Dimir Assassins · face-down theft

DeadlyEtrata

A blue-black Assassin deck that steals the table's cards. Every Assassin that connects cloaks the top card of that player's library under your control. Their creatures fight for you, and Etrata turns their other cards face up or casts them for free. This is the full manual: a guide in steps, every card with rulings, the combos, the numbers, and a real game to practise against.

Bracket 3
35 lands
3 lose-the-game lines
2 two-card combos

Start the guide Play a game

100: cards, 35 lands, 21 of them basics
2.3: average mana value of the nonland cards
13: Assassins and changelings in the 99 to trigger Etrata
6–6.5: out of 10 on the old power scale

How the deck wins

Hit, cloak, flip, repeat

1
An Assassin connects
Cheap evasive Assassins do the work: Changeling Outcast, Hookblade Veteran, Brotherhood Spy and Basim Ibn Ishaq.
2
Etrata cloaks
Each hit puts the top card of that player's library onto the battlefield under your control, face down as a 2/2 with ward {2}.
3
The cloaks join in
Roshan, Hidden Magister, Maskwood Nexus and Arcane Adaptation make them Assassins, so next turn they hit and cloak too.
4
Someone loses
Etrata, the Silencer, Strixhaven Stadium or a combo knocks a player out, and Ramses, Assassin Lord turns that into a win. See the combos.

The guide

Learn it track by track

Twenty-one short chapters, from the deck in 60 seconds to the trickiest face-down rules and the Bracket 4 aggro upgrade. Your place is saved on this device.

The guide · 17 tracks

Step by step

Tap a track to read it. Each one ends with the next.

Combos & kills

How to actually win

Most games are won by stealing the table's best cards and attacking with more creatures than they can block. When the board stalls, the deck has two two-card combos and two cards that say "loses the game". Ramses, Assassin Lord turns any of those eliminations into a win for the whole table, as long as that player was attacked by one of your Assassins this turn. Tell your table about the combos before you play.

Combo 1Mill and drainDuskmantle Guildmage + Mindcrank

You need: Both on the battlefield, {1}{U}{B} for Guildmage's first ability, and a starter: any life loss or any card going to an opponent's graveyard.
Total mana: 7 across two or more turns: Guildmage 2, Mindcrank 2, the ability 3.
Speed: Instant. Activate it at the end of an opponent's turn, or during combat.

Pay {1}{U}{B}: for the rest of the turn, whenever a card is put into an opponent's graveyard from anywhere, that player loses 1 life.
Start it: an Assassin hits them, or Guildmage's second ability ({2}{U}{B}) mills two cards.
They lose life. Mindcrank mills that many cards.
Each milled card makes them lose 1 life, which mills again. It stops when they're dead or out of cards.

Does it hit every opponent?

Guildmage's ability covers every opponent, but each chain needs its own starter. A combat hit on each player starts all of them at once.

What stops it?

Removing Guildmage before its ability resolves, or Mindcrank at any point. Once the ability has resolved, killing Guildmage doesn't stop it for that turn. Dispel and Counterspell protect the setup.

Combo 2Half, then allUnstoppable Slasher + Wound Reflection

You need: Wound Reflection on the battlefield before the end step, and Slasher dealing combat damage to the player.
Speed: One combat and one end step.

Unstoppable Slasher hits: the player loses its damage, then half their remaining life, rounded up.
At the beginning of the end step, Wound Reflection makes them lose the life they lost this turn again. That's always more than they have left.

Alt winsLoses the gameEtrata, the Silencer and Strixhaven Stadium

Etrata, the Silencer can't be blocked. Each time she deals combat damage to a player, she exiles one of their creatures with a hit counter, and that player loses once they own three exiled cards with hit counters. Then she shuffles into your library, so March of Swirling Mist phasing her out in response keeps her around. Strixhaven Stadium gains a point counter each time one of your creatures hits an opponent; the hit that brings it to ten makes that player lose.

Pair them with Ramses

With Ramses, Assassin Lord on the battlefield, a player who loses after being attacked by one of your Assassins this turn wins you the game.

The engine

Cloak calculator

How many cards you steal this combat, and how fast the face-down army grows when the cloaks are Assassins too.

