Deep Clue Sea precon upgrade guide

Every upkeep, Morska makes you a Clue. Every Clue costs two mana to read. And fourteen of this box's lands make you wait a turn before you can pay one.

Morska, Undersea Sleuth investigates at the beginning of each of your upkeeps, so there is a Clue sitting on the table every turn you take. Turning it into a card costs {2} and a sacrifice. That card is usually the second one you draw that turn, and the second draw is the thing this deck is actually built around: six cards in the box, the commander included, trigger on it.

So the running cost of this commander is two mana a turn, forever, on top of whatever you meant to cast. The box does not budget for it. Fourteen of the thirty-seven lands it ships enter tapped, which means that on more than a third of your land drops you spend the turn a mana behind the plan you just described. Almost everything below follows from that one collision.

What are the best Deep Clue Sea upgrades?

Mana first, answers second. That is the entire priority order here, and the prices are Scryfall market prices for default printings, checked 10 September 2026.

1. Eight untapped lands, $5.71 for the set. Adarkar Wastes, Yavimaya Coast, Brushland, Glacial Fortress, Hinterland Harbor, Sunpetal Grove, Seachrome Coast and Botanical Sanctum. Nothing else here changes as many turns for as little money.
2. Cyclonic Rift, $39.37. The box ships exactly one card that answers a permanent an opponent has already resolved. This one answers all of them, at instant speed, on the turn of your choosing.
3. Heroic Intervention, $15.96. The deck spends four or five turns assembling a board of tokens and counters, and one card in the box protects it.
4. Get Lost, $4.07. Two mana, instant speed, and it covers the creature, the planeswalker and the enchantment.
5. Beast Within and Generous Gift, $1.32 for the pair. Neither is elegant. Both answer any permanent at all for three mana.
6. March of Otherworldly Light, $0.92. Scales from one mana upward, and it exiles, which matters against the recursive commanders this deck is slow against.

There is a seventh upgrade that costs nothing at all, and it is the one I would make first: take Reliquary Tower out. That needs its own section.

Six cards are watching your second draw

Morska's third line puts two +1/+1 counters on her whenever you draw your second card in a turn. Five more cards read the same trigger:

• Detective of the Month makes a 2/2 Detective.
• Alandra, Sky Dreamer makes a 2/2 flying Drake.
• Ethereal Investigator makes a 1/1 flying Spirit.
• Jolrael, Mwonvuli Recluse makes a 2/2 Cat.
• On the Trail puts an extra land onto the battlefield tapped.

Twenty-one of the hundred cards in the box have the word Clue or investigate printed on them, so Clues are never the bottleneck. The mana to read them is. A Clue is {2} and a sacrifice every single time, and none of those five payoffs cares that you own a Clue. They care that you drew.

One card in the box fixes that outright. Tezzeret, Betrayer of Flesh has a static ability making the first activated ability of an artifact you activate each turn cost {2} less, which makes the turn's first Clue free. One card out of a hundred, and he is a four-drop.

The trigger says each turn, not your turn

Read Morska's third line again. Whenever you draw your second card each turn. Not each of your turns. If you can draw two cards during somebody else's turn, Morska gets her counters, Alandra gets her Drake and Jolrael gets a Cat, all on a turn you had nothing else to do. In a four-player pod that is four chances a rotation instead of one, and Clues are the ideal fuel for it, because cracking one is an activated ability you can use at instant speed.

Two cards in the box already understood that. Wilderness Reclamation untaps all your lands at the beginning of your end step, which hands you a full grip of mana across three opponents' turns. Innocuous Researcher does the same thing stapled to a 3/3, and its clause about not casting spells until your next turn is not the drawback it looks like: cracking a Clue is an activated ability, not a spell.

That is the case for keeping mana open, written out. It is also, word for word, the case against a land that enters tapped.

One land in the box repeats the commander's first line

Morska's first line is "you have no maximum hand size". Reliquary Tower's first line is "you have no maximum hand size". Your commander is available in every game you play, so for the overwhelming majority of turns the Tower is a land that taps for one colourless mana, in a three-colour deck that wants {U}{U} on turn three. Cutting it costs nothing, and it is the clearest single improvement in the box.

Temple of the False God fails from the other end. It makes no mana at all until you control five lands, and the turns this deck most wants a spare {2} are the ones before that.

Path of Ancestry is the third colourless offender. It enters tapped, and the scry only happens if the creature you just cast shares a creature type with Morska. Morska is a Vedalken Fish Detective, and six of the deck's thirty-two creatures are Detectives, Morska herself among them. No creature type here reaches half the board, the most common one is Human at seven, and there is no typal plan to build toward. Five casts out of thirty-one, then, and the rest of the time it is a Command Tower that costs you a turn.

