I employ several different styles of playtesting for Constructed tournament preparation. My favorite is the mock tournament, a practice tournament with decks and sideboards where several players show up with what they would bring to some conjectural event to see how the various ideas play out in even a loose format. Our first couple of mock tournaments over the past few months have shown great results, with my version of Jushi Blue winning the pre-States event and Chris Pikula's B/W Homebrew winning the pre-GP Philadelphia tournament (and followed up with similar results at the appointed real big shows). The only major drawback of the mock tournament is that it only works fairly late in the preparation process, when even the playtest decks are a little fleshed out. That, and the fact that you actually have to have a critical mass of interested testers to make the thing occur to begin with.
The most fun excuse for playtesting that I've found is to battle in the Tournament Practice room on Magic Online. As fun as that is - it in fact represents an addictive joy that keeps you up until the wee hours of the morning (and by "you" I mean "me") - casual room battle is inefficient at best. For one thing, there is no guarantee that your opponents will be packing the relevant decks you want to test... For amateur preparation, this can actually give you an idea of what an unknown opponent may bring to the table, but it is probably not a plus in National Championship or Pro Tour prep. The bulk of games I like to test is based on a technique called "extremes in metagaming" which was innovated by Justin Polin when he was still a Neutral Ground regular. The idea is to attack the furthest reaches of the metagame specifically; if you can withstand or even dominate the decks at those extremes, then the decks that fall between will... necessarily fall as well.
For example, during Kamigawa Block, I tested only against Gifts Ungiven (combo/board control), White Weenie (aggression), and to a much more limited extent, Mono-Blue (true control). In the actual PTQ I won, I played against zero Gifts Ungiven decks, which sucked because it was my best matchup, but also kind of didn't suck, because Gifts was arguably the best deck of the format (and who really wants to play against the best deck every round?). I played against White Weenie three times and got matched with Black Hand four times, including Top 8, Top 4, and Top 2. The fact that I didn't strictly prepare for Black Hand wasn't a big deal... In many ways the deck was the same as White Weenie (Hand of Cruelty is the same as Hand of Honor for a U/G deck's purposes); in more important ways, for my deck, it was flat-out worse. For example, I'd much rather have Sink into Takenuma resolve against me on turn 4 than Hokori, Dust Drinker. Hokori locks all my lands while the opponent probably has a tempo advantage, whereas Sink into Takenuma merely eliminates my hand - something I can rebuild with Sensei's Divining Top - while the opponent destroys his own board position.
Similarly, for States, I tested once again against White Weenie (including vestigial versions of Boros Deck Wins) from the aggression side and Gifts Ungiven from the control. I did a little light testing against board control decks like Fungus Fire, but because I only intended to play decks with 8+ counterspells (and the deck I eventually went with had 17 starting), dedicated board control with life gain seemed horrendous; had I gone with non-permission decks, combo might have been an issue, but decks with lots of counterspells love to play against Standard Heartbeat of Spring, so that saved some work.
Going into a format like PT Hawaii, there are probably close to 20 viable, individual, decks. Unless you want to spend 50 hours a week playtesting (I personally haven't been able to spend quite that many hours since college) for two months straight, there is no way you can hit or even think up every single iteration of every single deck. For the Standard that is going to be showcased starting tomorrow, exhaustive matchup testing would require explorations of Boros, Fungus Fire, Ghazi-Glare, Sullivan Deck Wins (and Wins and Wins), B/G beatdown, BUG beatdown, Heartbeat, Jushi Blue (New York and Japanese), Critical Mass (including no, Red, and Black splashes), AnnexWildfire, Hattori-Hanzo Tron, Battle of Wits, Enduring Ideal, regular Gifts Ungiven (spice.dec), Wild Gifts, and Karsten Gifts... and that list doesn't even take into account Guildpact additions! When you include all the expected Guildpact cards, from Izzet Guildmage to Steam Vents to the simultaneously hyped and hated Orzhov suite of cards, the number of individual viable decks probably exceeds 20! (At least until the Pro Tour and a couple of weeks of Magic Online cut that number down.)
