Friday, November 21, 2014

Cube Design: A Series of Shifting Archetypes

    Cube design has evolved significantly since I built my first cube in 2008.  At the time, the format was described to me as "all the best cards".  I went through all of the cube lists I could find, and if I owned a card, and someone else was cubing with it, it went into my cube.  That strategy didn't last very long, however, and I have slowly tweaked my cube over the past six years to create an environment where every card has a purpose.  Over that time the goal of my cube has changed significantly.  Instead of trying to play all the best cards, I'm trying to create a fun drafting environment that constantly provides new challenges for my playgroup.  As a result, I tend to view my cube as the sum of it's parts, instead of as a collection of individual powerful cards.

    Typical cubes have some elements of aggro, midrange, and control represented in each color.  The most basic cube might give each color an identity, and then you can mix and match the five color identities to create ten to fifteen different strategies.  Cube design, though, is much deeper than that.  Cube designers have the ability to shape the way each strategy plays out.  Even for a strategy as simple as red aggro, there are decisions to make.  One cube might focus on one drops, playing even questionable cards like Reckless Waif or Tattermunge Maniac.  Another might focus more on the burn spells in the deck, playing cards like the often inefficient but occasionally powerful Brimstone Volley.  There are a surprising number of ways to build even the most straightforward strategies, but the choices don't stop there.

Managing The Available Archetypes


    That power can be expanded to control over distinct archetypes.  I use the term archetype to broadly refer to any specific strategy being supported by the cards in the cube, whether it's something as specific as a combo deck (Splinter Twin) or as broad as a color strategy (Green Aggro).  I lump archetypes into three classes:
  1. Archetypes where all cards are good on their own
  2. Archetypes with one or two build-around cards, where other cards are good on their own
  3. Archetypes with cards that are only good in that archetype, but that are powerful when assembled together

Class 1 - Independently Good Pieces


    Reanimator is often one of the first archetypes beyond aggro, midrange, and control that cube designers run into.  It is one of the easiest combo decks to support in a cube, and most of the cards required to support it are good on their own in a variety of decks.  Discard outlets like Liliana of the Veil or Looter il-Kor see play even when you aren't trying to discard.  Reanimation spells have value in midrange attrition decks.  Reanimation targets can be selected so that ramp decks can easily cast them.  New cube designers are drawn to Reanimator because it's fun, interesting, and simple to add.

    Unfortunately there aren't nearly as many interesting archetypes in this class as cube designers might like.  It's rare that the key combo piece in the archetype (Reanimate, for example) is a perfectly fine card in an average midrange deck.  Still, there are several archetypes available in this class if you want to explore them (for example, some variations of Tokens).

Class 2 - Building Around a Card or Cards


    Wildfire is the perfect example of a build-around archetype.  In order to build a good Wildfire deck, you need three things: Artifact ramp, Creatures or Planeswalkers that don't die to Wildfire, and Wildfire effects.  Mana rocks, big creatures, and planeswalkers will always be good in a variety of decks, so the only fringe cards that get included in the cube in order to support the archetype are the Wildfire effects themselves.

    The quantity of effects you include varies depending on how you want to support the archetype.  If you only include Wildfire, the deck may come together when someone sees it in pack 1, but it's unlikely to get built if Wildfire happens to be in pack 3 because drafters can't draft early expecting to see it.  On the other hand, if you include Wildfire, Burning of Xinye, and Destructive Force, you're increasing the chances of one of the three being in pack 1 and encouraging drafters to pick it while simultaneously making it easier to pick artifact mana early planning to pick up a Wildfire effect later.  With multiple copies in the draft, it's far easier to count on seeing one.

    Increasing the density of the build-around card also increasing the quality of the deck once drafted, which increases the desire of drafters to play an offbeat strategy.  If the rewards aren't there for committing to something beyond a "good stuff" deck, people won't be incentivized to try something new.

