In Development - Red Teaming

“Playtest the dumb strategies.”

In 1981, Steve Jackson and Nick Schuessler published Game Design Vol. 1: Theory and Practice, a thin but thorough book that covered the basics of game design and development. The quote that opens today’s column comes from that book, and it has stuck with me since the first time I read it. In the book’s context they were mainly addressing the idea of playtesting a wargame, most often a historical wargame, and making sure that you tested options that didn’t make intuitive sense. Should having a supply truck in the same map space make your tank invulnerable to artillery fire? Probably not, but a flaw in the Panzerblitz rules makes that possible, and playtesting the “dumb” strategy of having supply trucks driving next to front-line tanks would have caught it. At the same time, you don’t want to rule out things that could have happened that make little sense, like an unsupported British infantry brigade holding off elements of two SS Panzer divisions for more than a week (hint — this happened).

The core lesson here, the one that is transferable to Magic, is to test before we decide. Looking at an environment dominated by unfettered Disciple-Ravager Affinity, it’s unintuitive to imagine that someone could top eight with a deck that hopes to win by casting a nine-mana spell – but someone did. Even someone who saw the land-searching potential in Mirrodin block might have concluded that the approach would be too slow, but the same person who took the time out to goldfish the deck a few times might have seen some glimmer of feasibility there and realized that a playable deck existed.

In approaching the new Shards-Zendikar Standard we may be tempted to write ideas off ahead of time, whether we’re discussing entire deck designs or specific card selections. Is Cancel still primarily a Block and Limited card rather than a Standard playable? Most likely – but it’s good to test these assumptions every so often, especially in the early, early playtest period when we’re just seeing if deck designs can hold up.

With that idea firmly in mind, let’s talk about testing.

Red Team
In business settings, a “Red Team analysis” is a stage of proposal development where a group that hasn’t been working on the proposal in question comes along with a fresh perspective and tries to attack the proposal, finding game-ending flaws. The “Red Team” label here goes back to military exercises, where it’s the now-somewhat-historic moniker for the opposing team, as played out by experts from your own side. The purpose of the Red Team analysis is to see if your training, strategy, or proposal can stand up effectively to an attack by the expected opposition, whether that’s a metaphorical attack in the business environment or a very real-world attack carried out in an urban combat training center.

In our terms, the “Red Team” phase is the point where we actually run our decks up against our best guess at some of the key opposing archetypes and see if they survive. This is a playtesting phase, but it is distinct from the kind of playtesting we do once we have a known deck and are planning for a specific tournament. We’re neither fine-tuning the deck nor improving our play skill with the deck. Instead, we’re still figuring out if it even works or not.

At the same time, we learn more about the true nature of the environment.

Last week, I laid out my understanding of the new Standard, which I distilled down to “cascade, black, vampires, and big spells.” To address this proposed environment, I put together a couple of super-low-curve aggro decks and a pair of Ascension-centric control builds, both sets standing in for a larger archetype. These archetypes subsequently went through our Magic-style Red Team analysis, which meant actual games featuring actual decks, and not just best guesses about play environment.

I’ve collected a number of prospective Standard builds from across the Magic writing community, and rather than linger too long on that topic, I’ll just recommend that you do the same. Specific decks included a number of cascade variants, some mono-black (aka vampire) aggro or midrange, mono-red aggro, and a couple of different variants of Baneslayer control. Once again, one would not go too far wrong by checking in with Josh’s “Zendikar inspirations” article and the builds therein. I’ll plan on describing in more detail how I collect deck lists for playtesting in a future In Development.

What we learn about the environment
One of the fun elements of Magic playtesting is that it tests your hypotheses about not just your own designs, but the play environment as well. My concern for the new Standard was cascade, vampires, and taking an Ultimatum to the face. Testing with my Red Team gauntlet let me know to what I extent my hypothesized nouveau Standard was going to actually match reality. Here are a couple things I picked up in the process.

