Magic is an incredibly difficult game. From drafting to deckbuilding, piloting the matches, and more, each aspect of Magic is rich with decisions.
Even the best players don't start from scratch when tackling these decisions. Heuristics – rules that are correct most of the time – are powerful tools that lessen the cognitive effort required by the game. But what separates the good players from the great ones is the ability to identify the scenarios where an accepted heuristic doesn't yield the best play.
There are a variety of heuristics commonly applied to Limited, and, by definition, they're right most of the time. My goal today is to lay out examples that demonstrate when it makes sense to deviate from these heuristics. The following four heuristics are ones I hear most often.
Three Sources for a Splash
The accepted requirement to splash a card is three sources. Some people say that two sources is enough, but I disagree. Your average game of Limited doesn't go past Turn 10. With only two sources, you're less than 70% to see one by Turn 10. If you add a third source, this number jumps above 80%. And if you do the math for earlier turns, say, Turn 6, two sources is 55% and three is 70%. This is such a significant delta that I'm convinced three sources is the line. However, like any heuristic, additional context can prove it incorrect.
1. Your deck has an abundance of card draw or selection.
In Dominaria, I often joked that Opt and Divination were fixing. These cards don't directly put lands of your splash color on the battlefield or in your hand, but they do each let you see two cards deeper into your deck. The reason this heuristic is for three sources rather than two is the math stated above. This math assumes you draw one card per turn. If you have a density of card draw, you can trim on sources for your splash.
2. Your splash card is better the earlier you cast it.
In general, don't splash a card unless it's good later in the game. However, some cards are fine in the late-game, but optimized in the early-game. Consider Rhythm of the Wild. This card is best on Turn 3, but if you cast it alongside a two-drop on Turn 5, that's still solid.
Unlike other splash cards, such as removal spells, Rhythm of the Wild is a terrible topdeck. However, it can still be a great card to splash with the proper synergy. A Simic deck with Skatewing Spy, Sharktocrab, Trollbred Guardian, and other +1/+1 counter synergy cards is always in the market for Rhythm. But because it is most impactful early, the card is lackluster with only three sources. I wouldn't look to splash it unless I had at least four, and preferably five.
3. Your splash card is castable without your splash source.
It's fine to play Deploy in an Orzhov deck with only two sources because Depose is almost always castable. The reason three sources is generally required for a splash is to mitigate the risk of drawing a dead card. Having a Lawmage's Binding stuck in hand for a couple of turns can be detrimental; it's basically a mulligan. When this is not a concern because the splash card isn't dead without the source, two sources or even one source is reasonable.
Choose to Play First
Games of Magic can snowball. Cards like Gateway Sneak can take over the game if left unchecked. Interacting with a creature in a tempo-positive fashion with cards like Arrester's Admonition and Savage Smash can put the player on the receiving end too far behind. And the majority of these scenarios benefit the person on the play.
This isn't to say that the play is strictly better; however, the accepted logic is that when the play is advantageous, it is by a large margin. And when going second is advantageous, it's by a smaller margin. This, coupled with the snowball examples above, means that, if you're not sure, you should just choose to go first.
1. You have a density of cheap removal.
Games of Limited tend to be wars of resources. The extra card is valuable, but the loss of tempo from the draw can mean that sequencing spells optimally is more difficult. The cheaper the interaction in the deck, the less likely it is to fall behind because it can fight a good curve. And once the downside of falling behind is mitigated, the extra card becomes more enticing and choosing to draw makes sense. Similarly, you can also choose to draw if you have a Wrath effect, since that type of card is so good at catching up from behind.
2. Both your deck and your opponent's deck are looking to play the late-game.
Unlike Constructed, who's the beatdown in a match of Limited isn't predetermined. Often, both decks are capable of taking aggressive roles and games boil down to jockeying of position in the combat step. However, it's not uncommon for a deck to win via other means than combat. Some decks plan on killing all their opponent's creatures, and then win with whatever filler creature they have left once the coast is clear. Decks that plan on winning very late in the game cannot afford to go second without a density of cheap interaction because their primary way to win requires not falling far behind. When both decks fit this description, neither player needs to worry about falling behind, and hence it's correct to choose to draw.
3. Your opponent chooses to draw.
This is a bit of an odd one, and it isn't a hard and fast rule. If your opponent chooses to draw, it means something about their deck. For some reason, they believe it is correct to draw. They are either right or wrong. It can't be correct for them to go second and for you to go first. If your opponent is choosing to go second, it is likely for good reason and I would strongly consider mimicking that choice in the following games.
Play Seventeen Lands
Every time I attend a draft at a store and somebody hasn't drafted before, this is one of the first things the other drafters tell them. The conversation usually starts by telling them to draft a two-color deck, pick mostly removal and creatures during the draft, and then play seventeen lands and 23 of the cards drafted.
With seventeen lands, you will make your third land drop in 90% of games. Missing that land drop is one of the most common ways to lose a game of Limited, so seventeen lands is the default.
1. If your deck functions well on only three lands, you can play sixteen lands.
2. If you deck actively wants access to more than six lands, you should play eighteen lands.
3. If your deck functions well on only two lands, you can play fifteen lands. And if you're mostly one color here, you can even play fourteen, although this is an anomaly.
For the most part, Limited decks don't want more than six lands and don't function well without four lands. Most Limited formats have seventeen lands as the default, with a couple of archetypes that fit the descriptions above. Don't be afraid to experiment with your manabase!
Use the Most Mana Every Turn
This is one of the first level-up heuristics. When most players start learning to play Magic, they play almost randomly. They look at their cards and make a decision without much behind it. Eventually somebody tells them to try to use all their mana each turn. This heuristic works well because, in general, cards are costed based on their impact on the game. A two-mana spell is less likely to be as impactful as a three-mana spell by design.
1. Something needs your immediate attention.
If you're looking down a Gateway Sneak, you need to act fast. If you have three mana, a Slimebind, and an Arrester's Admonition, using this heuristic, it's correct to cast the Admonition. But what if you have Slimebind and Aeromunculus? Aeromunculus is the most mana-efficient play, although if your opponent has a Gate, the Sneak gets through and they draw a card. The risk is too high and Sneak can run away with the game. Make the inefficient play because the current battlefield requires your immediate attention.
2. Waste your opponent's mana.
You have five mana, an Azorius Knight-Arbiter, and a Dovin's Acuity in hand. Your opponent has two Islands untapped. Usually it's correct to cast the Arbiter, given the delta in mana cost, but the Dovin's Acuity plays around both Quench and Essence Capture. Whether I cast the Arbiter or the Acuity is highly dependent on the game at hand, but considering the inefficient play in order to inconvenience the way your opponent can sequence their spells is important.
3. Power level discrepancy.
If the cheaper card is noticeably better on the current battlefield, it'll be the better play as long as your plays can be sequenced well in the following turns. This is not common and only happens with the best uncommons and rares. Play Theater of Horrors instead of a Catacomb Crocodile. It's just too much better to care about efficiency.