Optimal Hand Selection in Standard

Feature Article from Kuan-Kuan Tian
Kuan-Kuan Tian
submit to reddit » Print «

Welcome back to another installment of Kuanling's School of Mise. This week I'm going to talk about one of the most important gameplay decisions that Magic players commonly face: the decision whether to keep their starting seven or to throw it back. Proper mulliganing, in my opinion, is an important skill to master because the entire course of the game will depend on this initial choice. Furthermore, the majority of Magic players actually have terrible mulliganing habits and I would not be surprised if as high as 10% of their losses were directly related to incorrect mulligan decisions.

What this article seeks to do is to help you master the skill of proper mulliganing by outlining several principles that will greatly aid your mulligan decisions. These principles are not only simple to learn but they can also easily be committed to memory, especially since I have provided a bunch of examples from a variety of different formats that highlight their application. Of course, while you're reading this article, you should always keep in mind the distinction between “principles” and “rules”. Rules and their exceptions (exceptive rules) apply to a situation in an all-or-nothing fashion, determining the decision without any leeway for an alternative. Principles, on the other hand, not only have a tendency to conflict with one another but they also have a measure of weight. Often, two or more conflicting principles will apply to a situation, which means that we will often need to add up the weight of the principles supporting each alternative and balance them in order to find the correct answer.

Now that we have this important distinction in mind, let us move on to examine the principles of mulliganing…

Principle—When in doubt, mulligan
In general, PTQ level players are notorious for under-mulliganing. This, I think, is due to a combination of three reasons.

1)PTQ level players overvalue card advantage. The power of card advantage has been drilled into everyone's heads because it is the most measurable and easiest to understand resource. However, there are other resources in magic as well, such as mana, spells and tempo, which means that the loss of a single card is not as devastating a blow as most people imagine. Indeed, if it makes sense to trade tempo for cards with a card like Divination or to trade cards for tempo with a card like Chrome Mox, then it makes sense to go to Paris when your hand is especially low in other resources.
2)PTQ level players often convince themselves that their hand is keepable by imagining ways in which their hand can be “fixed”. Just think of the number of times that you've kept a borderline hand because you convinced yourself that the hand would be fine if you drew a Forest. Mulligan decisions should be based on mathematics and not wishful thinking.
3)Taking a mulligan requires effort. If you want to take a mulligan, you have to shuffle your deck thoroughly, present it to your opponent, and draw another hand. Indeed, if you're one of those deprived Magic junkies that just wants to “play a game” or to “get a game in”, then you are going to under-mulligan as well.

What this means then, is that it makes sense to mulligan borderline hands that you are unsure about because that is a positive EV decision if you generally under-mulligan. Of course, following this principle will cause you to mulligan a few hands that you would have won with, but this seems to be an acceptable consequence if you just think of all the games that you would have lost from keeping hands that you should have thrown back.

Principle—Mulligan hands with two spells or fewer
As I mentioned above, the number of cards you have isn't the only relevant resource in Magic; the number of spells you have is important as well. As such, it makes sense to mulligan hands that don't have access to many spells. The following hand, for example, taken from Tomoharu Saito's Jund Ramp deck (which can be found [here]), is a clear mulligan.

Hand: Sprouting Thrinax, Dragonskull Summit, Dragonskull Summit, Savage Lands, Savage Lands, Broodmate Dragon, Swamp

Don't let the presence of the Broodmate Dragon fool you. The card is undeniably good but it does not make up for the paucity of spells in this hand, especially since the Dragon is rather ill-equipped to deal with some of the commonly played cards in the format, such as Baneslayer Angel, Flashfreeze, and god Forbid, a turn one Steppe Lynx.

Principle—Mulligan hands with one land or fewer
No-one likes being mana screwed. And yet for some reason I see a lot of players keep one land hands. Seriously, if you kept a one land hand then you do not have the right to complain about losing because you weren't allowed to cast your spells. The following hand, from Bram Snepvanger's Boros deck (which can be found [here]) is a classic example of a hand that you should not keep.

Hand: Elite Vanguard, Steppe Lynx, Ranger of Eos, Path to Exile, Plated Geopede, Lightning Bolt, Arid Mesa

This hand is a mulligan even if you are on the draw. Bram's deck plays 24 lands which means that there is a 30/53 x 29/51 = 32% chance of you missing your second land drop. This means that 32% of the time, your deck will not be able to function properly at all. Furthermore, even if you draw a land, your hand isn't even that great. For example, you could draw a Plains or a Marsh Flats which means that you would still be unable to cast the two red spells in your hand. And even if the land you draw is a red source, your turn two Geopede isn't the most exciting play in the world since you'll need further land drops to turn the Geopede into something other than an overcosted Tundra Wolves. Oh, and by the way, you have a less than 9% chance of being able to cast the Ranger of Eos by turn 4.

