There are Pro Tours in which you have a deck you want to play, and Pro Tours in which you do not. PT Amonkhet was the latter. When you have a deck you like, you feel on top of the world—the PT is awesome and you can’t wait to play the Constructed rounds. When you don’t, you agonize for hours until it’s finally 11:30 p.m. and you have to submit something—you start lamenting the fact that you just blew your biggest chance of getting money and Pro Points, and you hope you can manage to go 9-6.
Choosing a deck for a PT is a rollercoaster of emotions that we put ourselves through every three months, and for the last six months I’ve found myself at rock bottom, only to be taken up again by a combination of 10% preparation, 10% talent, and 80% incredible luck. You see, I actually played what I believe was close to the best possible deck in each of the last two Pro Tours, but on neither occasion did I know that going into the tournament.
For PT Aether Revolt, our team tried a variety of options before settling on Mardu Vehicles the day before. In the end, Mardu Vehicles players took the top 6 spots in the tournament, and I was one of them. For Amonkhet, we again tried a variety of options before resignedly settling on Temur Marvel one hour before deck lists were due. Our deck was just an MTGO deck, changed by just one card—in 50 Pro Tours, this is something I’ve never done before. In the end, Temur Marvel players took half the slots in the Top 8, with EFro being one of them.
When I say it was 80% luck, I’m probably exaggerating a bit. We could have copied any of several different archetypes, but we chose to copy a Marvel deck. Among Marvel lists, we could have chosen a variety of colors and strategies, but we chose Temur Marvel with Gearhulks and Censors. We could have added any one card to our sideboard, but we added a Chandra, Flamecaller. Our testing didn’t give us any of these answers clearly, but it gave us enough information to choose the best of the options we had. It’s not ideal, but it’s something. I still feel incredibly lucky that we landed on two great decks, when we could easily have ended up with something unplayable (like the half of our team who played B/G Aristocrats. Sorry guys!).
Speaking of which, this is the deck I almost played:
This looks like a deck I’d never play, and the fact that I almost did is a testament to my desperation. I am, and have always been, a fan of good cards. Rather than hoping I can get a bunch of 3s working together to become a 9, I’d rather just have 8s to begin with. Some people on our team are not like that. Sam Pardee is legendary for his deep affection for bad cards. Cards like Loam Dryad, Blisterpod, and Zulaport Cutthroat are a bug for me, but a feature for him. Most of the time, his decks work. Almost always, actually.
It’s because of Sam that I almost played this deck. It felt decent, it was different and unexpected, and we had a good sideboard plan for all of the format’s most popular decks. In the end, though, I knew it probably wasn’t as good as the other options. In fact, we all knew that was probably the case, even Sam—he’s usually assertive that his choices are good (and they usually are), but this time he wasn’t nearly as excited about his deck. The main question was how much value we gained by being a surprise deck, and how much that made up for being a bad deck. If your deck is a 53% deck that becomes a 58% deck once people don’t know how to play and sideboard against it, then it’s better than a 55% deck.
The people who chose the B/G deck believed that they would get to 58%, but I didn’t. I thought we were starting at maybe 48%, and then we might get to 52% because people would be unprepared, but I believed both Mardu and Marvel were better than 52% to begin with.
I also worried about Zombies, both the deck itself and the reaction to it. We thought Zombies was a bad matchup, and we didn’t know how popular it was going to be, but I also thought people would be skewing their decks to beat Zombies, given its MTGO popularity, and the B/G deck would then get hit by all the splash hate in the form of sweepers. In the end, even if Zombies itself turned out to be a budget MTGO phenomenon and saw no play at the PT, I thought its mere existence would make the B/G deck worse. On top of that, Siggy and EFro, two of the best players on our team, were adamant that playing the B/G deck was a horrible mistake, which helped solidify my conclusion.