So for expediency's sake, I decided to cut that huge number down to a more manageable set for purpose of litmus testing. To get to "the next level" of serious consideration, my test decks had to perform reasonably well against the extremes first. Here is the pre-Hawaii litmus gauntlet I used: Boros Deck Wins
Michael Pohling -1st Place, 2005 New Jersey Champs
It might seem a little off to select this particular deck for the offensive wing of the litmus test, but I had very specific reasons in mind. Ghazi-Glare might kick this deck's butt, might have been the breakout of the World Championships, ultimately winning the whole shebang, but from the extremes in metagaming perspective, I had other lock- and control-oriented decks in my initial gauntlet; I don't need those elements so much as a brawling deck that can come out relentlessly, deal a lot of damage, and present a point or two of potentially scary Reach once the board has been stabilized. I wanted Boros because it was so popular at Worlds (behind only Jushi Blue in numerosity) and because I wanted to minimize building "both sides of a matchup" as much as possible - I would have had to bring my own biases into the equation if I were going to run a test G/R, for example. On top of all that, Boros was the #2 deck by population after Jushi, so I figured it was a more than adequate representative for Jitte aggression.
Pohling's Boros Deck Wins specifically is a little weird - Swiftblades main (instead of nowhere) and only two Chars - but it had a pretty special pedigree, having beaten my version of Jushi Blue to win New Jersey States (quite the opposite of repeated trials at New York States). I actually liked this deck for its low mana count... In testing, especially if you are not actually considering a deck but are just using it as a practice partner, having more spells and a low curve can have virtue. This version was very good at exposing bad mana balance and at punishing any poorly-tuned beatdown deck (for example most B/W attack decks I made), making it an ideal litmus tester, eliminating bad ideas before too much time was spent on them.Karsten Gifts
Frank Karsten - Top 8, 2005 Worlds - Standard
Just as Hari Seldon put his two Foundations at opposite ends of the galaxy, the White Weenie and Gifts Ungiven decks have been the monolithic opposites that defined and define extremes in metagaming testing since full set Kamigawa Block. The builds have changed, with White Weenie embracing Red and Gifts Ungiven moving from lock-combo to fatty-control to its latest iteration of fatty-lock, but the principles are the same: beat the deck that attacks you, beat the deck that you have to attack, and you can succeed. This gifts is absurdly more powerful than any other version because locking down the opponent's entire board (which Yosei + Greater Good does nicely) is much more powerful than just locking down creature damage with Ethereal Haze or just looping Cranial Extraction (see the whole "Hokori is scarier than Sink into Takenuma" discussion, above). For playtest purposes, Osyp removed the Last Gasp for a more versatile Mortify, which makes plenty of sense given the wider card pool and the nature of Gifts Ungiven the card.
Since States and certainly Worlds, some version of Mono-Blue has to make a first string test deck. Unlike Kamigawa Block, where Jushi Blue was a minority, if successful, anti-Gifts strategy, Standard Jushi Blue, correctly tuned, has proven successful not only as anti-Gifts, but even better as anti-beatodown. For purposes of litmus testing, I chose Marcio Carvalho's list as the initial enemy:Jushi Blue
Marcio Carvalho - Top 8, 2005 Worlds - Standard
I generally dislike Carvalho's version of this deck, but he was in the Top 8 of the World Championships, so you have to give his deck a little respect (one mark of a good playtester is to remove ego as much as possible from the deck lists... thinking "you are right" is the surest way of getting the enemy decks wrong). That said, the main things that I didn't like were 3 Mana Leak and no Remand (which makes it fairly weak against decks capable of playing second turn Hypnotic Specter), the presence of three Azami, Lady of Scrolls (a card I've never liked), and three Rewinds main (I played two Rewinds main and two side at States, but they were the worst cards in the deck and eventually left the list entirely).
Over the course of testing, Osyp Lebedowicz decided that most players would be really excited about Electrolyze, and citing especially the lack of any Cranial Extractions in the sideboards of the international Jushi Blue decks, thought most Blue Control opponents would splash Red rather than Black. I cobbled together a Red splash based on Carvalho's deck thusly:TEST DECK
I appropriated Carvalho's mana base philosophically and switched the spells to accommodate the new color. Along the way I added the fourth copies of Disrupting Shoal and Mana Leak (best counters), cutting the worthless Pithing Needles. Hideous Laughter became Pyroclasm (same thing), and Electrolyze substituted Last Gasp. Andrew Cuneo, originator of the Draw-Go strategy nearly ten years ago, pointed out that Tidings is just a much better card than Azami; I think it is hard, at best, to disagree. Niv-Mizzet took the place of Black finisher Ink-Eyes, Servant of Oni. We all agreed that Keiga, the Tide Star is better in a Wildfire environment than the fragile Clouded Mirror of Victory; that swap was dictated more by the environment than the natural synergies you might want in a Blue control deck.