Principle—The principles in favor of mulliganing have less weight if you have already mulliganed
The reasoning behind this is simple: the odds of getting a good hand drops with each mulligan since you have less and less cards to work with. What this means is that the two spells or fewer principle often does not apply for six card hands and the one land or fewer principle often does not apply for five card hands. As such, it would be correct to keep either of the following hands.

Hand: Sprouting Thrinax, Dragonskull Summit, Dragonskull Summit, Savage Lands, Broodmate Dragon, Swamp

Hand: Elite Vanguard, Ranger of Eos, Plated Geopede, Lightning Bolt, Arid Mesa

Of course, the odds of winning with either of the above hands are not particularly high and you'll have to get lucky to win. The reason why it is correct to keep them is because it is unlikely for you to get a better hand by Parising again.

Principle—Mulligan most of your color-screwed hands
Color screw is mana screw's slightly less evil cousin. Because it is slightly less evil, it only blanks out some of the cards in your hand. Since our hand is partly functional, we will need to consider additional factors in order to determine whether we should throw the hand away or not. These factors include the number of uncastable cards, the number of correctly colored mana sources we will need to draw to cast our spells, and how urgently do we need our mana. The following hand from Snepvanger's deck is a clear example of a hand that should be mulliganed due to color screw.

Hand: Scalding Tarn, Teetering Peaks, Elspeth, Knight-Errant, Plated Geopede, Steppe Lynx, Goblin Guide, Mountain

Exactly half of the spells in this hand are uncastable due to a lack of White mana. Furthermore, the urgency of needing White mana is compounded by the fact that Steppe Lynx is much more effective early than late. To top everything off, Elsepth is uncastable unless we are able to draw two sources of White.

Principle—Mulligan hands which are “virtually” mana flooded
Some hands have what would appear to be a good mix of land and spells but should be mulliganed anyway because they, for all practical purposes, contain very little action. The following hand from Saito's Jund Ramp deck is a good illustration of this point.

Hand: Rootbound Crag, Rootbound Crag, Verdant Catacombs, Rampant Growth, Rampant Growth, Siege-Gang Commander, Mountain

This hand is a mulligan because Rampant Growth, for all intents and purposes, should be characterized as half a land and half a spell. As such, this hand is “virtually” mana flooded and should be mulliganed due to the two spell or fewer principle outlined above. Don't let the accelerated Siege-Gang Commander fool you. If your opponent is able to deal with it with a Jund Charm, a Flashfreeze, a Day of Judgment, or even trade two for one with it (e.g. with a Lightning Bolt and Maelstrom Pulse), then you will be completely out of gas and in prime position to lose the game.

Principle—Mulligan hands with a lot of “dead cards”
This principle is similar to the principle above. Some hands, for all intents and purposes, should be mulliganed because they contain two or fewer “real spells”. For example, if you knew that your opponent was playing Joel Calafell's Jacerator deck (which can be found [here]), then the following hand from Snepvanger's deck is an easy mulligan.

Hand: Teetering Peaks, Path to Exile, Path to Exile, Kor Skyfisher, Elite Vanguard, Marsh Flats, Plains

The fact that this hand contains three lands and four spells is an illusion since Path to Exile is a dead card against the creatureless Jacerator deck. This hand, like the example above, is in effect a two-spell hand and should be mulliganed due to the two spell or fewer principle.

Principle—Land heavy hands are good in the Control Mirror
Hitting all of your land drops in the Control Mirror is important for two reasons. Firstly, having lots of land enables you to cast multiple spells per turn, which is important if you want to resolve your key spells, such as Ajani Vengeant, Cruel Ultimatum and Mind Spring, through any permission the opponent might have. Secondly, because control decks have a lot more answers than threats, having a bunch of answer cards in your hand is generally strictly worse than having lands. For example, if you were playing Philipp Summereder's Sphinx Control deck (which can be found [here]) against a mirror matchup, the following hand would be keepable.

Hand: Sejiri Refuge, Sejiri Refuge, Arid Mesa, Arid Mesa, Island, Flashfreeze, Jace Beleren

Even though this hand only has two spells, this hand is not at all bad. Jace Beleren and Ajani Vengeant are the two most important cards in the matchup and this hand has an answer to both of them. Furthermore, you should count yourself lucky that this hand has a bunch of lands instead of a bunch of useless spells like Wall of Denial, Day of Judgment and Path to Exile.