Class 3 - Putting It All Together


    Storm is an oft-maligned archetype that tends to polarize the cube community.  Unlike the archetypes from the first two classes, storm requires a lot of cards that are only good in the storm deck.  Like the second class, you need the build-around card for the deck to work (Mind's Desire, Tendrils of Agony), but the support cards (like Heartbeat of Spring or Lotus Petal) are often useless outside of the deck as well.  This has two effects: The storm deck can be very difficult to draft effectively, and the draft can get cluttered with cards that most decks don't want.  The upside of this is that when the deck comes together it can be very exciting and create the type of games that many cube designers want to encourage.

    Storm is probably the most extreme example of this approach.  Another example is the more popular Pox deck, which has otherwise unplayable cards like Reassembling Skeleton, but requires less space in the cube than Storm does.  Finding the right balance that makes an archetype in this category draftable but doesn't generate constant 15th pick cards is the most difficult part of including one of these archetypes.


Finding the Lynchpins


    All three classes of archetypes influence the other cards you choose to include in your cube.  A cube that wants to encourage reanimation, for example, might choose to play Terastodon over Worldspine Wurm as a high end green creature, since Worldspine Wurm can only support the ramp deck and Terastodon can support Reanimator as well.  A cube that wants to play Sneak Attack might play Sundering Titan over Inkwell Leviathan as a Tinker target, since ETB effects are more desirable if you don't plan for your creature to stick around.  The key to making your cube function as a cohesive whole once you've included many disparate archetypes is finding the cards like this that support multiple archetypes.  Too many cards that are only good in one deck will find their way to the 15th pick in the pack if an archetype goes undrafted.

    As an example, one popular archetype these days is Red/White Tokens.  In recent sets red has gotten several incredible token enablers (Ogre Battledriver, Purphoros, God of the Forge), and those cards get paired with white's efficient token makers to create a unique deck that gives red some identity beyond aggro.  A cube that is trying to support Storm might decide to also support Tokens in order to give Empty the Warrens multiple decks to shine in.  Similarly, a cube trying to support Pox might also support Tokens in order to find room for Goblin Bombardment.  Having multiple decks that support some of the narrower cards can lead to more drafters competing for them, which prevents them from being in the way during the draft.

    The interaction between niche archetypes and the more basic strategies can help archetypes feel different every time you include them based on how the cube around them is configured.  A cube that supports Pox and Land Recursion (Life from the Loam, Crucible of Worlds) is going to find the Pox deck playing very differently than a cube that supports Pox and Tokens.  The first build will find Pox decks focusing more on recurring their own resources to stay ahead, while the second build will focus on generating multiple permanents for a single card to accomplish the same goal.  The first build will find Pox decks leaning more towards green as an auxiliary color while the second might find them leaning towards white or red.  By mixing up the environment every time you play an archetype, you can keep even old strategies appearing fresh.


 Changing Archetypes Regularly


    Of course, mixing up the environment only works if you're not playing the same archetypes all the time.  Some of the more linear archetypes have the downside of a fairly repetitive draft experience.  Drafters can get tired of there always being a Wildfire or Tinker deck they have to beat.  Drafters who don't like one of the archetypes that requires lots of support can get tired of seeing so many cards they don't want in packs.  It's important to recognize this when it happens and work to keep your cube from getting stale.
    The best way to resolve this problem is to constantly rotate archetypes in and out, tweaking them each time so they look different when they come back.  Since the cube environment is constantly changing, a strategy that wasn't good in it's original iteration might become much stronger the second time around.  Often new cards will come out that change the way a deck plays or make previously bad cards better in context.
    By constantly trying new things, you also give yourself permission to fail.  Sometime an archetype you're trying to support just doesn't work out for your cube (I'm looking at you Green Aggro).  It's important for drafters to know that you're willing to give up on things that aren't working.  If it took you far too long to cut the last offbeat strategy that didn't work, your drafters will be less excited about trying your newest one.  When you try new archetypes, don't get so caught up in the excitement about adding them that you can't see if they aren't working.

