Today, I’m going to write about the fourth macro-archetype, combo. Hopefully there are no definition arguments this time. The archetype is pretty straightforward and, for most people, it occupies a completely different point on the spectrum (i.e. it’s not positioned between aggro or control, it’s something else altogether).
Combo decks have been around forever, but the first one I can remember is Mike Long’s ProsBloom, from Pro Tour Paris 1997:
Combo is a deck where all your cards have very little value, until you achieve a critical mass of them, at which point you win the game with them. In this particular deck, you get a Cadaverous Bloom or a Squandered Resources into play (or both), then you chain Natural Balances and draw spells until you get enough mana to Drain Life them for a billion.
The only goal of every other card in your deck is to get you to a point where this happens. If things don’t happen exactly like that, then none of your cards actually do anything—when you look at a game in which a combo deck didn’t work, then it looks like a very bad deck. Most of your games, you’ll be losing all the way up until you suddenly win—there is no gradual battle for resources, no burying them in card advantage, no squeezing in the last point of damage, it’s all or nothing. With most combo decks, the real challenge is to know when it’s time to “go for it”, and when it’s time to wait.
To me, there are two kinds of combo decks: ones that seek to get a specific combination of cards into play, and ones that rely on chaining multiple spells from your deck.
Two- and Three-Card Combos
The idea here is that you want a specific combination of cards—usually two or three, if your deck is good—that don’t do anything particularly important by themselves (not necessarily, obviously, but usually the way it goes), but, when assembled, win you the game. Here is Samuele Estratti’s deck from Pro Tour Philadelphia:
Most decks like this one will win the game when they find their combination, regardless of what they’ve been doing previously or what happens next—there is no luck component, no chance of fizzling. It really is all or nothing. As such, other than finding more disruption for their disruption, you don’t gain anything by waiting. A turn four Splinter Twin is going to win the game just like a turn seventeen Splinter Twin will when unopposed. This means that you don’t have any incentive to wait if you think they can’t stop you, and most of the time you should just go for it as soon as you can. It also means these decks are better than their counterparts at winning a game where they’ve been constantly disrupted. You only need those two cards, and not a critical mass of resources.
Playing these decks against a deck with no disruption is incredibly easy—it’s just a race. Can they bring your life total to 0 before you get your combo? If not, you win.
The majority of games are rarely like this, though, especially when your combo involves a creature staying in play. If they can interact with you, you must judge two things: how likely it is that they’re going to be able to disrupt you if you go off now, and how likely they’re to disrupt you if you try to go off in the future.
This might seem like a lot of guessing, and to some extent it is, but it’s always going to be an educated guess, and it’s not exactly rocket science. To know how likely it is that they’re going to be able to disrupt you, you need to know what kind of answers are available and widely played in the format, and in what numbers. If you’ve playtested at all, things shouldn’t go much differently than you would expect (i.e. “he is playing GW, most GW decks play four Path to Exile”).
Then you factor in how much they know—do they know of your deck? If so, then maybe they wouldn’t have kept such a slow hand without an answer, and it’s more likely that they have something. Did they just tap out last turn when you could have killed them? Then they probably don’t have anything, and they’re just trying to race you. Again, this is certainly guesswork, but trying to “read” what they have in hand will usually be, and the best you can do is to assign random probabilities in your head that you think are close to right (and, again, you don’t necessarily need numbers—you can just think in terms of “likely” and “not very likely”).
Then, once you have a general idea, such as “they will probably be unable to disrupt me if I go off this turn”, then you need to figure out if they'er more or less likely to be able to do it in the future, assuming that killing you counts as disrupting you. That’ll be a function of what you have in your hand and in your deck, and what they have.
To make things simple, assuming ample mana available, it will simply be a question of whether they have more answers to your combo in their deck than you have answers to their answers. If they have four removal spells and you have eight counterspells, then every turn you wait favors you. Whereas, if you have four and they have eight then you want to minimize the number of draw steps both players get. The overwhelming majority of the time, you’ll be favored here, because your deck is built for that—even if you’re not drawing answers, you’re drawing draw spells, which increases your likelihood of finding more as well. I’d say that, if you can wait, then most of the time you will wait.