Game plan

How to pilot it

You're the tempo deck at the table: cheap threats, counterspells held up, and a commander that gets better every turn you connect. Stay small and hard to read, take the table's cards, and let the opponents spend their removal on each other.

Turns 1–3The setup

A one- or two-drop Assassin: Changeling Outcast, Hookblade Veteran, Brotherhood Spy.
A mana rock if you have no Assassin: Sol Ring, Arcane Signet, Dimir Signet.
Etrata on turn 3, before combat, with an Assassin ready to hit.

Turns 4–6The heist

Attack every turn with evasive Assassins at the player with the fewest blockers.
Add a type enabler: Roshan, Hidden Magister, Maskwood Nexus, Arcane Adaptation.
Keep {U} or {U}{U} open for Dispel or Counterspell when the board is worth protecting.

Turn 7+The kill

Flip the stolen bombs with Etrata's {2}{U}{B}.
Assemble a combo, or set up Etrata, the Silencer or the Stadium, with Ramses on the battlefield.
Toxic Deluge resets a table that got ahead of you.

Practice

Opening hand trainer

Draw seven from the real 99 and see whether it's a keep. Your first mulligan in multiplayer Commander is free.

Keep

3 or 4 lands with both colors, or a rock that fixes them.
2 lands plus Sol Ring or a cheap rock, and a cheap Assassin.
Any hand that casts Etrata on turn 3 with an Assassin to follow.

Mulligan

0 or 1 land.
6 or 7 lands.
Only counterspells and removal, with no threats.

Example

A good draw, turn by turn

A sample goldfish with no interaction from opponents, to show how the heist snowballs.

T1
Swamp, Changeling Outcast
An unblockable Assassin
T2
Island, Dimir Signet. Outcast hits for 1
4 mana next turn
T3
Land, Etrata, Deadly Fugitive before combat. Outcast hits
First cloak
T4
Land, Roshan, Hidden Magister. Outcast and the cloak attack; the cloak has menace
Two more cloaks
T5
Four Assassins attack. Flip the best stolen card with Etrata
Seven cards taken from the table

Sequencing

Ten habits that win games

Cast Etrata before combat. Her trigger only counts hits while she's on the battlefield.
Attack every turn. Each Assassin that connects is a free card.
Look at your cloaks. Know what you have before you plan the turn.
Flip creatures for their mana cost. It's cheaper than Etrata's ability and nobody can respond.
Flip Roshan after blockers. Menace and Assassin types arrive when blocks are locked in.
Counter the wipe. Save a counterspell for the spell that clears your board.
Keep a blocker home. Etrata's deathtouch keeps you alive while the cloaks work.
Cast Spark Double, never flip it. It copies nothing when turned face up.
Line up Ramses. Knock a player out on a turn your Assassin attacked them.
Tell the table about the combos. Guildmage and Mindcrank, Slasher and Wound Reflection.

Matchups

Reading the table

Against board wipes

Hold a counterspell or March of Swirling Mist once your board is worth protecting. Keep one or two cheap Assassins in hand to rebuild.

Against combo decks

You're the deck with counterspells. Keep Counterspell, Arcane Denial or An Offer You Can't Refuse for their key spell, and pressure them first.

Against big creature decks

Etrata's deathtouch and your cloaks block well. Cloak their library: big creature decks give you big creatures. Reality Shift and Infernal Grasp handle the one that matters.

Against graveyard decks

Bojuka Bog exiles a graveyard. Cloaking the top of their library keeps those cards out of the graveyard too.

The setlist

All 100 cards

Tap any card for its full entry.

Glossary

Every keyword in this deck

What each term means, and which of your cards use it.

Filter the glossary

FAQ

Questions people ask about this deck

Setting up the table…

Opening the Arena…

Piano roll

The whole deck as a score

Every nonland card is a note: its lane is its role, its position is its mana value. Tap a note to open the card.

Mana curve

The levels

Nonland cards by mana value. X costs count as 0 plus their fixed symbols. Tap a bar to list its cards.

64 nonland cardspeak at 2 mana

Card types100 cards

Colors

Pips against sources

How much each color is asked for in mana costs, and how many of your lands make it.

Roles

What the 99 do

A card can fill more than one role, so the bars add up to more than 99.

Draw odds

What are the chances?

Exact odds from the real 99: how likely you are to see a card, or a group of cards, by a given turn.