Signaling


    Changing the supported archetypes in your cube on a regular basis does have it's downsides.  If your playgroup shifts regularly and you don't have the same drafters every time, or if you have drafters who don't keep up with your changes regularly, a constantly changing cube can be a recipe for a disastrous draft.  For example, my cube used to push Reanimator pretty strongly..  After the deck dominated for several months, my playgroup decided to back off the strategy for a while and cut the two best reanimation targets while keeping only three reanimation spells.  Reanimate was still a perfectly good card to play in midrange, but it wasn't part of a turn 3 combo deck like it had been.  A player who doesn't draft with us often came, opened up pack 1, and first picked Reanimate in an attempt to force a previously strong archetype.  His deck ended up playable, but nowhere near the power level he expected at the beginning of the draft.

    This is a cautionary tale because experienced cube drafters have expectations based on what they see in the packs.  If pack 1 contains Sundering Titan, it's reasonable to expect that Tinker is in the cube.  Seeing Splinter Twin can lead to the assumption that the combo is being pushed and that there are at least four creatures to be drafted that win instantly (while seeing Kiki-Jiki lets you know the combo is probably supported, but doesn't tell you how much).  Even something as innocuous as seeing Thran Dynamo could suggest that a big artifact ramp deck is supported, since Thran Dynamo is far enough down the list of quality mana rocks that it's inclusion suggests there are a significant number of them in the cube.  As you make changes to the archetypes you support, make sure to consider the archetypes you aren't supporting anymore, and look to see if there are any cards that send mixed signals.



Moving Forward


    I'm planning to write a series of spotlight articles on different cube archetypes, including lessons learned from my attempts at each one.  Some of the attempts have gone better than others, but I think there is much to learn even from the archetypes that didn't work out for my playgroup.  If there are any archetypes in particular you want to read about, let me know.  If I tried that archetype, I'll bump it to the front of the list, and if I haven't, maybe I'll give it a shot in my next update.  I look forward to trying some new things and writing about them as I go.

Monday, November 17, 2014

Tri-Color cards - A Practical Experiment

Background


    Recently I was inspired by this post by Channel Fireball's Jason Waddell to give tri-color cards another attempt in my cube.  Like many cube designers I had slowly eliminated all tri-color cards from my cube over the past several years as I tuned and refined my list, cutting the cards that were commonly picked last.  Still, I was inspired by Jason's ideas on how to increase the chances of cards being relevant, and with all the sweet new goodies from Khans block I figured I'd give it a shot.
    I'll be assuming you've read Jason's article in my comments below.

Tri-Color Options


    The first thing I needed to decide was which 5 Tri-Color factions I wanted to support.  I'm no mathematician, but I believe I figured out all of the available options.  Aside from the obvious options of playing just the five shards or the five wedges, there are ten other configurations available. 


    I started off with the desire to play both Temur and Grixis, because I wanted to include Nicol Bolas, Planeswalker, and Maelstrom Wanderer.  There are only two combinations in the list above that include both of those factions.  I settled on the version in the top left: Bant, Grixis, Abzan, Temur, and Mardu.

Cube Construction


    My original goal was to play two cards in each tri-color pair, and cut two cards from each relevant guild in order to make them fit.  I never was able to settle on a second Mardu card I thought would actually draw people into three colors, though, and so eventually I just decided to only play one Mardu card and only cut one card from Boros.  In doing so, I ended up with the following multicolor configuration.


    In an attempt to further increase the incentives to play the supported three color combinations, I also included the uncommon tri-lands for each faction (although I decided to use Murmuring Bosk for Abzan).

Drafting and Gameplay


    I did two ten person drafts with the cube in this configuration.  I didn't take specific notes for the first draft but our consensus in discussion afterwards was that the tri-color cards were fun, but there was very little noticeable incentive to choose a faction.  In a 465 card cube, 10 cards here or there simply wasn't enough to shift the draft strategies one way or another.  Roughly half of the three color drafters in that draft were in supported factions, which means the support provided had no real effect on the draft.  In the second draft, we saw the exact same results.  There were only four tri-color drafters.  Two of them were in the supported factions, two were not.
    In both drafts, however, tri-color cards saw play.  People splashed to include them, people wanted to play with them, and they made space to fit them in, even the ones that arguably weren't the best for their strategy, because they were fun and interesting.  The consensus among the people I spoke with was that they enjoyed having the three color cards, especially the splashy ones that did unique things they couldn't get elsewhere, but that they didn't think all the extra stuff I did made a difference.  Before the first draft I made sure everyone knew which factions were supported, and everyone agreed that once the draft started, they promptly forgot which ones I had mentioned and it never seemed to matter.