Sometimes, though, mana is a concern. Imagine that you have the Twin in hand with a Dispel, but it’s only turn four. By waiting until turn five, the worst that can happen is that your opponent draws a removal spell they didn’t have, but that’s going to be countered by the Dispel you can now cast. So, at worst, you’re even. At best, you’re going to counter the spell they had previously with the new Dispel, so you’re up. In this case, if they aren’t threatening to kill you, then you’ll definitely wait.
It’s also possible that they have a trump that you can’t deal with easily, and the only reason they haven’t played it is that they have no mana. When someone is sitting on Ghostly Prison, then once they reach their third land it’s going to be much harder for you to kill them, and you should just try to go for it even if there’s a chance you’ll be disrupted. This is especially true because once they do play the Prison, then whatever disruption they have that was stopping you from going off the previous turn will still be there, and you’ll still have to go through it to kill them (but now you’ll have to go through more).
If you’re in an early enough stage of the game, then you might just constrict them on mana in a way that allows them to play only one answer, even if they have multiple. The Twin deck plays Pact and Dispel, for example, which will be cheaper than whatever answer they can get. So, as a general rule, I’d say that, if the battle is for resources (disruption versus answers), then you wait as much as you can, but if mana plays a part in constricting them, then you should seize the opportunity and go off as soon as you can.
Another interesting thing with this kind of combo is that, unlike most Storm combo decks, you don’t actually lose if they disrupt you—sure, you’re going to three-for-one yourself, but it doesn’t matter much. Given enough time, you’ll amass your components again, and once you do, the concept of card advantage will have lost all meaning. You’re basically blanking their entire deck with two of your cards, which gives you ample room to spew before that happens.
By Storm Combo, I mean the sort of deck that chains multiple spells in a turn, even if it doesn’t necessarily use the storm mechanic. The Prosbloom deck above would be a storm combo, for example. Another example is the deck played by Conley and Matt Nass (and many others) in the same tournament Estratti won with Twin:
What differentiates these decks from the two-card combo decks is that you need a bunch of resources rather than just two, and often, once you start to go off, you’re not sure if you’re winning or not—you might just hit a clump of lands and die horribly. As such, you have even more incentive to wait here. Even if you know they have no disruption, every draw step you get and every extra land drop you make increases your chances of success.
You’ll be in this scenario a lot more with this sort of deck, because it rarely relies on creatures, so the chances people will be able to disrupt you are very low—you’re basically goldfishing in two-thirds of your matches, and every time you do that you should wait until you absolutely can’t wait any longer, and only then try to go off.
If the opponent has disruption, then things are going to depend on whether he has a clock or not. If they have disruption plus pressure, you’re in trouble—you’re probably not going to win if both players get an average draw. If they’re a control deck, though, then you’re actually fine, because those decks are a lot more resilient to permission if they’re given enough time.
Against a slow deck, then you just sculpt your hand as much as you can, and you’ll get to a point where it’s not going to matter if they counter a single thing or not, or even three “single things”. Every card in your deck is pretty interchangeable when you have a ton of mana, and they’ll end up just adding to your storm count.
With the two-card combo decks, you can just throw stuff into counterspells—they’re just one-for-ones. You can even throw stuff into disruption, because, again, you’ll be three-for-one’ing yourself in a deck that only needs two cards to win.
With this kind of deck, you can’t do that. If you get your last ritual countered, then you’ve just seven-for-one’d yourself in a deck that needs 7 cards to win—you’re probably not going to be able to get it all together again before they kill you. This means you can’t “test the waters” or “eh, I’ll just try it”. When you try it, you had better kill them, or you can assume you’ve lost the game.
