Deep Analysis – Rapid Prototyping

Read Richard Feldman every Thursday... at StarCityGames.com!
Thursday, April 3rd – Tournaments come and tournaments go, but one thing always remains the same: we never have enough time to fully prepare. Today’s Deep Analysis sees Richard Feldman explore some of the techniques employed to help manage our time wisely. By using and developing his U/G strategy from last week, he shows us deck design in action.

Writers often speak about time management as it pertains to in-game clock management or pre-tournament preparation rituals, but time management is not a concept that is often applied to deckbuilding. (Well, with the notable exception of Adrian Sullivan recent article).

In fact, when figuring out how much time to devote to a certain aspect of building or tweaking a deck, I recall seasons when I could only identify two states of time: “plenty of time left to try that,” and “no time left to try that.” In other words, I felt that I was either in the early experimental stages before a tournament, with scads of time to spare, or in full-blown Oh God There’s Only A Week Left And I Don’t Have A Sideboard Yet mode, as though somewhere in the middle of the preparation period the switch would just flip from one mode to the other.

More than likely, I wouldn’t have felt myself in Help Help Help mode as often as I did if I had better spent my time back when I had a lot of it. I remember back in Onslaught-Mirrodin Standard, I once ran an experimental deck through a sixty-game gauntlet just to make sure I wasn’t missing anything, when I really should have given up after it lost nearly all of the first twenty against the format’s two top decks.

For an example of how an individual deckbuilding decision can be put in the context of time management, I will turn to the Blue/Green control deck I arrived at last week. When I last talked about that deck, I was concerned about its matchup against Mono-Red, which is why I had included maindeck Bottle Gnomes and an Academy Ruins with which to fetch via Tolaria West. My hope was that I could throw everything I had at them on defense, keeping Tarmogoyfs and Spectral Forces back to preserve my life total, while setting up Gnomes plus Ruins to put myself out of burn range so I could assume an offensive role and finish the game.

I was naturally skeptical that this plan would pan out in practice, so I started playing test games to gauge its effectiveness. In one game, I led with Tarmogoyf, then Bottle Gnomes, then Transmute for Tolaria West, and then Spectral Force. That’s one of the best draws I could hope for with this setup against a Red deck, and I still didn’t get there. All it took was Magus of the Scroll and Keldon Megaliths to counteract my Gnomes each turn, while my opponent built up resources to swarm around my Force. Even if he hadn’t had the Megaliths, and I was netting one life per turn, he still would have overcome my defense in due time. In another game, I again had the Ruins and Gnomes, my opponent once again had a decent draw, and he simply toasted me before I could execute the loop more than once.

At this point, I had only played about five games, but was faced with the choice of whether or not to keep testing. Technically, is a two-game result significantly significant? Surely not. Is it enough information on which to decide a life-or-death decision? Also no. But is it enough to justify aborting a playtest session? That’s another story.

Let’s say I prove it the thorough way. I keep running games until I come up with thirty examples in which I got the Gnome recursion going. By the end of it, I can confidently say – one way or the other – that I either can or cannot depend on that plan to carry me against Red decks.

Now let’s say that I decide that what I learn is that I really can’t beat Mono-Red with this configuration, looping Gnomes or no. Time to try something else.

Maybe I just need to add some Spike Feeders. Let’s see, I’ll just cut the Spectral Forces for some Feeder action, play another couple dozen games… hmm, still looks unfavorable. Well, maybe it was that I replaced Spectral Force and not, say, the Cloudthreshers. Still unfavorable? Man…

I could go on all day. Even if I did find a configuration that worked against Mono-Red, I’d have spent so much time discovering it that I would have severely decreased my remaining time in which to test the other matchups before the relevant tournament. Even worse, what if I ultimately discovered that I wanted to play another deck altogether? How much time would I have left to practice with and tune that deck, having spent so much time definitively proving that the previous deck was off-base?