The six taplands that are not buying anything

I am not against lands that enter tapped. I am against lands that enter tapped and give nothing back for it. Of the fourteen in this box, eight are paying for something real and six are not.

The eight that stay, and what each is buying:

• Seaside Citadel taps for all three of the deck's colours.
• Azorius Chancery, Selesnya Sanctuary and Simic Growth Chamber each make two mana, which is exactly the price of a Clue.
• Temple of Enlightenment and Temple of Mystery scry, and both of them make blue.
• Irrigated Farmland and Scattered Groves carry basic land types, which turns out to matter a great deal once the new lands arrive.

That last point is why the check lands are the right purchase here. Glacial Fortress, Hinterland Harbor and Sunpetal Grove each enter untapped if you already control a land with the matching basic type, and this deck is full of them: fourteen basics plus Canopy Vista, Prairie Stream, Irrigated Farmland and Scattered Groves. In the upgraded list Glacial Fortress sees thirteen of the other thirty-six lands, Hinterland Harbor sees fifteen and Sunpetal Grove sees twelve. They are untapped from your second land drop onward in almost every game.

The six that leave:

• Path of Ancestry, argued above.
• Krosan Verge enters tapped, then asks for {2} and a sacrifice to fetch two more lands, also tapped. That is two turns of tempo to fix a mana base Hinterland Harbor fixes for thirty-one cents.
• Lonely Sandbar, Secluded Steppe and Tranquil Thicket are the three cycling lands. Their whole pitch is that late in the game you can pay two mana to draw a card instead of playing them. The box already counts twenty-five cards as card draw, against a normal range of eight to twelve. This is the last deck on earth that needs a land to draw it a card.
• Temple of Plenty is the one of the three Temples that makes no blue, and blue is the colour this deck is shortest on by a distance.

A goad payoff with one goad source, and a wipe that kills my own board

Six more cards come out, each for a reason specific to this list rather than for being a cheap common.

• Serene Sleuth investigates when it enters, then offers to investigate for each goaded creature you control at the beginning of combat. The only card in the deck that can goad anything is Ransom Note, and goading is one of three modes you pick once, when you crack it. The second ability is a blank.
• Merchant of Truth costs {2}{W}{W} and investigates when a nontoken creature you control dies. The creatures this deck fields in numbers are Spirits, Drakes, Cats and Detectives, all of them tokens. It was also the earliest double-white card in the box, which is why taking it out pushes the deck's white requirement back a full turn.
• Knowledge Is Power gives your creatures +X/+X where X is the cards you have drawn this turn. On a normal turn here that is two or three, so it is a five-mana anthem for +2/+2.
• Magnifying Glass is three mana for a rock that taps for colourless, in a deck already eight sources short on {U}{U}, and its investigate mode wants {4} and a tap. It gets counted as ramp and as card draw and it is barely either one.
• Killer Service makes Food, then asks for {2} and a sacrificed token at your end step. It wants the same two mana at the same point in the turn as the Clue does, and it eats a token that six other cards are busy counting.
• Organic Extinction destroys all nonartifact creatures for ten mana, discounted by tapping the artifacts this deck is trying to keep. The board it destroys is mostly mine, Morska included. Farewell and Fumigate are enough sweepers for a deck that wants to be the one going wide.

Six cards are counting the Clues Morska wants me to spend

Kappa Cannoneer, Mechanized Production, Shimmer Dragon, Tezzeret, Betrayer of Flesh, Thought Monitor and Whirler Rogue all care how many artifacts you control. The commander's engine asks you to sacrifice one every turn. That tension is the most interesting decision the deck offers, and it is why three cards that look like easy cuts are staying.

Mechanized Production enchants an artifact, not a creature, and it is a real win condition here rather than a cute one. Every Clue token is named Clue, so eight of them on the battlefield ends the game on the spot. Academy Manufactor turns each investigate into a Clue, a Food and a Treasure at once, and Adrix and Nev, Twincasters doubles whatever is being made. This deck can genuinely get there.

Nettlecyst arrives with its own body through living weapon, so it is not an Equipment waiting for a creature to protect. It counts every artifact and enchantment you control, which here is a growing pile.

Junk Winder is printed at {5}{U}{U} and almost never costs it. Affinity for tokens reads your Clues, your Food, your Treasures and every 2/2 you have made, so on the turns that matter it is a three or four mana 5/5 that taps down a blocker every time another token shows up.

One answer to a resolved threat is not enough

Swords to Plowshares is the entire removal suite. One card, against a normal range of five to ten for a Commander deck. Everything else the box offers as interaction is a sweeper, and a deck that wins by having the widest board is not thrilled about casting sweepers.

Five instants fix it: Cyclonic Rift, Get Lost, Generous Gift, Beast Within and March of Otherworldly Light. Four of those five cost under a dollar and a half each. Heroic Intervention joins them because the box protects its board with exactly one card, Koma, Cosmos Serpent, and this deck spends four or five turns building something a single Farewell erases.