When I was watching Luis play Tron versus Swathstorm, he kept winning what is a theoretically horrendous matchup because his opponents kept trying to go off as soon as they could, only to be stopped by one Condescend or even one Remand. They didn’t understand that, if they just waited until the point where they could afford to get a ritual countered, then they wouldn’t be able to lose. In this case, Luis’ deck had one kill condition and it cost
By trying to do it on turn three, his opponents were turning six or so of his cards (the counterspells) into actual kill conditions, because if he had them then they were just dead—even his glacially slow deck could just kill them with Colonnades if they got to one card in hand. Going off on turn two is cute, but it’s not necessary when you’re playing against a slow deck, and it’s not recommended unless it’s a very specific scenario.
Building a Combo Deck
The biggest question when building combo to me is not “what do I want in my combo deck,” but “should I play combo in the first place.” Most of the time, the answer is no. But, when the answer is yes, then it really is a resounding yes. Playing a different combo deck that people weren’t expecting and killing them in a weird way they never saw coming is one of the most rewarding feelings you can have when playing a tournament, and finding such a combo deck is hitting the jackpot in playtesting. But, this scenario doesn’t happen very often (to me, it happened only twice: when I played Enduring Ideal at Nationals and when I played Swans in Barcelona, and it almost happened in Philly with Blazing Shoal. To a lot of people I’d imagine it happened in Berlin, with Elves). Most of the time, the combo deck you’re deciding whether to play or not is an established deck everyone knows about.
Deciding if you want to play a combo deck will usually depend on other people’s perception of the metagame rather than your deck. Rarely will you be able to play it if everyone is gunning for you, no matter how good it is (with notable exceptions, such as Flash). Playing combo is accepting the fact that, as much as you get free wins, you’re going to get free losses—either the games in which you don’t draw your one piece, or the games in which they play a permanent you can’t deal with. It’s frustrating to have your opponent mulligan to four, and still beat you because you couldn’t find your Splinter Twin, whereas if you were just playing a bunch of Grizzly Bears you’d have easily won, and it’s very frustrating to auto-lose to someone who has a hand of six lands and a Leyline of the Void.
I remember at Worlds ‘03, in Berlin, Carlos was playing for Top 8 with Mind’s Desire against Goblins, and he lost the last game because he drew around 50 cards, but all four Desires were in the bottom 10 cards. At this point, it doesn’t matter what his opponent does, he is just going to lose—his opponent could have been playing Mono Green Giants and he still would have beaten him.
At Nationals, playing Twin, I lost a game in which I drew over 30 cards, had all four Splinter Twins in my hand but no Exarch. Sure, this is not likely, but it will happen, and it’s a risk you’re accepting when you play those decks—a risk that you don’t have in almost any other deck.
Building a combo deck is usually not that complicated (as opposed to building a sideboard for it, which is almost impossible to me). There are generally a lot of cards you
Two-card combos can usually play a lot more answers, whereas a storm combo deck will often ignore whatever it is the opponent is doing and just play cards that advance its own game plan, because drawing answers in the middle of your combo is very bad. Just take a look at the two lists from Philly. Estratti can afford to play Lightning Bolt, Firespout, Dispel, Pact of Negation, Remand and Spellskite, whereas Matt Nass has only Remand and Lightning Bolt, both of which have applications in the deck other than just being answers.
One mistake I see people make sometimes is to play less than four of their combo card, because it’s “slow” or “doesn’t do anything outside of the combo.” I remember seeing a lot of lists with three Splinter Twins, for example. This is just stupid—if you can’t accept having dead Splinter Twins in your hand, then play a different deck. The amount of games you’ll lose because you didn’t draw any of those is much greater than the number of times you’ll get Twin-Flooded (also look at Estratti, he has six!).