Opening hands

Lands in your first seven

35 lands in 99 cards. How often each count shows up in a fresh seven.

Strength

A fast, interactive Bracket 3

That's "Upgraded" in the official brackets, roughly 6 to 6.5 on the old 1 to 10 scale. Strong at a casual or upgraded-precon table, nowhere near cEDH. This is a judgement from reading the list, not from tested games.

Why it's strong

A 3-mana commander that starts stealing cards on turn 4.
Four counterspells, spot removal and Toxic Deluge.
Two two-card combos and two cards that make a player lose, all turned into a table win by Ramses.

What holds it back

Small creatures: one board wipe undoes several turns of work.
No ramp beyond mana rocks, and no tutors.
Stolen cards are random: some turns you cloak a land.
The Bracket 4 aggro upgrade fixes all three.

Before you sit down

Bracket 3 allows two-card combos as long as they aren't early-game ones. Mention Duskmantle Guildmage + Mindcrank and Ramses at the pregame talk, since some tables ban them.

Your other deck

Etrata vs Miku

They're close, and both are Bracket 3. The upgraded Miku deck is slightly ahead at about 6.5 to 7, with Etrata at about 6 to 6.5. Miku hits harder once it's set up. Etrata is faster, more interactive, and would give Miku a hard time head to head. This comes from reading both lists, not from playtesting.

1 v 1

Etrata is favoured

It can counter or kill Trostani and the overrun cards.

4-player pod

Miku is favoured

Its overruns can knock out several players in one swing, while Etrata hits one player at a time.

Your record

Your games against the bots

Saved on this device when you finish a game in the Play tab.

The upgrade · Bracket 4 aggro

From sneaky B3 to a Bracket 4 Assassin rush

Today's deck is a strong Bracket 3: it steals cards, counters things and kills with two combos. What it doesn't do is hit hard. The upgrade turns it into an aggro deck: 30 cheap creatures, 28 of them Assassins, two Assassin lords, a card drawn for almost every hit, and at the top end the fast mana, tutors and free counterspells of a Bracket 4 deck. Etrata's cloaks become the bonus, not the plan. You buy it in three stages, and each stage is a legal 100-card deck you can play in the Play tab.

50 swaps in 3 stages
30 creatures
10 Game Changers
~1,090€ for all three

Why change it

What holds today's deck back

From the deck list and from bot games in the Play tab's engine (120 to 800 games per number).

It barely hits

In bot games the deck deals about 16 combat damage to players per game, a third of what a retail precon deals. Most of its creatures are 1/1s and 1/3s, and only 13 of them trigger Etrata.

The engine dies with Etrata

Cloaks only turn into Assassins with one of four enablers, and flipping them needs Etrata alive. Remove her twice and the deck is a pile of 2/2s with no card draw.

Nothing finds the key cards

No tutors, and the card draw is slow: Mask of Memory, Reconnaissance Mission, They Came from the Pipes. Each combo piece shows up in only about one game in five.

Dead slots

March of Swirling Mist, Supernatural Stamina, Dispel, Aqueous Form and Training Grounds rarely do anything, and Wound Reflection costs six for half a combo.

The plan

Three stages, each one playable

0nowowned
Today's list: cloaks, counterspells, Slasher + Wound Reflection and Guildmage + Mindcrank.
Bracket 3
1~53€
The aggro core. 29 cheap swaps: Mari, the Killing Quill, Black Widow, Deadly Hunter, Achilles Davenport, Ezio, Blade of Vengeance and 12 more Assassins, Skullclamp, Kindred Discovery, Rooftop Bypass, cheap removal and both haste equipment.
Fast Bracket 3
2~297€
Interaction and mana: Rhystic Study, Fierce Guardianship, Cyclonic Rift, Swan Song, Mystic Remora, Deadly Rollick, Bitterblossom, Watery Grave, two fetch lands and Cavern of Souls.
Top of Bracket 3 (3 Game Changers)
3~742€
Fast mana (Mana Vault, Chrome Mox, Lotus Petal, Ancient Tomb), Demonic Tutor, Vampiric Tutor, Imperial Seal, Force of Will and two more fetch lands.
Bracket 4 (10 Game Changers)

Prices are MTGGoldfish USD × 0.85, looked up on 29 and 30 Sep 2026, cheapest regular printing. Cardmarket couldn't be read from here, so check each card there before buying: it's usually the same or cheaper. MTGGoldfish shows no regular price for Ezio, Blade of Vengeance (it was only printed in the Assassin's Creed starter kit), so it isn't in the totals. The stage 3 budget swaps on the Buy it page bring the whole plan to about 660€.