Cascade decks are, indeed, legitimate. I’m sure no one is shocked. You can basically port many of the cascade builds from Block, or even from the prior Standard, right into Alara-Zendikar and they work just fine. Any strategy you’re going to use needs to be able to deal with the possibility of the opponent’s deck vomiting up two to three spells in one cascade of doom. If you can’t deal with these bursts of card advantage, you’re going to lose.

Mono-red aggro is the litmus test for surviving aggression. If you haven’t done so yet, go proxy up the mono-red aggro deck from Josh’s article or a similar one from another source and just goldfish with it for a while. Then build a control deck, and realize that there’s not a ton of life gain to work with here, and the mono-red aggro builds start attacking on turn one and don’t let up.

Finally, Baneslayer. Baneslayer, all on its own, independent of deck context, is a performance benchmark for any deck design you’re going to try and field. If you can’t reliably handle a Baneslayer, you might as well stay home.

The presence of mono-red aggro and Baneslayer as twin survival tests for any deck has tended to militate against the viability of those big haymaker spells I was worried about earlier. They’re still there, but an Identity Crisis or even a Cruel Ultimatum is nowhere near as horrid as the second Baneslayer of the game or Goblin Guide into Hellspark Elemental into Ball Lightning (15 damage by the third turn, if you didn’t do the arithmetic).

So, if you wanted to proxy up just a bare bones gauntlet of decks to rapidly test prospective designs, I’d go with one Baneslayer control, one Cascade midrange, and one mono-red aggro. Cover those, and you’re probably good to go.

You may want to check out last week’s article to review the prospective decks I began with before moving on to the rest of today’s piece.

Testing Vastwood Aggro
Last week’s Vastwood Aggro builds were an attempt to see if low-curve aggro that was not particularly conscious of card advantage could prey on what I imagined might be a slower play environment. The very brief answer is “not so much.” But let’s talk about testing, and then about what probably went wrong there.

I rapidly discarded the Vastwood Jund build. It faced multiple dilemmas, starting with the issue of the mana just not working out exceptionally well. A Jund build needs to hit all three colors in the first two turns to reliably kick out a truly aggressive start. This is tremendously unlikely. At the same time, trying to hit these colors frequently meant that the marquee card of the deck, Vines of Vastwood, was hard to cast in kicked form. This, in turn, made the Slaughtermaster a terrible choice. I tinkered a bit more, including making the obvious choice of adding in Bloodbraid Elf, but the deck was frequently dead to a Baneslayer, and was being outrun by mono-red. Terrible.

The Naya version of Vastwood was more promising. It came out of the gates very quickly, but tended to be overwhelmed by cascade decks starting at the midgame. This problem was resolved by adding Ranger of Eos, which let it reload with Nacatls and other one-drops in the mid-game, and even offered some wins against Baneslayer simply by dint of having piled in a lot of damage already, and offering the possibility of massively “overcommitting” to an attack to win the game. Still, this was incredibly shaky, and I found myself comparing the adjusted Naya Vastwood build to earlier low-curve aggro decks to try and divine the missing pieces.

We can pause for a moment to check in with the “best build” for Naya Vastwood. I’m not recommending this deck, but wanted to give you a taste of where this deck went before I binned it.

Naya Vastwood Aggro (not recommended)

There were other variations, of course. I changed up the number of Rangers, replaced them with Bloodbraids, added some Woolly Thoctars, and so forth. But after due consideration, I don’t think this breed of ultra-low-curve Zoo can work right now.

The problem is burn. Rather, the lack of burn. Craig Jones’s Zoo deck from Honolulu ran fourteen burn spells, with three clocking in at two damage, four more at three damage, and seven at four damage. This yielded a deck that could run in some early attacks and then, when the opponent had the poor taste to drop a Keiga, simply launch burn over the top and torch the opponent out. In contrast, while we have some very exciting burn in Standard in the form of Lightning Bolt and Burst Lightning, we don’t have enough burn to work this kind of archetype.