Principle—Sketchy hands are keepable if they contain an important card
Sometimes, a matchup depends entirely on whether you have access to a certain card. Under such situations, the fact that a hand contains that card will often outweigh any other considerations you might have in your mulligan decision. For example, if you were playing a Bant Threshold deck in Legacy against an opponent playing two-land Belcher, the following hand would be an automatic keep.

Hand: Misty Rainforest, Misty Rainforest, Windswept Heath, Tundra, Tropical Island, Brainstorm, Force of Will

This hand is keepable because two-land Belcher is a deck that is designed to either win the game on turn one by casting and activating Goblin Charbelcher or to set up the win by making an excessively large amount of goblins with Empty the Warrens. The presence of Force of Will and a blue card to pitch to it means that you should be able to stop their nefarious plan. The corollary of this principle is that you should mulligan hands such as this.

Hand: Brainstorm, Tropical Island, Tropical Island, Tundra, Sensei's Divining Top, Counterbalance, Tarmgoyf

This hand is a mulligan simply because it does not have a Force of Will. Note that it is even correct to mulligan this hand on the play despite the presence of the Brainstorm. The reason for this is because the percentage chance of seeing a Force of Will in the top 3 cards is only about 21%, while the odds of seeing a Force of Will and a blue card to pitch to it in a new six card hand is much higher at 34%.

Principle—Combo decks need combo pieces
When stated like this, this principle may seem blindingly obvious but you'd be surprised at the kind of hands that an inexperienced player who had just netdecked a combo list is willing to keep. Combo decks need certain cards to “go off” and if you don't have those cards, then your deck is basically doing nothing and you might as well be mana screwed or mana flooded. In other words, if you were playing Yuuya Watanabe's Dredge deck (which can be found [here]) please do not keep hands like this.
Hand: Dread Return, Misty Rainforest, Watery Grave, Island, Stinkweed Imp, Golgari Grave-Troll, Glimpse the Unthinkable

Dredge decks need to have access to two things to get going. Firstly, it needs some way of putting a card with dredge into the graveyard. Secondly, it needs some way to accelerate the dredging. Typically, this comes in the form of card drawing spells such as Ideas Unbound and Drowned Rusalka although having an Abundance of mill effects, such as multiple Hedron Crabs or Glimpse the Unthinkables, would accomplish something similar. The above hand is a mulligan simply because it does not satisfy the second requirement. Sure, the Glimpse will get a dredge card into your graveyard but where do you go from there? You do not have any options available apart from dredging one card a turn, which is far too slow to race two thirds of the available decks in Extended. Furthermore, taking a mulligan does not hurt the dredge deck in any significant way at all since half of the cards in the deck don't actually do anything relevant when they're in your hand. Indeed, the primary resource for the dredge player is not the number of cards in hand but the number of cards in the graveyard and so it makes perfect sense to give up a card in an attempt to get a hand that fills up the graveyard at a faster rate.

Principle—The presence of sideboard cards often makes a hand stronger
This is a less extreme version of the principle that sketchy hands are keepable if they contain an important card. Because of the impact that sideboard cards can have on the matchup, they are sometimes worth multiple cards. This means that sometimes a spell light hand, for all intents and purposes, should be thought of as containing a good mix of land and spells. The most obvious example of this is in the Rubin Zoo vs Affinity matchup, where the following hand is obviously keepable.

Hand: Ghost Quarter, Grove of the Burnwillows, Marsh Flats, Misty Rainforest, Sacred Foundry, Ancient Grudge, Ancient Grudge

The above hand is virtually a nine-card hand consisting of five lands and four instant speed Vindicates. When you think of it that way, it becomes easy to understand why such a hand should be considered a “good mix” of land and spells.

Some sideboard cards however, do not have the same impact on a matchup as Ancient Grudge and so should not be considered as being worth multiple cards. As such, if you were playing Snepvanger's Boros deck against Saito's Jund Ramp deck, you should beware of hands like this.

Hands: Marsh Flats, Arid Mesa, Arid Mesa, Teetering Peaks, Celestial Purge, Harm's Way, Ranger of Eos

This hand contains two sideboard cards which are quite effective against Jund. Celestial Purge is a nice answer to Sprouting Thrinax, which is their best defensive creature while Harm's Way is an excellent answer to cards like Pyroclasm and Jund Charm, which they are sure to side in against you. However, Celestial Purge and Harm's Way do not impact the matchup so much that they should be considered to be worth multiple cards. Indeed, Celestial Purge is just an efficient removal spell while Harm's Way is nothing more than a cheap Undermine. As such, I would mulligan the above hand simply because it has no early action (see below).