Conclusions


    I've drawn two separate conclusions from the drafts we did with the cube in this configuration.

    If you want to support tri-color drafting in the way Jason suggests, you have to go further than I did.  Eliminate off color guilds entirely, instead of just backing off the support of them.  Actually decrease the available manafixing for those guilds and increase the fixing for the guilds you want to support.  Make enough of a change that it resonates throughout your cube.  Changing 14 cards in a 465 card cube simply wasn't enough to make a noticeable difference when you were looking at a pack during a draft.  The downside to this approach is obvious (and Jason mentions it in his article): The tri-color cards in existence just aren't deep enough to make it worth it.  You have to really warp your cube for the support make a difference, and when you do that you've made a huge change to your cube just so that someone can draft Siege Rhino.  Until we have better tri-color cards available to us, I suspect this approach won't catch on.

    The more interesting conclusion, to me at least, is the natural followup to the first one.  If changing 14 cards in an intentional effort to warp the cube in the direction of specific tri-color factions didn't make a noticeable impact on color balance, why not just play the few tri-color cards that are actually worth cubing with in a separate multicolor section, and forget about trying to balance it all?  There are few enough of them that you'd want to draft anyway that it simply won't matter.  Based on this experience, I have trouble believing that including only Nicol Bolas, Planeswalker and Maelstrom Wanderer in my cube would have any real impact on Izzet as a guild during a draft.  Two cards one way or another won't skew the draft to the extent that colors or guilds are any easier or harder to draft.

    I know some cube designers have long advocated unbalanced multicolor sections, and I'm coming around to their way of thinking.  The guilds are all deep enough at this point that I won't be unbalancing them anytime soon, but the next time I make changes to my cube, I expect to include a few of my favorite tri-color cards, and I won't worry about the color balance when I do so.

Sunday, November 16, 2014

GP NJ Top 16 Analysis

I was interested in looking over the numbers from GP NJ, threw this together in about half an hour.  Don't expect any deep conclusions, but I found the maindeck card breakdowns very interesting and I thought other people might as well

Top 16 Decklists

Archetype Breakdown

  • 3 Miracles
  • 2 Jeskai Stoneblade
  • 1 ANT
  • 1 Infect
  • 1 U/R Delver
  • 1 U/R Landstill
  • 1 MUD
  • 1 Omni-Tell
  • 1 Jeskai Delver
  • 1 Elves
  • 1 Sneak and Show
  • 1 Grixis Pyromancer
  • 1 Sultai Delver

Top 8 Maindeck Card Breakdown

(# out of a possible 32)
  • 28 Brainstorm
  • 7 Pyroblast (or Red Elemental Blast)
  • 10 Treasure Cruise
  • 2 Dig Through Time
  • 23 Force of Will
  • 7 Wasteland 
  • 0 Tarmogoyf

Top 16 Maindeck Card Breakdown

(# out of a possible 64)
  • 56 Brainstorm
  • 51 Force of Will
  • 15 Pyroblast (or Red Elemental Blast)
  • 21 Treasure Cruise
  • 10 Dig Through Time
  • 15 Wasteland 
  • 4 Tarmogoyf