Cyclonic Rift is $39.37, more than half of what this whole list costs. I put it in anyway. The deck's problem is not that its one answer is weak, it is that it has one, and Rift is the card that turns a losing board back into a winning one at instant speed. If you would rather not spend that, Aura Shards at $28.33 is the substitute I would look at first: this deck puts a creature onto the battlefield most turns, and Aura Shards turns each of those into a dead artifact or enchantment on somebody else's side. It answers less, more often.

Rift is also a Game Changer, so the upgraded list runs two of them alongside Farewell. Bracket 3 permits up to three, and the box was already sitting at Bracket 3 on Farewell alone, so nothing moves.

Blue goes from twenty-four sources to twenty-eight

The land count does not change. What changes is what those lands do on the turn they arrive: untapped goes from twenty-one to twenty-two, conditional from two to seven, and unconditionally tapped from fourteen down to eight.

Blue moves from twenty-four sources to twenty-eight, white from twenty-three to twenty-five, green from twenty-three to twenty-five. Twelve cards in the deck want {U}{U} by turn three, and the hypergeometric maths says thirty-six blue sources would make that a 90% proposition. Twenty-eight gets it to 78.5%, up from 69.4% in the box. I am not going to pretend a three-colour deck reaches thirty-six blue sources. That improvement is what is honestly available, and half of it costs under two dollars.

The rest of it: average mana value falls from 3.71 to 3.48, spot removal goes from one card to five and stops reading as low, card draw drops from twenty-five to twenty-three and is still roughly double what a Commander deck normally runs, sweepers go from three to two, and the bracket floor stays at 3 in both directions.

All fourteen adds together come to $67.35, and $5.71 of that is the eight lands.

The upgraded decklist

Here is the finished list with all fourteen swaps applied. Hover any card name on this page, in the list or in the prose above, to see the card and what it costs today.

Load into the deck builder Copy decklist

“Load into the deck builder” opens this list in your account, pinned to the printings the box shipped, ready to edit or save. “Copy decklist” gives you plain text for Moxfield, Archidekt or Arena — no account needed.

FAQ

Does Morska trigger on my opponents' turns?

Yes, and it is the most missed line on the card. Morska reads "whenever you draw your second card each turn", not each of your turns. Cracking two Clues during somebody else's turn gives her two counters and fires Alandra, Sky Dreamer, Jolrael, Mwonvuli Recluse, Detective of the Month and Ethereal Investigator as well. That is the whole reason this guide spends eight land slots buying untapped mana.

Why cut Reliquary Tower from a deck that draws this many cards?

Because Morska already does it. Her first line is "you have no maximum hand size" and so is Reliquary Tower's, and your commander is available every game. What is left is a land that taps for one colourless mana in a three-colour deck that wants double blue on turn three. Cutting it is the one upgrade on this page that costs nothing.

Should I be cracking Clues or holding on to them?

Both, and the split is the deck's real decision. Six cards care how many artifacts you control: Kappa Cannoneer, Mechanized Production, Shimmer Dragon, Tezzeret, Betrayer of Flesh, Thought Monitor and Whirler Rogue. The commander wants one Clue spent per turn to hit the second draw. The practical answer is to crack the one Clue that triggers Morska and bank the rest, and to prioritise Tezzeret, whose static ability makes the turn's first artifact activation cost two less.

How much does the whole fourteen-card list cost?

$67.35 at Scryfall market prices for the default printings, checked 10 September 2026. Cyclonic Rift alone is $39.37 of that. The eight replacement lands, which are the swaps that change the most games, come to $5.71 together.

Can I skip Cyclonic Rift and still fix the deck?

Yes. The eight lands and the four cheap instants come to about $12 and they address the two things the box actually got wrong, which are tapped mana and having a single answer to a resolved permanent. If you want a second big card instead of Rift, Aura Shards at $28.33 fits this deck well, because you put a creature onto the battlefield most turns and each one destroys an artifact or an enchantment an opponent controls.

Does the upgraded list still sit in Bracket 3?

It does. The box is already Bracket 3 because of Farewell, and adding Cyclonic Rift takes the deck to two Game Changers where Bracket 3 allows three. Nothing else in the fourteen swaps touches the bracket criteria, so the floor is 3 before and after.

How does Deep Clue Sea play if I never buy a card for it?

Slowly, and then well. It runs 37 lands, 32 creatures and an average mana value of 3.71, with 25 cards that draw against a normal range of eight to twelve, so it is never short of cards. It is short of mana and short of answers: fourteen lands enter tapped and Swords to Plowshares is the only card that removes a resolved permanent. Against other precons that is fine. Against a table that upgrades, the deck draws twenty cards and watches somebody else win.