Instead of spending all this time searching for measurable, technically precise back-up to prove or disprove my first hypothesis that Gnomes and Ruins weren’t enough, followed by even more extensive testing to find a replacement, I could have followed through with my initial impression. Although doing this would technically increase the chance that I would make the wrong decision, the time cost that I would spend proving it out completely would almost certainly outweigh the benefits of knowing the answer with more certainty. After all, if choosing differently really had that big of an impact on the deck’s potential, wouldn’t I have a good chance of reconsidering it later once I had more information?

Sure, in an ideal setting, I’d test out everything thoroughly – but no one has that kind of time; there are far too many decisions that come up in testing, and one slight metagame shift can invalidate them all and make you start over. If I don’t have enough time between now and my next tournament to pore over every single deckbuilding decision, then my two options are to make only a few decisions, but be very sure that I have evaluated them correctly (while running out of time to make rest of the decisions)… or to make as many informed decisions as I can, devoting more time to accurately solving the problems whose answers I am not sure I know (such as how the Reveillark matchup goes – at this point I haven’t even considered it!), and less time to those questions I feel confident I can answer already with little experimentation.

Back to the U/G deck. I ended up deciding, based on the two games in which my strong anti-Red draw (Bottle Gnomes and Academy Ruins included) got crushed by the Red deck’s average draw, that I needed to seriously reconsider some of my deck’s fundamentals.

Before making any further changes, though, I decided to try it against Reveillark, the other matchup I was concerned about, to see where its weaknesses in that matchup lay. Unfortunately, like Mono-Red, that matchup also seemed rocky. In talking with Zac about the U/G deck, he expressed concern about Lark’s many bounce effects and that they would be able to kill me with their random dorks before I could stick a Force or Thresher. As it turned out, I found myself getting comboed out more often than not; since I had no way to disrupt their graveyard, it didn’t much matter that I could keep my life total afloat, because ultimately they would just find the last piece of the combo and ruin me in one big turn.

With that in mind, I set off to…

… wait! What’s this? Fish? Damn near every Standard article on StarCityGames.com last week? Really? Guess that’s one more to add to the gauntlet; if it’s not big now, it may well be soon, and it would be dangerous to ignore it.

So I try out the matchup against Adrian Sullivan build, and it turns out the Islandwalking is really big. One Tarmogoyf will hold off a lot of fish until they bounce it (Adrian’s has an actual fish for a bounce effect, though other builds rock Cryptic Command instead), but one Island on my side of the table means my Goyf has to race – and he’s really not up to the task by himself.

In thinking about the matchups against Mono-Red, Reveillark, Fish, and Faeries, I realized that Pyroclasm was a pretty fantastic removal spell for a deck like this. It was more awesome against some of those matchups than others, but it was also quite exciting against things like Elves, Goblins, and Kithkin, and even non-dead against G/R Big Mana because of Siege-Gang Commander.

It was at this point that I tried one of the many crazy deckbuilding ideas I have: splash Pyroclasm into the U/G deck, while removing all the Islands. Hey, who cares if they Islandwalk if I have no Islands?

The manabase ended up here:

4 Tolaria West
4 Yavimaya Coast
4 Shivan Reef
6 Snow-Covered Forest
1 Boreal Shelf
1 Snow-Covered Island
1 Highland Weald
1 Urza’s Factory
1 Academy Ruins

4 Search for Tomorrow
2 Into the North (replacing Mind Stone from the previous list)

While coming up with this mess, I decided that it would be okay if I played one basic Island (for Search for Tomorrow’s sake), as long as I didn’t play it when paired against Islandwalkers. I snuck in a Boreal Shelf (essentially a worse Island, in most cases) so that I could fetch it with Into the North against Fish.

Now, it only took me about two minutes to draw up this manabase. How long does it take me to poke holes in it? Even less. For one, I now have eight painlands and six comes-in-tapped lands, up from four and four, respectively, and here I was trying to improve my matchup against Mono-Red. Second, I’ve made Cryptic Command’s triple-Blue casting cost a pipe dream. Third, I’ve introduced scenarios in which I cannot fetch out a second Blue source (let alone a third) with Search for Tomorrow because I’ve already drawn the one Island.