Notes/Conclusions

  • There was significantly less Treasure Cruise/Dig Through Time in the top 16 than I expected.  My initial thought was that the prevalence of maindeck pyroblast might have something to do that, but the dominance of Brainstorm makes that conclusion look fairly flimsy.  If you look at the actual decklists, it becomes clear the Treasure Cruise is an incredible card in Delver decks that previously would run out of cards, but that blue control and combo decks have better tools to accomplish their goals.  Dig Through Time fills that slot in some of them, but either way Treasure Cruise doesn't seem to be the multi-archetype all-star it has been rumored to be.  
  • 14 decks maindecking 4 Brainstorm in the top 16.  Time for your regularly scheduled reminder that Treasure Cruise isn't actually the problem.  When it comes to a 15 round tournament, the consistency Brainstorm has to offer simply can't be matched or eclipsed.
  • 15 Maindeck Pyroblasts in the top 16.  7 players maindecked two, one player maindecked one.  Neither of the two nonblue decks that made the top 8 played Pyroblast.  Four decks (the 3 Miracles decks and one of the Jeskai Stoneblade decks) appeared to be playing red ENTIRELY for Pyroblast.  Even though all four had incidental red cards in the sideboard (Wear/Tear and Pyroclasm) it's incredibly unlikely any of them would've splashed for red for those sideboard cards alone since similar but less efficient effects exist in the colors they were already in.
  • I've been an outspoken critic of the lack of ability to brew in legacy for a while lately, but Eli Kassis and Lam Phan did a pretty impressive job of going outside the box and getting rewarded for it here.  If you're willing to limit your brewing to decks that contain Brainstorm and Force of Will, there may still be room to do neat stuff.
  • This is the fewest Wastelands I've seen in a top 16 of any major event in as long as I can remember. 

Edit History:

  • Added Wasteland to the maindeck card breakdown
  • Added Tarmogoyf to the maindeck card breakdown

Sunday, November 2, 2014

Storm Archetype Breakdown

The summary below was written in response to this question on reddit.

Right now I'm of the opinion that there are 3 major storm variants that are viable in the metagame. TES, ANT, TNT. Since I have no idea how much you know about each of them, I'll give a quick overview and my perceptions of their strengths and weaknesses.

TES (The Epic Storm)
Overview
  • TES is a 5 color storm deck that intends to cast Empty the Warrens for more than 10 goblins by turn 2 or 3. Turn 1 is definitely possible.
  • Because it's trying to go off so fast, it uses Rite of Flame instead of Cabal Ritual, since Threshold is harder to attain that early, and it can use Silence which is easier to go off quickly with than discard.
  • Empty the Warrens is plan 1, Tendrils of Agony is plan 2. The deck is not constructed to reach ten storm as consistently as other options.
Pros
  • TES is the best Ad Nauseam deck in the format. It plays cards like Chrome Mox which are much worse when trying to storm naturally because it's card disadvantage, but much better after an Ad Nauseam because you'll have extra spells to use for the imprint.
  • Probably the fastest storm deck that actually protects itself. You can get faster with Belcher or Spanish Inquisition, but not with the protection that TES features.
  • Because of the 5 color manabase, it can play Silence instead of Duress. Not all variations do, it depends on the metagame (it's less popular right now), but sometimes that's an advantage.
Cons
  • More all-in than something like ANT or TNT. Since Empty the Warrens is plan one, you can combo off, make a bunch of tokens, and lose a race to a Batterskull. If you make 12 tokens on turn 2 on the play, and they play stoneforge on turn 2 on the draw, they win the race (even if they crack two fetches or crack 1 fetch and force once)
  • Substantially more vulnerable to all of the TNN/Elves hate out there. Most decks are playing Zealous Persecution, Golgari Charm, Rough/Tumble, Pyroclasm, Massacre, etc.
  • Typically runs 0 basics, so more vulnerable to wasteland. Since it can go off from 1 land, that's not always a big issue, but one land + ponder hands can get dangerous quickly.

