Revolt and Improvise: The Math and the Brews

The brewing season for a new set is always fun. Last week, I presented various decks based around cards like Siege Modification, Sram’s Expertise, Rishkar, and Felidar Guardian. This week, my brewing and number crunching efforts will be focused around the two new mechanics from Aether Revolt: revolt and improvise. I’ll first do some math and then brew up one deck around each mechanic.

How many enablers do you need for a turn-2 Greenwheel Liberator or Metallic Rebuke?

Let’s start by running some numbers. A Terrarion or other enabler on turn 1 allows you to play a 4/3 Greenwheel Liberator or Mana Leak on turn 2. That’s pretty good. But without an enabler, these cards are highly underpowered. This naturally raises the question: What is a good number of enablers to have for a dedicated revolt or improvise deck?

This question is close to the problem of how many Dragons you need for Orator of Ojutai. I investigated that problem in my Magic Math article on Dragons of Tarkir—my analysis today will be very similar. The only difference is that you need to play the enabler on turn 1—if you draw it on turn 2, it’s too late. But the formulas from that article are straightforwardly adapted to account for that.

In line with that article from two years ago, the number that I’m interested in is the probability of drawing at least one enabler (such as Terrarion) in your top 8 cards, conditional on having at least 1 payoff card (Greenwheel Liberator or Metallic Rebuke) in that top 9. So intuitively, I’m looking at turn 2 on the draw, I tacitly assume I always have 2 lands or more, and I only care about whether or not I have an enabler if I have at least 1 payoff card. For a deck with 4 payoff cards, the results are as follows.


Click to enlarge.

For Greenwheel Liberator, I would like to trigger the bonus in at least 80% of the games where I can play it on turn 2. If it’s less than that, then I don’t expect the card to be good enough. By my calculations, 12 enablers gives 82.8%, so 12 is the number I would recommend. The revolt deck in this article will indeed contain 12 enablers.

Metallic Rebuke is the type of spell that you may not even want to play at the first possible opportunity. Playing it off-curve is not the worst, so the enabler count is less stringent. The improvise deck in this article will still contain 12 turn-1 enablers, but it’s fine to run Metallic Rebuke in a deck with less. You’ll just have to accept that you won’t be able to consistently play it on turn 2 in that case.

Now, let’s do some brewing!

4-Color Revolt Aggro

I could’ve just made a green-white good-stuff deck, but I decided to try and push the limits. Let’s go over the cards one by one.

The 12 enablers that I chose for the deck all allow for an on-curve revolt creature without losing card economy. They also put a card in the graveyard for Renegade Rallier and help fix the mana for this 4-color deck.

The benefit that these cards have over alternative revolt enablers like Greenbelt Rampager or Implement of Combustion is that they don’t require mana on the revolt turn. Renegade Map, Terrarion, and Unbridled Growth allow you to set up a Greenwheel Liberator on turn 2 or a Renegade Rallier on turn 3—the alternatives don’t.

These are the payoff cards. If you could guarantee revolt every single turn, then all of them would be above the curve.

These 2-drops allow you to trigger revolt every single turn—the Monkey is exiled at end of combat, and Aethergeode Miner can blink. If these creatures die, then they can be brought back by Renegade Rallier.

The red is speculative. The mana base is… ambitious, even with twelve 1-mana fixers. And I’m not sure that this pump spell is worth it. But it synergizes very well with Solemn Recruit (which has double strike) and Kari Zev (either to boost the 1/3 first striker when they double-block or to boost the Monkey token that was going to depart anyway). It also offers a way to beat Ishkanah, Grafwidow, which the deck would have a lot of trouble with otherwise.

If the red splash turns out to be too ambitious and/or your opponent has cheap removal for the Solemn Recruit/Invigorated Rampage combo, then you could board out Kari Zev, Solemn Recruit, and Invigorated Rampage for Aethergeode Miner, Woodland Wanderer, and Declaration in Stone, respectively. Those non-red replacements would all fill a similar role. I even made sure you can swap Mountain and Needle Spires for Aether Hub. Besides that craziness, the sideboard also has more normal cards, like Walking Ballista. Its main purpose is to stop the Saheeli Rai/Felidar Guardian combo by pinging Saheeli, but it can also trigger revolt at will.

Now, I don’t advocate going into a tournament with an alternative plan in your sideboard in case you decide during the day that the fourth color was ultimately too much, but it’s not a terrible approach if you didn’t have time to test. Back in 2006, Dutchman Ruud Warmenhoven showed up the night before Pro Tour Honolulu without a deck because he had been backpacking around the world. We handed him a black-white aggressive deck, which he played more or less blind. After he lost his first couple of matches, he concluded (if I recall correctly) that his losses were partly due to drawing too many Ravenous Rats, which were terrible. He boarded them out for Castigate every match after that, improving his deck in the process, and easily cruised to the Top 8. All I’m saying is that it has been done before.

The Verdict

Will my revolt brew be competitive? It looks fun to play, but I’m not holding my breath yet. Compare it to delirium. Even when their keyword demands are met, Narnam Renegade and Greenwheel Liberator are worse than Gnarlwood Dryad and Grim Flayer, and delirium decks never had a big problem getting 4 card types in their graveyard. Moreover, you could at least play the delirium creatures on curve with a straight face, whereas a 2/1 Greenwheel Liberator on turn 2 is just embarrassing.

What’s more, there is no midgame payoff like Ishkanah. You have… Airdrop Aeronauts? Aid from the Cowl? These cards don’t come even close.