Also, as what turns out to be a side issue, Vines was almost never a deciding factor, or even a big player, in any games. So it goes.

At this point I returned to Gatherer – whose use I’ll describe in a lot more detail in a future piece – and reconsidered my aggro approach in light of these testing results. Early beats from small creatures were clearly good, since the non-aggro decks in this format get an awfully slow start (no one’s Mana Leaking your dudes, e.g.), but a deck that’s going to succeed needs more reach than mere burn can provide. Thus, I wanted to go for somewhat fewer early beaters, while keeping them as powerful as possible, and then to find a way to include a bit more reach.

Then I saw a card I’d been overlooking, and this new design came together very quickly.

Nayamorphic

In a perfect world, I’d be magically running four friendly-color fetches in this deck. Here we adopt the compromise solution of running four copies of Terramorphic Expanse and occasionally accepting a tapped land on the second or third turn as a fair trade for giving us super-powered one and three drops.

The card I’d previously overlooked that now gives us the third quartet of cats in Nayamorphic is Steppe Lynx. Initially, I wasn’t so keen on Steppe Lynx since, perhaps in a fit of worst-casing I decided that something that was often a 0/1 was just too awful. As it happens, this is not true in a deck that runs eight fetchlands, and I’ve been regularly beating in for four (or five, with the occasional exalting from Pridemage) on turn two. It’s majestic, and it’s far more damage than the prior Vastwood builds ever managed that early.

This deck also leverages its fetch-happiness into superpowered Knights. I admit I’d been shying away from Knight of the Reliquary in making new designs since I’ve been rocking the Knights for the past couple months and do like to switch things up. But in the current environment, and especially in a deck that sometimes has two to three lands in the graveyard as you’re casting a Knight on the third turn, it’s too good to pass up. Knight also has the pleasing advantage of letting you power up Lynx while simultaneously thinning your deck and pumping the Knight itself for the following turn.

I’ve similarly relented and gone with Bloodbraid Elf. In this deck, Elf is better than Ranger, as Ranger offers an attractive form of late-game reload, but Elf can generate immediate victory by offering itself as a hasted creature plus a free burn spell or game-ending Naya Charm.

The “reach package” has been advanced significantly by the addition of Naya Charm and Elspeth on top of the extant burn. Both tend to generate wins, even through the Baneslayer benchmark. The application of Naya Charm for a game-ending swing is obvious, but consider Elspeth powering up a Knight that likely already started at 5/5 or so. An 8/8 flying Knight beats a Baneslayer in a head-to-head fight.

I have been tremendously satisfied with Nayamorphic as my aggro deck of choice, and will be focusing on it in the weeks ahead.

Testing Ascension Control
Last week I wrote about two potential approaches to Ascension Control, one more oriented toward “do nothing” card drawing and countermagic, the other toward planeswalkers and board control. The questions here were twofold. First, is Luminarch Ascension really as much of a house as I think it should be? Second, what’s the proper host for Ascension?

The Esper Ascension Control build rapidly collapsed under an expanded testing regime. It’s decent enough against cascade and other slow control decks, but mono-red aggro and some other aggro builds slide right in under its insufficiently effective countermagic and murder it. I learned a number of useful things from the crashing and burning of the Esper variant. First, Ior Ruin Expedition is a terrible card, and is not even as good as a poor suspend spell. Second, Traumatic Visions really is nice in its dual role of early-game Sylvan Scrying and late-game counterspell. Third, Luminarch Ascension is terrible when you can’t stop someone from punching you in the face.

The Planeswalker Ascension build did rather better, standing up reasonably well across the board. It could weather an early aggro assault much of the time, and could come out on top against Baneslayer builds. However, as in many G/W decks, it often found itself tapped out with few good ways to get back into the game. This is especially critical against cascade decks, since if both decks go into topdeck mode, the cascade deck tends to win unless you have card advantage on your side as well. Given this issue and the successful application of Visions in the otherwise failed Esper variant, I decided to try a Bant approach.