So, on Thursday before the tournament, our team split in three groups: the Hopeful, who thought B/G was going to be the best choice for us (Sam, Matt Nass, Hayne, Joel, and Josh Mcclain); the Resigned, who settled for the next-best-tested deck Mardu Vehicles (Oliver Tiu, Ivan, Steve, Ondrej, Petr); and the Brave, Handsome, Smart, and Daring, who locked themselves in a room trying to find an alternative to those two things (Siggy, EFro, Shuhei, Jacob, and myself).
There is also a secret fourth group, the “0-5s” (composed of just Ben Stark), who played B/G Energy for whatever reason. After Day 1 of the tournament I messaged Ben to see if he wanted to have dinner with us, and this conversation ensued:
In the end, the Brave, Handsome, Smart, and Daring group analyzed a bunch of options and settled on the “Control” version of Temur Marvel. It’s not really a control deck, but it plays more control elements than other Marvel decks (it has a counterspell!), and plays Torrential Gearhulks as an alternate plan for when they Dispossess or Lost Legacy you.
We mostly chose Marvel because it was the safest option—we believed Standard to be a bit of rock–paper–scissors, and Marvel was the best in this scenario. If we chose Zombies, we ran the risk of everyone playing Marvel and our deck being bad. If we chose Mardu, we ran the risk of everyone playing Zombies and our deck being bad. If we chose U/R Control, we ran the risk of everyone playing Mardu, and our deck being bad. If we chose Marvel, then we thought we were bad against U/R (almost unwinnable pre-board, below 50% post-board), but we thought U/R was very unlikely to be the most popular of those four choices, and we thought we were good against both Zombies and Mardu.
In the end, we were mostly correct, and U/R Control ended up being only about 6% of the metagame, whereas each of the other three decks approached 25%.
This is the list Siggy, Efro, and I played:
Jacob and Shuhei played a slightly different mana base and two different sideboard cards.
This version is almost identical to the one that was played in an MTGO PTQ—the only difference is that version had 1 Anticipate and 3 Ulamogs. It’s not every day you cut a 2-mana selection spell for a 10-drop creature, but in this case I think it’s correct because finding Ulamog is so important for you. In fact, I’d say that anyone playing fewer than 4 Ulamogs is making a big mistake. Sure, you will draw Ulamog more often, which is bad, but you just have to suck it up and accept that you mulliganed when you do.
Here’s a short breakdown of the sideboard:
3 Tireless Tracker: Tracker is the best “fair plan” available in Standard. I played post-board games in which I went Tracker, removal, removal and just won because my opponents didn’t have removal of their own. I wouldn’t expect anyone to take out all their removal against Marvel, since you still have a number of creatures, but people will generally trim removal, and drawing even one or two extra cards is good in the grindy, 1-for-1 sideboarded games.
3 Negate: Negate is not the best counterspell against control (that’s Dispel) or the mirror (Rejection), but it’s the only one that’s truly good in both matchups, so it’s here in multiples. It’s also the best against random things like the New Perspectives deck (which is a very bad matchup in game 1).
2 Dispel: The best counter against control decks because it lets you force through a Marvel much more easily. I also like to bring it in against most Marvel mirrors post-board, since they will have their own counterspells and cards like Glimmer of Genius and Harnessed Lightning as well.
2 Aether Meltdown: Strictly against Mardu Vehicles.
2 Radiant Flames: Mostly against Zombies. This is an interesting slot that can be given to Kozilek’s Return or Sweltering Suns, each of which has pros and cons. Kozilek’s Return is easy to cast and can be Gearhulked back, but it also doesn’t kill most of the Zombies if they have a Lord or a Liliana’s Mastery. Sweltering Suns can be cycled or Marveled, but double-red is hard to get with our mana base. We ended up choosing Radiant Flames because it’s the easiest to cast and it’s customizable (you can avoid killing your own Virtuoso, for example). It’s annoying that it doesn’t work with either Marvel or Gearhulk, but we thought being able to cast it for sure on turns 3-4 and killing everything through a Lord was more important.
1 Ceremonious Rejection: This is strictly for the mirror, but I think it’s the best card for that. The way most post-board mirror games go is that the person on the play can tap out on turn 3 for a threat (Rogue Refiner, Whirler Virtuoso, or Tireless Tracker), then the person on the draw cannot do that because they expose themselves to Aetherworks Marvel. As a result, being on the play gives you a huge advantage. Rejection mitigates this a bit because it lets you play a threat and still keep up mana to counter Marvel a turn sooner.