Principle—Mulligan hands with no Early Action
Tempo is another important resource in magic. If you keep hands that are severely deficient in early game interaction, then you not only give your opponent plenty of time to set up his or her game-plan, but you also severely increase the risk of losing to a fast creature rush. For example, the following hand from Snepvanger's deck is an easy mulligan.

Hand: Ranger of Eos, Ranger of Eos, Plains, Kor Skyfisher, Arid Mesa, Elspeth, Knight Errant, Mountain

This hand is obviously bad because it doesn't do anything until turn four since the Kor Skyfisher cannot enter the battlefield without ruining your tempo even further.

Of course, the definition of early action also depends on what kind of deck you are playing. For example, an aggressive deck like Boros would want to spend their opening turns summoning creatures to get in as much damage as possible. As such, they should avoid keeping hands like…

Hand: Scalding Tarn, Scalding Tarn, Lightning Bolt, Lightning Bolt, Lightning Bolt, Path to Exile, Plains

On the other hand, these exact seven cards would be an unspectacular but keepable hand if you were playing Summereder's Sphinx control deck. The reason for this is because that deck wants to spend its opening turns defending against the opponent's threats, and a hand with four removal spells is quite well equipped to do that.

Principle—Consider keeping speculative hands if your draft deck is outclassed
Sometimes, whether it is due to drafting errors, poor luck, or a mixture of both, your draft deck will turn out to be a very average pile that contains a lot of filler cards. In such situations, you'll often need to “get lucky” to win which means that it is often okay for you to keep speculative hands that give you a good chance to stealing a win if you are able to draw what you need. Take the following hand for example…

Hand: Greenweaver Druid, Oran-Rief Survivalist, Oran-Rief Survivalist, Vampire Lacerator, Vampire Hexmage, Hagra Diabolist, Swamp

As this hand will explode all over the table if I'm able to draw a forest and a third land within the first few turns, I will keep the hand on the draw if my deck is quite bad. Even though the chances of me hitting the exact cards I need aren't exactly high, keeping is the correct decision because it is extremely unlikely for me to get a six-card hand that is more likely to win the game than this one.

Principle—Strong sealed decks can keep average, land-heavy hands
Since sealed is a much slower format than draft, it is often okay to keep land-heavy hands since the likelihood of you being punished for it is quite low. If your deck is particularly strong and contains plenty of bombs, then more often than not, it will “get there” by itself if you have the mana to cast your spells. Take the following hand for example…

Hand: Cancel, Vampire Nighthawk, Swamp, Swamp, Swamp, Island, Island

This was a mulligan decision that I had to face in Grand Prix: Melbourne. My deck was completely insane. Apart from the Nighthawk, it also had Malakir Bloodwitch, triple Hideous End, a bunch of blue fliers, Bala Ged Thief supported by Umara Raptors and Nimana Sell-Swords, and a strong late game consisting of double Reckless Scholar and a Heartstabber Mosquito. Because my deck was so much better than my opponent's, I decided to keep the hand because I felt that my chance of losing the game with even a mana-flooded draw was smaller than the chance of me losing the game because my six-card hand had no land in it.

This principle, of course, does not apply to draft decks since the speed of the format is not conducive to letting hands develop by themselves. As such, I would not keep the above hand in a draft game unless…

Principle—Sometimes you have to go all-in
If your draft deck is pretty bad and you happen to have the best card in your deck in your hand, consider going all-in on that card and hope that your opponent is unable to deal with it. Indeed, the hand above would be a clear keep if there was nothing in your deck approaching the power of Vampire Nighthawk, especially since you also have a Cancel to defend it against cards like Burst Lightning.

Principle—Three color decks don't mulligan well
If your deck has pretty bad mana, then your average six-card hand will generally be worse than a bad seven-card hand that contains all your colors. Take the following hand for example…

Hand: Swamp, Swamp, Mountain, Mountain, Plains, Plated Geopede, Shepherd of the Lost

My second draft deck at Pro Tour: Austin was a complete trainwreck. It had no removal outside of a lonely Torch Slinger and it contained hits such as maindeck Bog Tatters and Unstable Footing, and double Lavaball Trap (awesome in sealed, unplayable in draft). It was also three colors with an 8/6/4 manabase. In one of the games, I was dealt the above hand and I decided to keep because with two swamps, two mountains and one plains, I could cast every spell in my deck. Of course, it also helped that the two spells I drew were both quality cards.

As it turned out, turn two Geopede followed by turn four Crypt Ripper and turn five Shepherd was good enough to win that game. Funnily enough, I also 3-0ed that draft with my three color pile since everyone else's deck was also really bad.

Heh, everyone loves a happy ending

Until Next Week,
Signing Out

submit to reddit » Print «