I’m not recommending the following deck, but for clarity’s sake, here you go:

Bant Ascension Control (not recommended)

This is not the very first build I went with, but I fairly quickly added in Jace as a way of having durable card advantage that could help pull you into your late game plan from the mid-game. It also plays into the key discovery that came up in testing the G/W Planeswalker Ascension build, which is that the value of Ascension builds tremendously when your opponent has to decide which of your threats to stave off with their damage. Obviously, damage directed at you gains them the dual benefit of stalling your Ascension and moving them toward winning the game, but when you have one or more planeswalkers in play, they may well be giving up control by focusing on you. In testing, both the G/W and Bant variants frequently saw board positions with Ascension and one or more of Nissa and Elspeth in play, and the aggro opponent in a real bind trying to figure out how to stay in the game.

Unfortunately, this deck picks up a better mid and late game, critical against cascade and Baneslayer builds, at the expense of its early game. This proved to be a killer flaw against the fast aggro we have to expect in the new Standard, as the deck is basically dead against mono-red aggro much of the time, play or draw.

Our goal, then, is to deploy the fantastic power of Ascension as a game ender, maintain some way of staying in the game against other card-advantage laden decks, and nonetheless survive the early onslaught of hyper-speed aggro. In addition, we’d really like a way to deal with an opposing Ascension or any number of other problematic enchantments, as these are looking to be major concerns in control mirrors and against midrange decks.

Again, I’m going to show you a deck that I’m not recommending, with the recommended concept to follow. Here’s what I tried next:

Ascension Pulse, version 1 (not recommended)

The idea here is to live through the early game via Paths, then win in the late game with some combination of planeswalkers and Ascensions. The deck was reasonably sound in testing, but kept running up against its somewhat top-heavy design and clunky mana requirements (witness its need to hit WW, BBB, and GG). These two flaws in turn meant that it was also folding more often than I liked to mono-red aggro, and that meant it wasn’t good enough.

I walked away from the testing process for a while here and considered a bit of history. I do this quite a bit, and I’ll talk more about it later. In this case, I referred back to Mihara’s Gifts deck from Honolulu and how it addressed its own mana requirements. I return to my tinkering and put this together:

Ascension Pulse, version 2

This is the product of a couple of cycles of testing against Baneslayers and Goblin Guides, and it fairly neatly combines early and late game efficacy. In the earliest of the early game we have Nissa’s Chosen as a reusable speedbump, and Path followed by Pulse to deal with opposing beaters. This combination of removal also helps us against the likely spread of opposition in control decks, as we can Pulse or Path Baneslayers, and Pulse opposing Bloodwitches. I shaved some weight from the curve by removing one of each of Elspeth and Nissa and by ditching Sorin, who was cute but nearly uncastable.

The biggest addition here are the Harrows. There’s been a lot of debate about Harrow and its value as an accelerator. In this deck, Harrow is not so much about accelerating from three to five mana as it is about making sure you’re in the appropriate colors to continue into the mid and late game. As a special and helpful bonus, Harrow’s lands come into play untapped, which means that you can Harrow on turn three and still Path a Ball Lightning away on your opponent’s next turn.

The mana base here, while supporting reasonably heavy triple-color requirements, is quite resilient against Goblin Ruinblaster, which I foresee representing a significant threat to other many-color control builds that rely on a swatch of nonbasics to meet their mana needs.

This deck did well in my Red Team environment, and I’ll be using it as my control builds in the coming weeks.

Next week I’ll talk about building a sideboard, and put together prospective sideboards – and possibly revised main decks – for Nayamorphic and Ascension Unfortunately, I have to miss the first of our area PTQs this coming weekend due to business travel, so I won’t have any exciting stories of Zendikar Sealed to share until a little later on. If you’re within a couple hours of a PTQ this weekend, I encourage you to go. They’re great fun, and a good chance to find unexpected ways to use all these new cards.