Buy it stage by stage See the final list

How the aggro deck wins

Attack, draw, attack again

1
Turns 1 to 3: cheap Assassins
Twelve creatures cost 1 or 2. Hired Poisoner, Changeling Outcast, Thrill-Kill Assassin and Guildsworn Prowler attack from turn 2, and deathtouch makes every block a trade.
2
Every hit draws
Black Widow, Deadly Hunter draws for each deathtouch creature that connects, Mari, the Killing Quill gives all your Assassins deathtouch, and Ezio, Blade of Vengeance, Kindred Discovery, Gix, Yawgmoth Praetor and Skullclamp stack on top.
3
Lords and evasion
Ramses, Assassin Lord and Achilles Davenport give +1/+1 each, Interceptor, Shadow's Hound gives every Assassin menace, and Rooftop Bypass adds a menace Assassin for each player you hit.
4
The kill
Unstoppable Slasher and Virtus the Veiled halve a life total per hit. Mari exiles every creature that dies with a hit counter, so Etrata, the Silencer often needs a single hit to make a player lose, and Ramses turns that into a win. Overloaded Cyclonic Rift clears the way for the last swing.

The Mari and Silencer trick

Etrata, the Silencer makes a player lose once they own three exiled cards with hit counters. Mari puts a hit counter on every opposing creature that dies, so a few trades in combat get a player to two. From then on, don't cash in that player's hit counters for cards: the Silencer's next hit exiles the third and that player is out. With Ramses on the battlefield and an Assassin attacking them that turn, you win the game.

Before and after

The deck in numbers

Counted from the three lists.

Bracket check

Why stage 3 is Bracket 4

Bracket 3 allows up to three Game Changers and no early two-card combos. Stage 2 stops at exactly three (Rhystic Study, Fierce Guardianship, Cyclonic Rift). Stage 3 adds seven more, and with three tutors Duskmantle Guildmage + Mindcrank becomes a combo you can find on turn 4 or 5: both put the deck in Bracket 4. It's not cEDH: no Thassa's Oracle line, no Mox Diamond, no mass land denial.

Game Changer	Stage	What it does here
Rhystic Study	2	A card for most spells the table casts
Fierce Guardianship	2	Free counter for the wipe aimed at your board
Cyclonic Rift	2	Overloaded, the last attack meets empty boards
Mana Vault	3	Turn 2 Etrata or Mari
Chrome Mox	3	A free extra mana from turn 1
Ancient Tomb	3	A land that taps for two
Demonic Tutor	3	Mari, Black Widow, the Silencer or a combo piece
Vampiric Tutor	3	The same, at instant speed, on top
Imperial Seal	3	Turn 1 tutor
Force of Will	3	Free counter for anything

In the game

How the bots do with it

All three lists are in the Play tab: pick Etrata aggro (stage 1) or Etrata B4 aggro in the deck picker. The full list also sits at the table as a Bracket 4 bot. These are bot-vs-bot results from the game's engine, not human games: 800 games per row, Etrata's seat against three bots, random turn order.

Deck (bot-piloted)	vs precons	vs Bracket 4	Combat damage to players
Today's list	7.5%	8.4%	16 a game
Stage 1	11.0%	9.5%	28 a game
Stage 2	11.2%	not run	not run
Stage 3 (B4)	6.8%	3.6%	22 a game

A fair seat in a four-player game wins 25%. The bots pilot every Etrata list badly, and stage 3 worst: in tests, swapping its fast mana and tutors back for lands and Signets brought it back up, so the bots lose value on the cards a human uses best (a Vampiric Tutor to the top, a Chrome Mox imprint, when to spend Force of Will). Read the stage 3 row as a limit of the bots, not of the list. The stage 1 row is the useful one: the same bots with more Assassins deal two thirds more combat damage and win half again as often.

