A heuristic technique, or a heuristic is any approach to problem solving or self-discovery that employs a practical method that is not guaranteed to be optimal, perfect, or rational, but is nevertheless sufficient for reaching an immediate, short-term goal or approximation. Where finding an optimal solution is impossible or impractical, heuristic methods can be used to speed up the process of finding a satisfactory solution. Heuristics can be mental shortcuts that ease the cognitive load of making a decision.
(Myers, David G. (2010). Social psychology (Tenth ed.). New York, NY.)
Magic is a very hard game, and in every game you play you will have to make a lot of decisions. Some of these decisions will be completely new to you, scenarios that you’ve simply never seen before, and you will have to think them over from scratch. The majority of decision points, however, will be moments that have at least a passing similarity to other situations you’ve already encountered. As you practice more and build up your repertoire of games and deck styles, you start realizing that, even though some situations are fundamentally different from each other (e.g. two different decks and formats), the way you think about them is very similar.
As you think through each sort-of-similar situation again and again, you start building up what I think of as your heuristics bank. Eventually, when a similar situation happens another time, you will no longer need to think everything through from scratch — you will identify the similarities to your previous situation and go from there. This is extremely useful, because it’s basically impossible to think everything through every single time — you will spend all your time and all your brainpower and not progress past Turn 5 of each game.
In Magic, you need shortcuts, even if they’re not necessarily conscious shortcuts. You need to recognize patterns and have starting points for your decisions. These shortcuts are heuristics, and today I’m going to talk about heuristics that I use in every tournament that I play – thoughts that I’ve had so many times that I now just apply to my actions and use to kickstart my more specific decisions.
It’s important to know that, with heuristics, they aren’t necessarily absolute. As we can see from the Wikipedia definition, they’re not guaranteed to be optional, and it’s important to know when to deviate from them. However, I would much rather have a good heuristic and follow it incorrectly the couple of times that they should be deviated from than not have a good heuristic to begin with.
For example, a very simple heuristic that we’re taught early on is that, if your hand has a good mix of lands and spells, it’s a keep. We could potentially evolve that into “keep hands that have between two and five lands”. If a new player follows this heuristic, it will probably produce a better result than not following it at all. Similarly, “play seventeen lands in Draft” will also produce better results for a newer player than trying to figure out how many lands to play each time. On the other hand, a good player will know to mulligan a hand of five lands and two Dance of the Manse, or to play a different number of lands based on the format, their mana curve, how many mana sinks they have and so on.
So, what are the top heuristics that I use in my matches of Magic?
Heuristic #1: You should first play a land, then attack, and then cast spells.
The goal of this sequencing is to give your opponent as little information as possible to inform their blocks. If you cast a spell before you attack, then your opponent knows two things — they know which spell you’re casting, and they know you’re not using that mana during combat. Therefore, you should almost always cast your spells after combat.
The land bit is a bit trickier. You could think about it and reasonably conclude that you should also withhold playing a land, because putting a land onto the battlefield gives your opponent information that this card is that specific land, as opposed to it being possibly anything in your hand. However, I’ve found that, most of the time, playing a land increases the number of things your opponent has to play around, since you have more mana available, which actually decreases information.
For example, imagine you’re playing Dimir Rogues and you have three lands on the battlefield. In this case, playing a fourth land means your opponent has to play around Zareth San, the Trickster, or around any two of Brazen Borrower + Drown into the Loch + Heartless Act, or double Soaring Thought-Thief, or a combination of Thief + removal. If you only have three lands when you attack, a lot of these possibilities no longer exist.
If you’re going from five to six lands, then you could be representing Zareth San plus Drown in the Loch, or three interaction spells. This is especially true in Limited, where often playing a land would represent the possibility of a second trick, or a different trick that wasn’t possible before – sometimes you’re even giving yourself an entirely new color of mana to work with.
This heuristic specifically is one of the ones that requires the most individual judging, as sometimes playing the land gives up more information, but I’ve found that playing the land before combat usually gains and is therefore a good heuristic.
Heuristic #2: Sideboarded games are much more interactive than Game 1s.
As a general rule, Game 1s of a match are a lot more about doing your own thing than Games 2 and 3, since most people bring in disruption elements after sideboarding. This is something that is commonly true across the board and impacts how you build your deck, how you mulligan, how you sideboard, and how you play the games.