In practice, I found myself siding in either the second Gearhulk and the Chandra or the Tireless Trackers in most matchups—sometimes I sided in both. Many decks can interact with Marvel after sideboarding, so I would usually make my deck less Marvel-focused by taking out a number of Woodweaver’s Puzzleknots and a couple Ulamogs, but I never took out the Marvels themselves. I also often cut 1 Glimmer against fast decks like Mardu, and Magma Sprays against decks with few targets for them. When I absolutely didn’t know what else to cut, I cut a Rogue Refiner or a Whirler Virtuoso, depending on whether I expected them to interact with my Marvel or not (if you’re getting Dispossessed, then you shouldn’t cut Virtuoso because it’s your other way of using energy).
Playing the Deck
• In pre-sideboarded games, your plan is simple: you’re a combo deck. You want to get to 6 energy, find Marvel, spin it and find Ulamog, and win. You have other ways to win, mostly via Rogue Refiner and Whirler Virtuoso beatdown or the occasional Torrential Gearhulk, but Marvel itself is your primary route to victory and should be your plan for most matchups.
• Post-sideboard, things change. People bring in disruption (Negate, Transgress, Dispossess, Lost Legacy, etc.) so you need to have an alternate plan. On the play you can Censor their Dispossess, but on the draw you cannot (since they can just wait until turn 4), which should also factor into your plans.
• Assuming an even distribution in the deck, you’re 35% likely to hit an Ulamog in a Marvel spin (with 3 Ulamogs, this number drops to 27%). If it’s turn 4 and you haven’t drawn an Ulamog yet, the chance is around 41%.
• If you’ve already Marveled a couple times, it’s often better not to cast an Attune with Aether, because this will shuffle all your non-Ulamog cards back in your deck. Thinning a land out of your deck is good, but not when you already know that 10 of the bottom cards aren’t the ones you want.
• If you board in Tireless Tracker, be careful with your land drops. You want to make sure you can have a turn 4 of Tracker plus Magma Spray or Tracker plus Dispel most of the time, but you also want to play Tracker before you play your land for the turn, so make sure you keep the right red or blue land in your hand.
Looking back on this PT, I think Team Genesis had the best Marvel list simply because Chandra was better than Gearhulk. Chandra isn’t good in the mirror or against control, but Gearhulk isn’t great in those matchups either (especially game 1 where your flashback options are limited), so I think Chandra is just a better Gearhulk overall and should be adopted over it, especially because Zombies is going to grow in popularity.
I liked Censor, even if it made me keep some hands that didn’t pan out. The flexibility of countering a turn-4 Gideon or Marvel was great. If you go the 4 Chandra route like Genesis did, then you likely need Servants, but you can still afford to play a couple Chandras without Servants. I don’t think Servant is bad per se, but I don’t like giving people targets for their Fatal Pushes or Magma Sprays if I don’t have to, and I think their version is a bit lacking in ways to find Marvel. Our deck had 4 Censors to cycle and 4 Glimmers, and they had zero of each (yes, I know Chandra theoretically digs toward Marvel too—that helps a bit, I guess). We were trying to be the best Marvel deck, and they were trying to be the best deck when they didn’t draw Marvel. The best answer is likely a compromise.
All things considered, Marvel is just the best deck right now (though it’s possible U/R is the best choice for a given tournament). If I were to play a Standard tournament tomorrow, this is the build I’d play:
Temur Marvel, Updated
Paulo Vitor Damo da Rosa
No Gearhulk makes your Glimmers a bit worse, so I’m fine cutting 1. You still want some number of search effects, but you no longer need to be able to flashback something, so you don’t need to play 4 Glimmers. With a second Mountain in the main to support Chandras, you can probably play Sweltering Suns instead of Radiant Flames, but I’m still not positive you should play a second Mountain even with Chandras in your deck.