Shopping list

Cut this, add that

Every swap, stage by stage: the struck-out card leaves, the new one comes in. Tap a card for its wiki entry, or the price to search Cardmarket. Tick what you've bought; your ticks stay on this device.

Budget

Cheaper stage 3

Stage 3 is 742€, and four cards are 461€ of it. These swaps keep the deck in Bracket 4 (tutors and free counters are what matter) for about 430€ less.

Instead of	Play	Saves
Imperial Seal (~155€)	Diabolic Intent (~13€): sacrifice a creature, tutor to hand. Skullclamp food works.	~142€
Chrome Mox (~139€)	Keep Mind Stone	~139€
Ancient Tomb (~113€)	Keep a basic land	~113€
Force of Will (~54€)	Pact of Negation (~16€): free now, pay {3}{U}{U} next upkeep	~38€

With all four swaps, stage 3 is about 310€ and the whole plan about 660€. The deck then has 6 Game Changers: still Bracket 4, which starts at four.

The finished deck

Etrata B4 aggro, all 100

32 lands, 30 creatures, 8 mana rocks and rituals, 9 card draw cards, 3 tutors, 11 counterspells and removal spells, and 6 other artifacts and enchantments. Tap a card to read about it.

Design notes

What stays, what goes, what didn't make it

Kept from today's deck

Every evasive one- and two-drop. Changeling Outcast, Hookblade Veteran, Mothdust Changeling, Universal Automaton, Brotherhood Spy, Desmond Miles and Basim Ibn Ishaq are exactly what an aggro deck wants.
The finishers. Ramses, Assassin Lord, Etrata, the Silencer and Unstoppable Slasher get better with Mari and Virtus.
One cloak engine. Roshan, Hidden Magister and Maskwood Nexus stay, so cloaks still become Assassins. Spark Double copies Etrata, Mari or Black Widow.
Guildmage + Mindcrank. Four mana for a second win condition the tutors can find.

Cut

Slow value. They Came from the Pipes, Reconnaissance Mission, Strixhaven Stadium, Scroll of Fate and Glitch Interpreter cost a turn without adding pressure.
Morph tricks. Kheru Spellsnatcher and Willbender need six and five mana of setup; free counterspells do the job.
Narrow spells. Dispel, March of Swirling Mist, Supernatural Stamina and Aqueous Form.
Wound Reflection. Six mana for half a combo. Virtus the Veiled is the same halving on a 3-mana creature.

Considered	Why it's not in
Orcish Bowmasters	A great Bracket 4 card, but it's an Orc, not an Assassin, and costs ~41€. First pick if you want an eleventh Game Changer.
Mox Amber	~80€ for a mox that only works with a legendary creature out. Fellwar Stone stays.
The One Ring, Necropotence	Strong draw, but both slow an aggro deck's turns. Black Widow and Ezio draw more here.
Opposition Agent, Notion Thief	Game Changers that don't attack well.
Jacob Frye, Brotherhood Ambushers	Good Assassins, just below the 30 in the list. Swap them in for Mistwalker or Lydia Frye if you like.
Toxic Deluge	A great wipe, but this deck is the board. It leaves at stage 2 for Cyclonic Rift.

DeadlyEtrata

Hit, cloak, flip, repeat

Learn it track by track

Make it a Bracket 4 aggro deck

Step by step

How to actually win

Pair them with Ramses

Cloak calculator

How to pilot it

Opening hand trainer

Keep

Mulligan

A good draw, turn by turn

Ten habits that win games

Reading the table

All 100 cards

Every keyword in this deck

Questions people ask about this deck

The whole deck as a score

The levels

Pips against sources

What the 99 do

What are the chances?

Lands in your first seven

A fast, interactive Bracket 3

Why it's strong

What holds it back

Before you sit down

Etrata vs Miku

Etrata is favoured

Miku is favoured

Your games against the bots

From sneaky B3 to a Bracket 4 Assassin rush

What holds today's deck back

It barely hits

The engine dies with Etrata

Nothing finds the key cards

Dead slots

Three stages, each one playable

Attack, draw, attack again

The Mari and Silencer trick

The deck in numbers

Why stage 3 is Bracket 4

How the bots do with it

Cut this, add that

Cheaper stage 3

Etrata B4 aggro, all 100

What stays, what goes, what didn't make it

Kept from today's deck

Cut