A very good example of that is the Gruul Adventures in Zendikar Rising Standard. Gruul Adventures is mostly trying to goldfish you Game 1; it has a few interaction spells, but it’s using those mostly as a way to clear out blockers and it still operates fundamentally as an aggro deck. For Games 2 and 3, it often morphs into a midrange and sometimes even a control deck.
For example, take this opening hand against a traditional Dimir Rogues deck on the play:
To keep this hand in Game 1 would be unthinkable – it’s simply too slow. You need to pressure them a little, since they win the longer games, and there are several cards you will draw that are not going to be good, such as Embercleave and The Great Henge.
Now, imagine the exact same hand in Game 2. You no longer need to be as aggressive; instead of cards like Embercleave and Brushfire Elemental, you have cards like Ox of Agonas; Vivien, Monsters’ Advocate; Phoenix of Ash; and Klothys, God of Destiny, which mean you actually win the longer game. In fact, you’re not even capable of being as aggressive as you were before, given the composition of your deck.
You switch roles in the matchup, and therefore you should switch your approach to mulliganing as well. This, for me, is a hand that I would not consider keeping in Game 1 but that I would probably reluctantly keep in Games 2 and 3. Even if this is not a hand you would end up keeping, the fact that you’ve sideboarded should make you think much harder about it.
You can also use this heuristic to inform deckbuilding. One relatively recent example is the latest Historic Arena Open, where my team and I ended up having to choose a deck without playing very many games of the format. We settled for a Temur Reclamation build (this was before the bans) and the biggest question was whether we wanted Blast Zone or Field of Ruin as our colorless utility land. Blast Zone was great versus Auras (both Mono-White and the Azorius version) and versus Mono-Blue, and Field of Ruin was obviously great in any Field of the Dead mirror match, whereas each card was horrible in the opposing matchup.
Ultimately, what made us decide for Blast Zone was the fact that we didn’t believe Game 1s in the mirror were going to be about Field of the Dead. There was simply very little interaction in both builds, which meant the main gameplan of Wilderness Reclamation + Expansion // Explosion was going to decide the game the overwhelming majority of the time, as opposed to the sideboarded games where everyone had counterspells and the Zombie tokens were actually relevant.
Because of this, we thought that Field of Ruin was actually not going to be very important in Game 1s, even though it’s supposed to be a “trump card” for the matchup, whereas Blast Zone would be relevant all games in the matchups it was meant to be played against, and therefore the better choice.
How many games of the matchup did we play before we decided on this? Zero. We simply looked at the decks and used the heuristic of “presideboarded games are less about interaction and more about doing your own thing.”
This heuristic was also very important in Standard Temur Reclamation mirrors and informed how most good players approached the matchup. As a general rule, in Game 1s, I’d usually just slam Wilderness Reclamation with many of my hands; there wasn’t much interaction that they could have and they were less likely to play around your own interaction if you kept open mana. In Game 2s, however, both players had a lot more interaction, so just slamming Wilderness Reclamation early on was often a death sentence, and you had to play more conservatively, which you could do because you knew your opponents would have to respect open mana a lot more.
It’s for this reason that some players (famously Jean-Emmanuel Depraz) did not think Shark Typhoon was a good maindeck card — because Shark Typhoon is a Plan B and the matchup in Game 1 was all about the Plan A. If no one is interacting with each other’s Plan A, then Plan A will win the game for one player or for another, which means you don’t really need a backup of any kind.
Heuristic #3: Quantity decks mulligan worse than quality decks.
One big classification I try to always have in my mind is whether a deck is a quality deck or a quantity deck. Quality decks have specific, very powerful cards that make the entire deck work. These specific cards empower the rest of the game and make up for big gaps in raw card quantity. For this reason, these decks should mulligan more aggressively, since they have a card to bail them out; you should not settle for a suboptimal hand with a quality deck.
One example of a quality deck that I played a lot was Faeries. Faeries had the card Bitterblossom, and if you had Bitterblossom on the play, all your other cards became better — Scion of Oona had creatures to pump, Mistbind Clique had Faeries to champion, and all your counterspells were more powerful because you were progressing your battlefield without cost and could actually afford to sit on them. A Faeries deck with Bitterblossom was many times more powerful than one without, and you were often much better off mulliganing to a five-card Bitterblossom hand than having a mediocre seven without it.
