesn't slow anybody's development, cost them a card, or stop them casting anything; it just makes you the least attractive player to attack. That's a meaningful distinction socially as well as mechanically: a pillowfort deck redirects the game, while a resource-denial deck stops it. Most tables that dislike stax are perfectly happy with pillowfort.

Ghostly Prison ↗

Attackers have to pay {2} each to come at you. It doesn't stop anyone doing anything except attacking you specifically, which is why it's the most socially acceptable card on this page.

Ensnaring Bridge ↗

Creatures with power greater than your hand size can't attack at all. A stax deck dumps its hand onto the battlefield by design, so this reads as symmetrical and plays as one-sided.

Crawlspace ↗

No more than two creatures can attack you each combat. A hard cap rather than a tax, which matters against go-wide decks that can simply afford Ghostly Prison.

Silent Arbiter ↗

One attacker and one blocker per combat, for everyone. It doesn't just protect you. It flattens the whole table's combat step into something that can't kill anybody.

Propaganda and Windborn Muse are the blue and white-creature versions of Ghostly Prison; Sphere of Safety scales the tax with your enchantment count, which makes it the Enchantress → deck's version. Note that Ensnaring Bridge and Silent Arbiter are genuinely symmetrical: they flatten the whole table's combat step, not just yours, which is the line where pillowfort turns back into real stax.

The weakest category, and the one with the clearest failure mode: a player who can afford the price just pays it.

Punisher effects permit everything and charge for it in life or cards. Against a slow, mana-hungry table Manabarbs is a genuine clock; against a deck with a Sol Ring and 40 life it's a tax that gets paid without comment. What makes them worth knowing is the combination: a punisher card next to a real lock stops opponents buying their way out, and the ones that charge mana rather than life (The Tabernacle at Pendrell Vale above all) hold up as locks in their own right.

Manabarbs ↗

Tapping a land for mana deals 1 damage to you. It permits every play and prices all of them, which in a 40-life format means it's a clock rather than a lock.

Painful Quandary ↗

Every opponent's spell costs them 5 life or a card. Five is a large enough number that the choice is genuinely painful, which is the whole design of a punisher card.

The Tabernacle at Pendrell Vale ↗

Every creature has an upkeep cost of {1} or it dies. A land that quietly taxes the entire board every turn, and one of the few punisher effects strong enough to function as a real lock.

Zo-Zu the Punisher ↗

Two damage to a player each time they play a land. It doesn't stop the land. It makes developing a mana base a race against your own life total.

Soft locks vs. hard locks

Once several pieces are down, the position has a name. A soft lock leaves opponents able to act, just badly. One land untapping a turn under Winter Orb is still one land. A hard lock means they cannot take any meaningful action at all and no draw will change that: Stasis with a way to pay for it forever, or Smokestack with more counters than anyone has permanents.

The distinction is worth naming because it's the difference between a slow game and a stopped one. A soft lock is a legitimate, interesting board state that opponents can still play through. A hard lock with no clock attached is the thing that turns Commander night into forty minutes of nobody doing anything, and the convention at most tables is simple: if you've established a hard lock, you should be winning within a turn or two, or you should say so and let the table decide whether to keep playing.

How stax decks actually win

Locking the table down isn't a win condition. It's the setup, and it's the part stax decks most often forget to finish.

Because the lock is the hard part, the kill is usually small: a single evasive creature swinging for two a turn, a planeswalker ticking toward an ultimate nobody can pressure, or a two-card combo assembled behind a wall of taxes that stop anyone answering it. In cEDH the last of those is standard. The stax package exists to buy the turns the combo needs. In casual play the first is more common, and it's why stax games run long.

The practical deckbuilding note: a stax deck needs a faster clock than its designer thinks. Once the lock is complete the game is decided and everything after that is bookkeeping, so the difference between a good stax deck and an exhausting one is entirely how many turns it takes to convert a won position into a finished game. A commander that attacks, or a wincon that doesn't need the board, is worth more here than one more tax piece.

Playing against stax

Stax decks lose to the same four mistakes not being made. Whatever you're piloting, these are the habits worth building:

The card-level answer is broad, cheap, and preferably instant-speed. Most stax packages are artifacts and enchantments, which is a narrow enough range that four slots covers it:

Krosan Grip ↗

Destroy an artifact or enchantment, with split second so the stax player can't respond by comboing off or protecting the piece. The single best answer to a lock that has just been assembled.

Vandalblast ↗

One mana to destroy one artifact you don't control, or {4}{R} overloaded to destroy every artifact you don't control. Many stax packages are almost entirely artifacts, and this is a one-sided sweeper for them.

Farewell ↗

Choose any combination of artifacts, creatures, enchantments and graveyards, and exile them. Expensive, but it answers a lock that spans three card types in one card, and exile beats recursion.

Boseiju, Who Endures ↗

A land that channels for {1}{G} to destroy an artifact, enchantment, or nonbasic land. It costs you no deck slot worth mentioning, which is why it belongs in every green deck.

Stax and the table

Stax is the archetype most likely to cause a genuinely bad evening, and pretending otherwise doesn't help anyone. It reduces what other people get to do in a direct, visible way, and in a casual pod it often produces long turns where three players do nothing. That doesn't make it wrong to play. It makes it something to raise before the game starts rather than reveal on turn four.

The Rule 0 conversation → is exactly the tool for this, and the honest version is one sentence: "this deck taxes and locks, and it includes mass land denial, so is that a game you want tonight?" Brackets give the same conversation a shorter vocabulary; a dedicated stax or lock package is bracket 4 territory, and mass land denial is explicitly not expected below it. One objection at the table is enough. That's the norm, and it applies here more than anywhere.

Two things make a stax deck welcome rather than tolerated: a real clock, so the game ends once it's decided, and a willingness to sleeve up something else when the pod says no. Both are easier if you know how your pod's games actually go, which is what tracking them tells you.

Frequently Asked Questions