I replaced Bottle Gnomes with Pyroclasm to see if it would at least help against the Fish, and it did – quite a bit, in fact. However, I still had an overarching curve problem. Too many games would see them leading with a fish (well, okay, a Changeling), another fish, and then another fish or two – while my first business spell of the game was turn 3 Cryptic Command, while still waiting for Ancestral to fire off. Worse, I often couldn’t even cast the Cryptic Command on turn 3 because my manabase was so Blue-light.

Again, I could have plugged away at this build, proving beyond a shadow of a doubt that it was not the way to go with the deck… but I didn’t. In the interest of time, I went with the Rapid Prototyping approach, quickly shifting to an alternate approach to the implementation of the strategies that worked best about this build, to see if I could stumble upon something better. I scrapped the U/G direction altogether, backed up to the fork in the road where I chose to investigate that build, and took a different path.

Zac suggested that Primal Command might be a good solution to both the Red decks and to Reveillark (with the “gain seven life” and “Feldon’s Cane you” modes, respectively), and I wondered how it might work as a maindeck card. However, the thought of mashing Primal Command, Cryptic Command, Spectral Force, and Cloudthresher into the same deck struck me as practically suicidal for a deck that already needed to lower its mana curve.

Well, if Cloudthresher is important and Spectral Force is important and I want to try out Primal Command… sounds like Cryptic Command is, surprisingly, getting the boot for this build. That leaves Ancestral, Pact of Negation, Rune Snag, and Tolaria West as the only Blue cards left in the deck. What if I tried some other color? What about… say, black?

4 Graven Cairns

4 Terramorphic Expanse

4 Spinerock Knoll

1 Highland Weald

1 Tresserhorn Sinks

1 Snow-Covered Swamp

2 Llanowar Wastes

4 Snow-Covered Forest

1 Snow-Covered Mountain

4 Search for Tomorrow

3 Into the North

4 Distress

4 Thoughtseize

4 Tarmogoyf

4 Spectral Force

3 Cloudthresher

4 Pyroclasm

4 Primal Command

4 Harmonize

Here, Ancestral is replaced by Harmonize (which doesn’t help with the mana curve problem, but Harmonize does typically yield its cards faster than Ancestral in this deck), Primal Command by Cryptic Command, Rune Snag by Thoughtseize, and the Pact of Negation / Summoner’s Pact / Ana Battlemage / Seedborn Muse suite with the straightforward 4 Distress. I also cut a Thresher for an extra Into the North, as I did not feel the full four were necessary if I was already running a set of Pyroclasms.

This deck most notably misses out on Cryptic Command, Rune Snag, Ancestral, and Pact of Negation, but the manabase is much better. Besides getting to play the amazing Graven Cairns (I nearly exploded with joy when I saw the news about Shadowmoor), this deck is in a fantastic position to abuse Spinerock Knoll. Between Spectral Force, Cloudthresher, and four Primal Commands with which to fetch them, the odds that this deck will connect for seven points of damage in one turn are quite high. Finally, with no Blue component, there are obviously no Islands with which to subvert Tarmogoyfs that would otherwise stand in the way of those pesky fish.

Again, I played some test games to get a feel for the deck.

I immediately identified some new problems. For one, Tarmogoyf is commonly Elvish Warrior in this deck; every spell in the deck is either a Sorcery or a Creature! Getting to 3/4 almost always requires some contribution on the opponent’s part, and higher than that did not happen often. For another, Distress and Thoughtseize were just not the same as countermagic. We’ve known for years that countermagic is more effective at gaining tempo than hand destruction is, as only the former requires that the opponent expend mana before losing his spell. This was relevant in every matchup; if this deck has such trouble keeping up because of its high mana curve, is it any surprise that the tempo loss from converting countermagic into discard magic was problematic as well?