Mulliganing with a Combo Deck
Mulliganing with Combo is not very hard, because you’re generally on your own, as it happens with combo a lot of the time. You don’t need to consider your opponent, at least not game one, only whether the hand is good for your deck or not. When playing combo, I’ll tend to take riskier keeps, because, if you get there, you just win—for example, a hand of Island, Hulk and Flash is excellent—as soon as you hit your second land you’ll get full value out of all of your cards (i.e. they’re going to die). It doesn’t matter if it’s turn two, four or nineteen. I won’t, however, keep a hand of one land and Bitterblossom, because if I miss for two turns then chances are my Bitterblossom won’t be able to get me back in the game, and its value will be greatly diminished if it’s played late. As a general rule, you’ll probably keep if you’re missing one combo piece, and you’ll probably never keep if you’re missing all of them.
Sideboarding with a Combo Deck
Sideboarding with Combo decks is by far the hardest facet to me. Again, there are a number of cards you absolutely must have, and that gives you very little room to change anything. In a deck full of answers such as Estratti’s, then you have more room to maneuver, because specific answers are good against specific things—you don’t need Firespout against a control deck, and you don’t need Dispel against a deck with no instants that interact with you, so you just swap those.
In the storm decks, though, you need everything! Still, there are some things you must board in, because if you don’t you just lose to certain cards, and the idea here is to board something generic, such as a bounce spell—it won’t do to board in Ancient Grudge for Canonist and lose to Rule of Law. You should also not board a lot of those cards in, because if you keep drawing them then the opponent will not need any hate to beat you—you’ll probably have a lot of draw spells you can use to get to them if you need to.
A different approach you can take is the “transformational sideboard,” which involves radically changing your strategy to avoid their countermeasures entirely. Your deck will never be as good as it was before (or you’d be playing that deck to begin with), but the idea is that your opponent’s deck will also not be as good, since he has tons of answers to cards you no longer have (and ideally zero answers to the ones you do have), plus you might have gotten a free win game one out of the deal, so maybe the “worse” deck is good enough to win one out of two.
The one thing you must avoid is transforming into something that loses to whatever it is they were boarding anyway—if you’re playing Dredge (which is an entirely different kind of combo deck), it’s silly to board in Tarmogoyfs and Tombstalkers and expect to kill the opponent with them, because they’ll use the graveyard removal they boarded to stop you.
It’s also silly to try to transform when your deck has a lot of cards that are very narrow—again, Dredge is a good example: sure, you’re going to add 15 cards, and what are you going to do with all the Stinkweed Imps, Narcomoebas and Putrid Imps left in your deck? Surely they can’t be good in your new strategy. Belcher has two lands in it. What is it going to board into, a bunch of Memnites?
One transformational sideboard I recall is the one for Swans, in Barcelona— here is Joel Calafell’s winning list, which I also played:
You see that our sideboard is somewhat transformational, with Countryside Crushers as a backup plan. However, we also have answers for their answers—tools to fight in “Swans” ground if we want to, with Maelstrom Pulses, Elders and Auras for their Runed Halos and whatnot.
A different group (namely Luis and Gerry) had a different approach. They had Rain of Tears, Fulminator Mage, and I believe Deny Reality in their sideboard, to try to board into a land destruction deck. Theirs was a full transformational sideboard—they had no answers to hate, instead trying to circumvent it by radically changing their strategy.
Another example I particularly like is Max Bracht’s Heartbeat deck from Hawaii, which aimed to circumvent hate by boarding into Kudzus, Iwamoris and Melokus, cards that would obviously not make a good deck, but that are just good enough when the opponent’s deck is now full of enchantment removal and no way to kill creatures:
Yet another example are the people who were boarding in Splinter Twin and Exarch in their Grapeshot/Pyromancer decks, since they theoretically take out creature hate, but that I don’t like—you still lose to counterspells, discard and enchantment removal, all of which they will board against you no matter what.
Whatever you do, though, if you’re playing a transformational sideboard (or even if you’re playing a deck that could have a transformational sideboard), always shuffle your sideboard and your deck together and then remove 15—this way, you leave them clueless as to what you’re doing, as opposed to telling them “you don’t need to leave in creature removal against me” when you only side in two cards as Pyromancer.
Well, this is it for today—I hope you’ve enjoyed it, and see you later this week!