Pat Chapin sent me a U/R deck, so I eventually switched to his version for litmus testing. I try to keep myself out of playtest decks as much as possible because I have very entrenched opinions that other players don't necessarily share; whether or not I am right is irrelevant because the goal is never to beat the best decks, but merely to beat the decks that show up. Here's Pat's version of U/R Control:U/R Control
Balancing mana bases with the Karoo (Golgari Rot Farm et al) cycle is very interesting. Karoos not only produce double mana, they ensure your next land drop, kind of like a Kodama's Reach. As such, playing with them presents new challenges both during deck construction and the tactical execution. Pat's deck has lots of short-term card drawing and four Izzet Signets, but it was still easy prey for our version of AnnexWildfire due to having only 22 lands.
For a long time I wanted AnnexWildfire to be the best deck. Brian David-Marshall thought up a cool new card in a totally different color for the deck, and I began a torrid love affair with the re-tooled sixty, which was sort of an American Dream - nine American Dreams really - for a victory in Honolulu.American Dreams
The structure is based much more closely on Masashiro Kuroda's winner from The Finals than the original Eminent Domain that took the Wisconsin State Championship in the hands of Adrian Sullivan. Kuroda's major innovations were to replace the Spectral Searchlights with all two-mana accelerators and remove all the Karoos (Dimir Aqueducts), while tuning the numbers more towards a control aware format rather than random damage sources.
This deck was one of the best decks for beating Karsten Gifts, and was much better than any version of AnnexWildfire in the past at beating beatdown decks. The inclusion of synergistic answers like Faith's Fetters did multiple things at the same time. Not only is Faith's Fetters a sort of super combo with Dream Leash and Confiscate via Three Dreams (the aforementioned brain child of BDM), but the versatility and life gain it represents shore up many of AnnexWildfire's old problems. For example, Adrian's original deck was comparatively weak against Boros Deck Wins, and PT Champ Shu Komoro's list from The Finals had to resort to playing Pithing Needle.
Three Dreams is just absurd when it works. Sometimes you only get one Aura with it, but in those games you are obviously so far ahead it hardly matters. Other times, and much more commonly, you play a "regular" game with your two-for-ones the first couple of turns... and then get the scoop as soon as you reveal three more two-for-ones (the "six for one"). It's the kind of card to which a designer can really grow attached, you know, emotionally (never good).
Ultimately, the Dreams deck - which I liked perhaps too much during testing - had two big problems. The first was structural, and we never reconciled how Kuroda was able to get past it... Mighty and synergistic AnnexWildfire at once has not enough lands... but too much midgame mana. Sometimes it doesn't hit its third land drop, yet it can be flooded with Signets. This couldn't be addressed by brining faux kill card Searchlight back in... The whole incentive of the deck is turn 3 Annex! Even when it hits second turn Signet and third turn land, the three color nature of the deck, combining somehow with Miren, Mikokoro, and in some lists Boseiju, means that it inexplicably can't cast Annex. This is depressing, but not necessarily lethal. The other issue is that AnnexWildfire can't buy a game against a tuned version of Blue Control, that is the re-tooled original.
My Blue stand-in went from Carvalho's deck, to a Red version of Carvalho's deck, to Chapin's Red splash, to, ultimately the Blue deck I might have actually considered playing were I qualified:Flores Blue
This is basically my States deck, with 4 Keiga and 2 Meloku instead of 3 and 3 (due to the now-Wildfire format), and Tidings in place of Rewind. It's a pretty good test control deck, and excruciatingly difficult to beat with mid-range decks or pure beatdown. That said, I expect that people will be ready for the Jushi strategy come Hawaii. I never really got why the Japanese Azami deck tipped and mine lost popularity over the past four months or so... As far as I can tell, the defining element of, say, Carvalho's deck is the ability to side into Boseiju + Persecute in the mirror while trading so many of the incentives that made the New York States deck so effective for ITS time. For example, the Dutch version with Jitte and Dark Confidant commanded card drawing past the relatively fragile Jushi Apprentice, but lost percentage against the relatively easy beatdown matchups. I still don't get decks that don't play Remand.