The Pyroclasms were awesome, but I really wanted to keep that countermagic around. I wanted Blue back, and considered compromising on Cryptic Command for the sake of the mana curve. I wondered what the deck might look like if I cut out the Commands for Delays or Remove Souls.

Remembering the problems I had had against Fish, I wondered if I could try another no-Island (or rather, one-Island) configuration now that Cryptic Command’s triple Blue was out of the equation. Though Pyroclasm was as awesome as I’d hoped it would be, I also wanted to try a different splash color – just to see where it led me. I tried White, and ended up here.

4 Tolaria West

4 Yavimaya Coast

6 Snow-Covered Forest

1 Snow-Covered Island

4 Adarkar Wastes

1 Boreal Shelf

1 Snow-Covered Plains

1 Arctic Flats

1 Urza’s Factory

4 Search for Tomorrow

3 Into the North

4 Tarmogoyf

3 Cloudthresher

4 Spectral Force

4 Ancestral Vision

4 Rune Snag

4 Remove Soul

4 Glittering Wish

2 Pact of Negation

1 Tormod’s Crypt

The advantage of White as a third color is that it is allied with both Blue and Green, meaning I can play both Boreal Shelf and Arctic Flats. If I had chosen Black, I could have gotten an allied color with Blue – which would have made me happy, as Blue-allied colors are enablers my no-Island addiction – but having just experimented with Black in the B/G/R deck, I wasn’t excited about what it had to offer my overall strategy.

In this prototype, I tried Glittering Wish where I would have otherwise run Primal Command or perhaps Pyroclasm. However, a quick look at the targets it could fetch revealed that my options were pretty limited. Teferi’s Moat was exciting against Elves, Goblins, and G/R, pretty irrelevant against Faeries and Reveillark, and reasonable against Mono-Red and Fish depending on how much bounce and countermagic their build incorporated. Saffi was a cute Wish target to combo with Cloudthresher, Gaddock Teeg, and Mystic Enforcer seemed unexciting, Mystic Snake looked a bit expensive, and Vanish into Memory struck me as too situational for even a sideboard Wish slot.

If not for its uselessness against Faeries and Reveillark, two of the matchups I was most concerned with beating, I probably would have gone ahead and maindecked the Moat instead of bothering with Glittering Wish. Unfortunately, none of the other Wish targets available to me seemed any good against Lark or Faeries either (okay, except maybe Saffi into Thresher against Faeries), so it looked like I was on to another prototype already – though the U/G/w build did get bonus points for its sideboard options of Forge[/author]-Tender”]Burrenton [author name="Forge"]Forge[/author]-Tender and Mana Tithe.

Like the original U/G/r, one-Island, 4 Cryptic Command build, I turned down this build before it even made it to the playtest arena.


By now I’ve gone through a lot of different builds with this deck. Most have been rejected, ending up back on the shelf before I’d even played fifty games with them, but each time I have learned about what works well and what needs work with this archetype.

Pyroclasm is fantastic in this deck. I really want to play it. The countermagic is also awesome; I really want that, too, even though a G/B/R configuration lets me play Graven Cairns. Ancestral is also way better than Harmonize, and another strong argument for keeping the Blue component. Bizarrely, I want to play Primal Command even more than I want to play Cryptic Command.

If I have Pyroclasm, I can probably get away with only two Cloudthreshers and still do well against Faeries – and reducing my Cloudthresher count (at the top of my curve) lets me cut an Into the North, as well, decreasing my chances of mana flood. Finally, I am excited by the prospects of playing Forge[/author]-Tender”]Burrenton [author name="Forge"]Forge[/author]-Tender and Teferi’s Moat, but not so much that I want to maindeck either card.

All these prototypes have led me to one last build, that I feel comfortable running through its paces in a real gauntlet.