There have been many other famous examples in history — Miracles with Sensei’s Divining Top; Mono-Red Aggro with Hazoret the Fervent; and recently Simic decks with Oko, Thief of Crowns. As you can see, this is something that can apply to all types of decks, from aggro to control.
Their counterparts are the quantity decks, which want raw card numbers but don’t necessarily care which cards they are. These decks usually don’t have specific cards that empower the whole deck; rather, all the cards work with each other and you just want as many pieces as possible to combine. These decks do not mulligan well and you should be willing to keep worse hands with them.
The quintessential quantity deck in my mind is Modern Burn. With Burn, all your cards do almost literally the same thing; they deal three damage to the opponent. All you care about is amassing enough of them to deal twenty, but whether that’s done with Lightning Bolt, Lava Spike, or Skewer the Critics is not very relevant. Other examples of classic quantity decks are Valakut, the Molten Pinnacle decks that just need to find a critical mass of lands and tribal aggressive decks like Merfolk or Humans, which usually scale quadratically (meaning each subsequent lord adds even more to the battlefield — for example, one creature will be a 1/1, but if you have three creatures they’ll be three 3/3s instead, so all you need is a high quantity of creatures).
There are also decks that tend to turn the match into a quantity match regardless of what the other person is playing. Jund Midrange in Modern is a deck like this; between discard and removal, it has so much disruption that any individual card is unlikely to resolve and stay on the battlefield forever, so what you need is simply more cards. When you’re playing versus a deck with many targeted discard spells, you should be much more liberal in which hands you keep, because it’s likely your best card will be gone.
For example, imagine a sample hand like this in a Wilderness Reclamation deck:
This is a very good hand, and I assume everyone would keep it. However, if you’re playing against a deck with many discard spells – say, a Mono-Black Aggro deck with Duress, Kitesail Freebooter and Agonizing Remorse – there are people who would look at this hand and say you shouldn’t keep, because you only have one actual spell and it’s going to get discarded. I’ve seen this argument many times.
What this argument misses is that the hypothetical discard spell will not disappear if you mulligan — it’ll still take away your best card, but now you’ll only have five other cards to work with. What you should do is the opposite — keep hands like this specifically because you’re playing versus a deck with discard and mulliganing versus them is punishing.
Now, why is this heuristic useful? It is useful because you don’t need to think through the situation every time. Whenever I’m playing a deck (or sometimes a matchup), I classify that deck as a quantity deck or a quality deck in my mind, and then when it comes to mulligan decisions I already know how to adjust my thinking. It doesn’t make the decision for me, but it lays the framework for the decision to be made — with quantity decks I mulligan conservatively, and with quality decks I mulligan aggressively.
Heuristic #4: In a mirror match, the person on the play should stick to their proactive gameplan. The person on the draw will be forced to react.
A lot of the time, whether we’re on the play or on the draw should inform how we sideboard and how we play the games. This happens for every matchup, but the difference is most striking in mirror matches. As a general rule, the person on the play is better off keeping to their original proactive gameplan, but the person on the draw doesn’t have that option, because if they do that, then they will always be second to whatever play is happening. Because of this, the person on the draw is forced to assume a more reactive gameplan.
The best example of this in recent memory was Simic Food mirrors. In that matchup, I remember that Ondrej Strasky’s sideboard plan in the mirror was very minimal — I believe he changed one card. He reasoned that, on the play, all he needed to do was be the best Simic Food deck he could be. On the draw, however, this plan just wasn’t good enough, because if both he and the opponent focused on being the best Simic Food deck they could be, then the opponent would just win by virtue of being on the play. Because of this, he needed more reactive cards, like Mystical Dispute, to at least try to put the brakes on the opponent. On the play, there wasn’t even a window to play these cards, as he was likely to want to tap out for his own plays.
The same is true for the Gruul Adventures mirror in the Zendikar Rising Standard. For example, here are our team’s sideboard plans for the Gruul Adventures mirror:
Out (on the play):
In (on the play):
Out (on the draw):
In (on the draw):
This is a radical change in approaches. On the play, we wanted to keep our deck as proactive as possible, so we had zero copies of Scorching Dragonfire. On the draw, we were forced to be more reactive — cards like Embercleave and Rimrock Knight came out, and the deck had the full four Scorching Dragonfires in it. If you pay attention to many mirror matches throughout history, you’ll see that this is a very common heuristic in high-level play.