To give you an idea of process from the other side, I decided pretty early that the best decks for the Hawaii format would be proactive, even if they were control oriented. That means that they have to be doing big and exciting things, even if those things are aimed at controlling the board. One of the first interactions I came up with was this one:Flores Purification
I remembered back to two Standard seasons ago, when Wayfarer control with Cloudposts and Temple of the False God was one of the best decks . "Why can't it be that way again?" I thought, especially as this generation of Cloudposts conveniently keeps the Wayfarer running by bouncing up existing lands rather than spreading the mana base.
The idea of playing Razia's Purification with three Karoos in play was too awesome to miss... I could even use Boseiju mana, float a White, and run out a post-Purification Wayfarer or Divining Top! How exciting! Unfortunately the deck was less super awesome than the initial idea. "Razia's Purification is no Wildfire," said Osyp. The deck would often be in a position where Purification was the only anti-swarm card readily available, and it would leave clock, land, and Jitte in play. In addition, Cuneo pointed out that Faith's Fetters and Purification is a terrible combo, which it, quite unavoidably, is. However the death of a terrible deck made way for a halfway decent implementation:Flores B/W
Somehow, I don't think this is what R&D had in mind when they minted the Orzhov.
I really liked this deck, because if you could get one use of Weathered Wayfarer off, you were generally in great shape against a control deck (they don't like it when you use Mikokoro, untap, and show them Persecute behind a Boseiju) and all the spot elimination and life gain was great against beatdown decks; B/W Wayfarer ran about 7/10 against both the Boros test deck and whatever Blue deck I was running that week. Unfortunately, the B/W was decidedly lackluster when facing Karsten Gifts, invalidating it from the beginning. I still think Wayfarer control is a reasonable archetype for, say, team Standard play, because its proactive plan is so powerful (you can't play Karsten Gifts every round), and because it doesn't generate a lot of conflicts (but I guess the Pro Tour will show us exactly what kind of B/W is the best way to spend a team's Godless Shrines).
That said, at the mock tournament we held last week, I did see my teammate Tim McKenna empty an opponent's hand, untap, and play Angel of Despair only to lose to a topdecked Overwhelming Intellect (you can go ahead and wince)... so I guess anything is possible. B/W Wayfarer is extremely fun to play to say the least (I like searching, shuffling, and drawing extra cards), quite different (Wayfarer much?), possessed of the best Sensei's Divining Top engine in Standard (about half the time)... and yet somehow desperate enough to play Dimir Aqueduct (I got nothin'). On the other hand, you can pretty much chalk one up in the win column every time your unknowing opponent cracks a Sakura-Tribe Elder... Unless it is dropping Greater Good into Morning Star, at which point you had better have a Mortify in hand or be playing a sideboarded game.
While the proactive elements of this deck are pretty solid (always having the best card in play after turn 6, Boseiju into Persecute with great regularity, Wayfarer activations in general), it is nevertheless a showcase for potentially inefficient deck design given an open format. The two plus two split between the Extractions and Persecutes is a noticeable one, for one of my decks at least. I generally play one or three Persecutes depending on my tutor availability, and almost unflinchingly sideboard all three of my played Extractions; the two plus two split - as with most such in Constructed deck lists - is indicative of not knowing which will be good in a format in general. Not surprisingly, this B/W probably would have packed one or more of both in the side, along with sundry other defensive or disruptive elements, had it gotten to the big show. No offensively transformative, um, anything.
Ultimately, the goal of this article, other than to list a couple of somewhat unusual lists for Pro Tour decks, was to give you an idea of how one deck designer approaches the task of building decks for a fairly large format, and explores new ideas when expecting highly-varied opposition. Picking decks at the far ends of a predicted metagame as initial punching bags - or in some cases, bad boyfriends - is a good way to save effort in the alpha stages of design, and can give you an idea of which decks are really worth exploring without expending too much time on what will ultimately be irrelevant matchups.
Plus, Constructed playtesting is more fun than drafting. Just ask Osyp if you don't believe me.
Enjoy the Honolulu coverage this weekend... I can't wait to get back in the booth!