With Modern Horizons now fully spoiled, it's time to wrap up this series of spoiler articles. Last week, I said that the set was a bit underwhelming. At the time it looked like a Standard set with some added complexity. Immediately afterward, the most potentially powerful cards in the set were revealed, causing me to eat my own words.

Onslaught Cycle Lands

Since they were revealed, the enemy-colored Horizon Canopy cycle has received a lot of press. It makes perfect sense; Canopy is a great card and used in many decks. It fixes mana and can be cycled late game, which makes it a contender for best land ever. Now more decks have access to the effect, and it stands to reason that it will rock the format. However good the new Horizon lands actually are, they're not the most powerful lands in Modern Horizons. They're very good, but they'll never be broken or abusable; they cost too much time, both in mana and land drops.

The reprinted Onslaught unicycle lands, on the other hand, have been absurd before and may be so again. The unicyclers always come in tapped, which isn't good in Modern. However, being lands is more of a bonus. They're meant to be cycled and increase velocity . This is similar to the Horizon lands, but cycling doesn't require a land drop. Instead, these cards combine with Life from the Loam to produce a card advantage engine that cannot be out-grinded. Every turn, interested pilots can draw up to n-2x cards, where n is the number of lands they control and x is how many times they cast Life.

Loam Is Real

When the bicycle lands were debuted in Amonket, my testing showed that they were not good enough for Modern. The mana investment was too high, the engine too inefficient, and the rest of the shell too dependent on the engine to be functional. I predicted then that it would take the unicyclers to make it work. Shaving a mana off anything makes it far more powerful (consider Counterspell vs. Cancel), and when it's part of an engine, the increase is exponential. Therefore, I feel very confident that Assault Loam is a real deck now.

History supports this theory. Loam decks proved a force in Extended for as long as Onslaught Block and Ravinca were legal. And this was a format where Mind's Desire and Chrome Mox were legal, graveyard hate was almost non-existent while dredge and Ichorid ran free, and Affinity had artifact lands. Loam's run started as a prison engine in CAL, evolved into Assault Loam, and survived until the end of its legality. Modern is a very different format form old Extended, but if the Loam engine could hang back then, it may well still be good.

Drifting Away

Loam isn't the only deck the unicyclers could resurrect. Astral Slide is a beloved deck that never had much impact beyond its Standard run, and now has another chance. Back when I was first getting into competitive Magic, Slide was a powerful and plodding board control deck that used the namesake card to contain opposing creatures, cheaply un-morph its own Exalted Angels , and then close the game with Lightning Rift and Decree of Justice. It never had much impact outside of Standard because it was so slow, and Living End fills a very similar role, but is faster. However, I know a lot of players that loved Slide back in the day and lamented that Modern doesn't include Onslaught block so they could keep playing their deck.

Astral Drift is Slide with upsides, and with the unicyclers back there's potential for a resurrection. However, several problems need to be overcome. First and foremost, the speed problem must be solved. The old Slide deck was glacial, and a reborn version would still need to tap out for a three-mana enchantment that does nothing on its own. If you're still alive by then, you start cycling cards to dig you deeper and blink out attacking creatures.

This brings me to the second problem, one of payoffs. Lightning Rift was cheap to cast and activate. The best analogue is Faith of the Devoted, which isn't as versatile and costs more. Same problem with Drake Haven. Even if those weren't big enough hurdles, why not just play Living End? It offers the same durdly gameplay, but adds competitive precedent.

Snow-Thing Awakens

The other big problem I previously mentioned was a lack of reasons to play snow card. The snow basics are essentially free, but the actual spells required a lot of hoop-jumping to make work. There needed to be more of a reason than Ice-Fang Coatl to bother. That reason has appeared in Marit Lage's Slumber.