Heuristic #5: Early during a set release, try to maximize the best-case scenario. Later on, try to minimize the worst-case scenario.
There are two kinds of people during preview season — those who think everything is broken because they only think about the best-case scenario, and those who think everything is bad because they only think about the worst-case scenario. For a long time I was the first type of person, but I’ve since realized that, much like with almost everything, the correct approach is to just have a balance.
Early on, when a set has just been released, the best approach is to be a “maximize the best-case scenario” kind of person. You want to give the new cards everything they need to be able to succeed to see if they’re worth pursuing — this means, for example, only playing cards that advance your own gameplan as opposed to some sort of removal spell. If the card isn’t good enough even when everything is stacked in its favor, you can dismiss it outright.
If it is good enough in this situation, then your approach should change; you should then try to make it so that the worst-case scenario isn’t as bad. Sure, this deck wins when everything goes right, but is there a way to make it win when things go wrong? Is there a way to win through removal, discard, counterspells? At this stage, you should then consider everything that can go wrong and see how your deck behaves in the face of that, and you know this will be something worth trying because you know that your deck is very good when it works.
The most important example of this for me personally is Elves in Pro Tour Berlin, a tournament that happened over a decade ago. In case you aren’t aware, that was a tournament that was dominated by a breakout deck: a combo Elves list with Glimpse of Nature, Heritage Druid, and Nettle Sentinel that could realistically kill on Turn 2.
I actually had access to the deck before almost everyone else in the tournament, but I made the mistake of dismissing it because I only considered the worst-case scenario; I focused on all the things that could go wrong outright and that made it impossible for me to see the true power of the deck.
The other teams (which had access to the deck later than I did, for the most part) did the opposite — they saw the best-case scenario, saw that it was broken, and only after that they spent all their time trying to minimize the worst-case scenario, arriving at several Plan Bs and alternate win conditions that bypassed the issues they were having. Since I skipped the first step, I didn’t feel like spending all my time working on solving this problem and I missed out on a very big opportunity.
So, now, my first approach when looking at a new card or a new deck is to think about it in the best possible light, and only then, if that seems good enough to me, I start thinking about the potential problems.
Heuristic #6: When playing a slower deck versus a deck full of discard, discard your spells and keep your lands.
A long time ago, when I was testing with Luis Scott-Vargas and Paul Cheon, they told me a story from their US Nationals testing times. They were both in the Top 8 playing Solar Flare (an Esper tapout deck full of big creatures), and they were supposed to help each other playtest their respective Top 8 matchups — Luis was playing against a generic opponent and Paul was playing versus a discard-heavy deck (something like Mono-Black Discard). They playtested Luis’s matchup for a couple of hours, and then when the time came for them to playtest Paul’s, Luis said, “Actually, we don’t need to test. I played against that guy in the Swiss. All you have to do is discard your spells and keep your lands and you’re going to win.”
Paul did exactly that and won the match 3-0.
This is a lesson that, in my mind, is applicable to this day. Whenever people play against a heavy discard deck, they lean towards keeping their best spells in hand, which makes sense — after all, we should try to play our best spells, not discard them. Most of the time, this is the incorrect approach.
I saw this a lot when Orzhov Doom Foretold (Yorion) was a deck, a while ago — people would discard their lands and hold onto their powerful cards (Elspeth Conquers Death, Expansion // Explosion, etc.), and then inevitably these cards would get discarded too, and the person would be left without lands or spells. A while after they would topdeck spells again, as one is wont to do against discard, but they would not have the mana to play them, because they discarded their lands early on in an attempt to hold onto their most powerful spells.
The correct way to play these matchups is usually to just make your land drops, discarding your expensive spells and relying on the fact that you will draw more spells over the course of the game. Then, when you do draw them, you’ll be able to actually cast them, instead of just discarding them a turn later. I’ve found that this is the case with almost any deck that has powerful spells against decks that have a heavy discard component.
All in all, I’m sure there are many heuristics that good players use, but these are the ones that I use that might not have been entirely intuitive when I started. Again, please remember that they are only heuristics and not absolute rules — you will have to deviate from them from time to time, they’re merely guidelines that should inform your thought process so it doesn’t have to start from a blank slate every time. If you keep that in mind, I think adopting those might cause meaningful improvements on several aspects of your game.