Another issue is that thanks to the stringent demands on the enablers, you’ll have plenty of hands where you don’t know whether you are mana screwed or mana flooded. That’s what you get when you play 12 cantrips and 18 lands.

So I’m not convinced yet. The card I actually like the most from this deck is Unbridled Growth. I tried to brew up an Enchantress-esque deck with Sram, Senior Edificer, Unbridled Growth, and Weirding Wood, but there just doesn’t seem to be a critical mass of engine cards yet. Maybe after Amonkhet.

Blue-Black Improvise

It’s not quite Modern Affinity, but I at least did my best in trying to port Ornithopter to Standard. I ended up with this blue-black list after first exploring builds with Thraben Inspector and Maverick Thopterist. Those cards are good, but I wasn’t happy with the resulting mana base. The deck needs double-black and double-blue on turn 3 and can’t run that many lands, so sticking to two colors seemed better. Let’s go over the cards one by one.

Herald of Anguish is the main inspiration behind brewing this deck, as it can easily dominate a game. It’s a big body, and all of his abilities are valuable. Servo Schematic and Terrarion are awesome to sacrifice for his -2/-2 ability, by the way.

As I was curious how consistently the deck could cast him on the early turns of the game and whether I had put enough artifacts in the deck, I set up a simulation. The Java code can be found here. Some of the key assumptions and simplifications are that all lands are Underground Seas, that you mulligan any hand with fewer than 2 or more than 5 lands and won’t use the free scry, that you are always on the draw, and that you have a fixed sequencing (e.g., play Cogworker’s Puzzleknot or Herald of Anguish over Reverse Engineer, if possible).

I disregarded games where we didn’t draw Herald of Anguish or where we couldn’t cast him in time because we missed land drops (because then the artifact count wasn’t the problem). Here, “in time” means turn 4. I kept the games where we cast Herald of Anguish or where we were stuck with him in hand because we didn’t have enough artifacts in time. The resulting set of non-disregarded games where we drew Herald of Anguish in time was split as follows:

  • Probability of a turn-2 Herald: 0.01%
  • Probability of a turn-3 Herald: 18.62%
  • Probability of a turn-4 Herald: 61.21%
  • Probability of not having enough artifacts for a turn-4 Herald: 20.15%

I’m satisfied with these numbers. The deck can cast him on turn 4 with reasonable consistency—4 lands, a Cogworker’s Puzzleknot, and another artifact is already enough—and can cast him on turn 3 surprisingly often. And the dream start of 4 Ornithopter, 1 Terrarion, 1 Herald of Anguish, and 2 Swamp is highly unlikely, but a turn-2 Herald is not impossible.

These are the big payoff cards next to Herald of Anguish. Including Herald, there are 11 total, which is the same as an Affinity deck with 4 Cranial Plating, 4 Arcbound Ravager, and 3 Steel Overseer. You often much have to mulligan if you don’t have a big payoff card, but Merchant’s Dockhand, Reverse Engineer, and Aethersphere Harvester help to boost the number of keepable opening hands.

The card I’m most excited about is Tezzeret’s Touch. It’s a more expensive Ensoul Artifact, but at least you get the artifact back if your opponent has a removal spell. Servo Schematic and Terrarion are sweet targets, and Ornithopter and Heart of Kiran are also happy to turn into 5/5 flyers.

The way I look at it, Reverse Engineer is Thoughtcast (an on-theme card draw spell) and Metallic Rebuke is Galvanic Blast (an on-theme interactive card). You want some amount of these effects because they’re pretty powerful, but you can’t play too many because these slots come at the expense of lands, enablers, and true payoff cards. I felt 6 was the maximum.

Naturally, we need a lot of artifacts. I tweaked the numbers for a while until I got 12 artifacts of 0 or 1 mana. That’s good news for Metallic Rebuke and the potential of a turn-3 Herald of Anguish.

I considered a build with Bone Saw and more Cathar’s Shield, as they are like Moxen if you’re holding a lot of improvise cards. But in most games you’ll only draw 1 or 2 improvise cards, so it seemed dangerous to cram the deck full of do-nothing artifacts. I think my list strikes a good balance.

Cogworker’s Puzzleknot and Servo Schematic both put two artifacts on the battlefield for the price of one card, which is exactly what you need to get a turn-4 Herald of Anguish with reasonable consistency. The deck can activate Cogworker’s Puzzleknot via Spire of Industry, Terrarion, or Tezzeret. The two Vehicles round out the curve and the deck.

The sideboard contains Battle at the Bridge, Fatal Push, Yahenni’s Expertise, and Tezzeret the Schemer against aggro decks. The sideboard also has Lost Legacy, Walking Ballista, Metallic Rebuke, and Murder against Copy Cat decks. Finally, Scrapheap Scrounger and Reverse Engineer can be boarded in against control decks.

The Verdict

Maybe I’m biased, but I have higher hopes for this improvise deck than for the revolt deck. Whether it will be competitive will likely depend on how much removal for Herald of Anguish people will have. If everyone plays Stasis Snare or Unlicensed Disintegration, then the payoff may not be worth the effort. But if everyone goes for Shock and Fatal Push, then they will have a hard time answering the 5/5, in which case an improvise deck like this one could be good enough.

What do you think? Is blue-black improvise going to be the new U/R Ensoul? If you have any ideas for improving these brews, then let me know in the comments!


Scroll to Top