As a fixed (uncheatable) version of Dark Depths, Slumber is notable first as a signal that Depths isn't getting unbanned, which is for the best at this point. Secondly, it may end up an exceptional control card. As a two-mana blue enchantment, it directly competes with Search for Azcanta. This is not a fight many cards would relish, but I think Slumber has a chance. Search only triggers once, on your upkeep. Slumber triggers whenever a snow permanent hits play. Combine the basics and some snow creatures and you can scry multiple times a turn. Secondly, the payoff for flipping Search is recurring card advantage. Slumber potentially wins the game outright. Not many decks will be able to make Slumber work, but the one that can may be greatly rewarded. The only control deck that already relies on basic lands is Blue Moon, so that's where I'd start testing.

The Best Thalid-Ninja-Hound Ever

Prior to Modern Horizons, Mutavault was the best changeling in Magic. Considering the competition, that wasn't a hard-fought victory; Chameleon Colossus and Mirror Entity see some niche play, but always in small numbers. Mutavaultmust now step down from its throne, because Unsettled Mariner is absurd. Every tribal deck that can run this card should be doing so.

The fact that it benefits from and triggers any tribal synergies is good, but if that was the only criteria, then Mothdust Changeling would see play. It's the ability that makes the Mariner. Mana Tithe-ing anything that targets your creatures is worse than hexproof, but not everything can be Spirits. It's still a very powerful ability that very effectively protects your creatures for two mana.

Spot removal in Modern is meant to break up attacks and slow down creature rushes before the more powerful cards come down. It frequently takes several in a turn to survive a tribal onslaught, and Mariner makes that a difficult task. Midrange and control will struggle against this card, and I predict they go more sweeper-heavy to compensate.

However, that's not the only benefit of Mariner. It also protects you. This gives him utility against combo decks that similar protective creatures like Kira, Great Glass-Spinner cannot match. While most tribal decks are already fine against combos like Storm, they tend to struggle against Valakut, the Molten Pinnacle. With Mariner out, Valakut can't just go for the kill; they must remove it or have enough mana to pay for all the taxes. This is achievable, but takes extra time. And time is all a tribal deck needs to win.

How Many?

However, there is a question of how many to run. Against another creature deck or Tron, Mariner is just a bear. The current metagame doesn't have a lot of targeted removal or combo win conditions. It is great against Dredge, since each Conflagrate target gets its own trigger. It also forces UW to actually hit its sweepers on curve rather than banking on Path and Detention Sphere buying time.

I would run a full set in UW Merfolk because it's perfectly on-curve and replaces Kira, but I'm not sure about other decks. Humans' maindeck is fairly set, with one permanent flex slot. Thalia or Kitesail Freebooter sometimes get shaved if the metagame is unfavorable.  That's not much room there, and Mariner is the kind of card you maindeck. It's decent but very medium in the sideboard, and Humans' sideboard is already stacked. Mariner is good enough for Humans, but the metagame question will dictate how many the deck needs to pack.

Sick Machinations

Sticking to the subject of Humans, we've also received Plague Engineer. Following in the tradition of Prime Speaker Vannifar and Cabal Therapist, Plague Engineer is Engineered Plague (see what they did there?) given legs to power it down. Humans has been highly successful for approaching two years now (thanks to favorable metagame conditions), so it makes sense to curb it with anti-tribal hate. However, Wizards clearly didn't want it to be too hateful. Tribal decks are a cornerstone of Modern, so I appreciate Wizards making a more counter-answerable version of Plague.

That said, I doubt that Engineer will see widespread play or do much to actually curb Humans. Engineer costs three, and by then Humans can grow enough for Engineer to merely shrink them rather than wipe the board. Any subsequent Thalia's Lieutenants or Thalia, Guardian of Thrabens will die on entering, but that won't stop them from growing already-existing copies of Champion of the Parish or Lieutenant. Humans also plays a lot of answers in Reflector Mage, Dismember, and Deputy of Detention, so relying on Engineer is very risky.