ANT (Ad Nauseam Tendrils)
Overview
  • Despite it's name, ANT is a Past in Flames deck. It's critical turn is a full turn later than TES, aiming for turn 3-4 instead of turn 2-3.
  • ANT is essentially a two color deck with two small splashes. With only one color of rituals (black), it only needs red for Past in Flames or the occasional Empty the Warrens, and green out of the board for Xantid Swarm, Carpet of Flowers, and Abrupt Decay.
  • As a result of the later intended critical turn, ANT plays Cabal Ritual over Rite of Flame. It intends to cantrip for an extra turn and therefore is more likely to turn on threshold.
  • Tendrils of Agony is plan 1. Often it's the only plan, sometimes you'll see Empty the Warrens in the deck but not consistently.
Pros
  • More consistent than something like TES, which runs fewer cantrips and fewer lands.
  • Because the deck is built to cast Tendrils for 20, you tend to win when you successfully combo off, which isn't true of all storm decks because most other storm win conditions are easier to interact with than tendrils.
  • More consistent manabase means you can fight things like Blood Moon and Wasteland a little more effectively. Typically ANT runs 1 Island and 1 Swamp to fetch for.
Cons
  • Significantly less explosive than TES.
  • As a result, hate bears like thalia, meddling mage, ethersworn canonist, and gaddock teeg are more effective against ANT than they are against TES because your opponent more reliably gets a turn 2 to deploy them.
  • Running fewer win conditions makes it more difficult to react to unexpected cards from your opponent. If you have Tendrils as the only win condition in your deck and your nonwhite opponent puts a Leyline of Sanctity into play that you weren't expecting, you might just be dead. As a result, you typically have to sideboard something like Chain of Vapor in for most/all matchups to prevent those situations from killing you. TES and TNT both manage to avoid this problem by playing Burning Wish and diversifying their threats.

TNT (Burning ANT)
Overview
  • TNT is a bit of a hybrid between ANT and TES. Essentially it's ANT with Burning Wish. The slots typically occupied by Preordain/Sensei's Divining Top/Grim Tutor become ~3 Burning Wishes. This decreases the quality of your cantrips (because the Preordains that are good turn 1 plays become Burning Wishes that might clutter your hand), but increases your threat density.
Pros
  • Increased Threat Density. Playing 4 Infernal Tutors and 3 Burning Wishes means you're substantially less vulnerable to discard.
  • Burning Wish provides access to game 1 hate answers, so you're not just dead to a game 1 Gaddock Teeg or a game 1 Thalia.
  • Sideboarding doesn't change your deck as much because you still use burning wish for access to most of your hate answers, even if you side in 1 or 2 cards your general plan stays together
  • Access to Empty the Warrens out of the board as a plan 2 means you don't have to side in Chain of Vapor in instances where you don't expect Leyline but it's possible.
Cons
  • The Big One - Burning Wish can't get a storm engine. If you're comboing off and all you have is Burning Wish, there's no good engine to wish for. Since Ad Nauseam is an instant and the burning wish exiles itself and is therefore not available to be flashed back with Past in Flames, it's a pretty mediocre card. Most successful TNT lists have a Grim Tutor in the board to wish for, but it requires a LOT of mana for that to work. Some lists run 4 Burning and 3 Infernals with the 4th in the board, but I personally don't like that at all, since Infernal is your best card, I want access to them all without wishing.

I've ignored DDFT (Doomsday Fetchland Tendrils), SI (Spanish Inquisition), and Belcher, all of which are variant storm decks that have the occasional good showing but that I don't think are as good as these three. Also pretty much everyone who plays storm seriously is brewing with Treasure Cruise to try and find the Treasure Cruise storm list. I don't think anyone has succeeded yet, but I won't be surprised if there's a Cruise list in the near future that puts up a few results.

They're all good decks, and if you want to play storm they all share most of the same expensive cards, so I recommend picking one to start with but picking up the pieces to build all of them. I'd choose each deck for a different metagame.
TES - Best in a metagame full of countermagic and hate because of it's speed. If people are prepared for you as the game goes longer, playing the fastest version is the best way to do it.
ANT - Best in a wide open metagame where you don't know what to expect. It's the most consistent deck but the least versatile. It's the hardest to hate out by non-blue decks because of it's lack of reliance on Empty the Warrens.
TNT - Best in a metagame full of discard. As I discussed in it's Cons section, Burning Wish can't get Past in Flames and win from there. However, if most of your opponents are playing Thoughtseize and putting Infernal Tutors in your yard, the Burning Wish CAN find an engine. Also since you're getting thoughtseized more, you'll want the threat density so you can show them hands with no single card that shuts you down.