I’ve cut out a lot of the original deck’s slowest elements. Gone are all the one-of Summoner’s Pact targets, along with the Pact itself. Urza’s Factory has been cut in order to make the mana more consistent (I still have yet to activate that card in this deck), and Academy Ruins went out with the Bottle Gnomes. I’ve fit in the full four copies of both Pyroclasm and Primal Command, two cards that were very strong in testing the various iterations of the deck.

The top end of the deck’s mana curve has decreased from four Cloudthresher (he’s a six-drop against most everything but Faeries and sometimes Reveillark), one Ana Battlemage, one Seedborn Muse, two different Pacts, and a full set of both Spectral Force and Cryptic Command, to a trimmer configuration of one Pact, two Thresher, and four each of Spectral Force and Primal Command. The remainder of the deck is all two-drops and lower, ranging from Ancestral and Search for Tomorrow on turn 1 to any of Into the North, Rune Snag, Remove Soul, or Pyroclasm on turns 2 and later.

It seems strange to have such a wide chasm between the two-mana and five-mana marks (with the exception of transmuting Tolaria West), but the eighteen “two-mana” cards (Search for Tomorrow nets out to costing two if it resolves) mean that I will often be able to maximize my mana usage by simply casting two spells when given four mana to work with.

Finally, this setup gives me access to “nine” (counting Search and Into the North, but not Tolaria West, which can technically fix mana as a last resort) White sources, with which I can access what I perceive will be an excellent sideboard card in Teferi’s Moat. (I opted for Dragon’s Claw instead of Forge[/author]-Tender”]Burrenton [author name="Forge"]Forge[/author]-Tender, as I’ve heard it is more effective against Mono-Red, and can imagine that this is especially true when I cannot consistently turn 1 him.) In the course of accomplishing this splash, I managed to arrive at yet another one-Island build of the deck.

Notice how similar this manabase is to the U/G/r configuration I mocked earlier in the article – and yet how this build has eliminated nearly all of the problems that made me deride the former as a “mess.” I complained that that manabase had too much pain against Red decks, not enough Blue for Cryptic Command, and introduced the possibility that Search for Tomorrow could not be used to locate a second Blue source. While this manabase does indeed have more pain, I’ve halved the number of Cloudthreshers and upgraded the three Bottle Gnomes to the massively Red-hating Primal Commands. Cryptic Command’s triple-Blue is a non-issue because it has been replaced with a two-mana counter, which also helps ameliorate the previous concern. Only the third complaint – the occasional Search for Tomorrow blip – has not been addressed, and it was by far the least relevant.

Assuming this proves to be the core around which I take this deck to its final build, I’ll have this rapid prototyping process to thank for it. If anyone can honestly tell me that you looked at last week’s U/G build and thought, “I bet next week this evolves into one-Island U/G/r/w splashing only Pyroclasm main and Teferi’s Moat in the board, with Primal Command and Remove Soul instead of Cryptic Command and the Summoner’s Pact package,” you get a cookie.

Then again, stranger things have happened; if you recall from the last article, this all started with Ajani Goldmane and my inability to resolve RTFC targeting Molten Disaster. Believe it or not, though, this is the messy process by which many decks are born. You find major weaknesses (Mono-Red and Reveillark, anyone?), the metagame changes (Fish, anyone?), and you roll with the punches. You try a lot of crazy card choices and builds, identify and discard the unprofitable ones as quickly as you can, and use what you learn about your deck’s strategy along the way to strengthen the final core.

So remember, if you ever find yourself busting your chops to definitively prove a decision is correct or incorrect, stop and ask yourself – is this the best use of my time? While it’s obviously important to inform your decisions with as much data as you can reasonably gather, the key is being careful not to cross over from reasonable gathering to the land of sharply diminishing returns. None of us has unlimited time to work with, and you might be surprised how much more you can learn about your deck by simply choosing a path and acting on it than by spending time evaluating one decision beyond a shadow of a doubt.

See you next week.

Richard Feldman

Team :S

[email protected]