Where I see Engineer seeing play, ironically, is in Humans. Right now, Humans struggles against other go-wide tribal creature decks. Thanks to Noble Hierarch, Engineer can land on turn 2 and devastate against Elves, Goblins, and sometimes even Spirits. Most of the creatures in Affinity and Hardened Scales are constructs, so an accelerated Engineer is quite good there too. The tribal hate card may end up significantly boosting the best tribal deck.

Things Get Weird

However, the problem with Humans relying on Plague against creature decks is the mirror. The Humans mirror is either decided by one player going off with multiple Thalia's Lieutenants or by dominating tempo with Reflector Mage. It's not the worst creature mirror (Merfolk is), but it is very frustrating and slightly brainless. Plague Engineer promises to stifle the former by shrinking the team and killing any subsequent Lieutenants before they can grow, which is a very strong plan without downside since Engineer isn't symmetrical.

But that clause may also be the Engineer's downfall. Another part of the Humans mirror is Phantasmal Image. Image copying opposing Lieutenants and Mages is incredibly good, and is a reason there's some fear about making the first move in the mirror. It is often necessary to play Mage or Lieutenant to get the deck moving. However, doing so provides a very tempting option for Image, and that card is a four-of. If your opponent is missing a critical card, it's often a good idea not to play yours so their Image can't get them back into the game. You want to do exactly enough to win, but not enough to let opposing Images wreck you.

Engineer further complicates such circling. There's almost certainly not room in the Humans sideboard for more than two Engineers, or three in a tribal heavy meta. That means the odds will always be better for the opponent to have Image than for you to have Engineers. The question then becomes whether the damage you do by playing Engineer outweighs the risk of letting your opponent have one. This dance may lead to a lot of metagaming and next-leveling in the mirror.

Mystic with Gears

My final card is another engineer, though of a more traditional variety. Goblin Engineer is a throwback card, and interestingly not to just one. Engineer presents like a reference to Goblin Welder, but combines that reference with Stoneforge Mystic. The former reference is the activated ability replacing artifacts in play with ones from the graveyard (though nerfed and making sense with the rules). The latter comes from the tutoring ability, though it sends the tutored card to the graveyard rather than the hand.

From there, Engineer can do a decent Mystic impression. If you use Engineer to tutor for Sword of Fire and Ice or similar, you can then trade it for an artifact in play, which is functionally the same as Mystic (though easier to disrupt). I could see this working as a red splash in Death and Taxes that makes clue tokens to feed Engineer. Whether this is good I can't say, but it might see some play while Stoneforge Mystic remains banned.

Even if that is farfetched, there are plenty of other fair, value-centric uses for Engineer. Tutoring for and then resurrecting Affinity cards come to mind. However, this sort of effect never ends up being played in fair decks.

Fixed it Broken

Goblin Engineer wants to be broken. Its predecessor, Goblin Welder, was broken way back when thanks to the busted artifacts from Urza block, and it couldn't even tutor for targets. Engineer will never return a Platinum Angel or Mindslaver to play since its ability is restricted to CMC 3 or less. However, it can find any artifact. Tutoring is a powerful mechanic, and so is reanimation. That Engineer combines both makes it very dangerous.

I don't know how to break Engineer. A lengthy Gatherer search for cheap artifacts contained a lot of enablers, but no obvious engine to get the combo moving. However, maybe Engineer doesn't need to do the work itself. Refurbish and Trash for Treasure exist, so there are ways to be a way to cheat in something huge and devastating on turn three. The best I could think of is Sundering Titan, which seems mediocre. If only Blightsteel Colossus could hit the graveyard.

There's also the possibility that Engineer's activated ability is combo-ready, too. I don't know of a great and cheap artifact engine right now, but I could be wrong. Even if I'm right at the moment, Wizards could always print something that is busted for Engineer to tutor for and then cheat in. Watch this card carefully; I can't imagine it won't get abused at some point.

Modern's Upheaval

It is hard to predict how Modern will react to Horizons. There's so much going on that I can't definitely say how the dust will settle. However, I can test decks, and use the ensuing results to extrapolate about the format. Tune in next week to see the results of those experiments.