Last week’s article ended with a cliffhanger. It was game 3 of the finals of the Magic Online Championship playoff tournament, and I was on the play in the Temur mirror match with this opening hand:

The question was first, should I keep this hand? And second, if I do, how should I play it?

I received great feedback in the comments section, with answers spanning what are essentially all five available options. What makes this situation so remarkable is that all five available options seem close enough in quality that I’m unable to rule out any of them. Let’s go through the possibilities:

Option 1: Mulligan. I need to draw a mana source to operate, and can’t produce all of my colors with just my opening seven cards. There’s a relatively large failure rate, and throwing this one back falls in the “reasonable” range of the mulligan spectrum.

Option 2: Keep and Attune for Mountain on turn 1. This sets up my best turn-2 play (Harnessed Lightning), and gives me the best chance of curving out perfectly.

Option 3: Keep and Attune for Island on turn 1. This gives me a high chance of casting Rogue Refiner on turn 3, which is the card that will really get the wheels turning for me. It also keeps the door open for turn-2 Harnessed Lightning if I draw an untapped red source, or turn-2 Servant of the Conduit if I happen to draw it.

Option 4: Keep and hold Attune with the intention to search for a Mountain on turn 2 if I fail to draw a land. Waiting on Attune gives me the highest chance of assembling all three colors of mana. If I draw Botanical Sanctum, I can search for Mountain, and if I draw Rootbound Crag, I can search for Island. The advantage of choosing Mountain over Island is that I can at least cast a spell if I miss my third land drop.

Option 5: Keep and hold Attune with the intention to search for an Island on turn 2 if I miss. The advantage of choosing Island over Mountain is that I can realistically win the game without casting Harnessed Lightning right away, but I will almost certainly need to cast Rogue Refiner if I’m to have a functional hand.

Since this was Magic Online, I had no need to fear slow-play rules, and I sat with my opening hand for about five minutes before deciding what I would do. When I finally made up my mind, I did so without full confidence that it was right. More than that, I knew full well that many players as good or better than me would disagree with my choice.

I chose option 4. I kept my hand, played a Forest, and passed the turn. (Two details that some might consider relevant are that I had sideboarded out Longtusk Cubs, and that my opponent mulliganed to 6 and scryed to the top). I drew a brick, Attuned for Mountain, and played it. My opponent had no 2-drop, and then I drew Botanical Sanctum and was off to the races.

I’ve already summarized what I found appealing about this option. Declining to cast Attune on turn 1 slightly increases my chances of finding a land on turn 2 because I do not thin my deck of a land before my draw step. But this effect is marginal, and it’s not the reason that I chose this line of play. Waiting until turn 2 gives me a chance to draw a land, which would inform my decision and increase my chances of finding all three colors of mana.

The reason I chose to search for a Mountain instead of an Island is that a Mountain is more helpful if I fail to draw a land on turn 3. If my opponent plays Longtusk Cub or Servant of the Conduit, I can miss my land drop and Harnessed Lightning it, and that’s still a game I’m capable of winning if I hit my land later on. Alternatively, if I’m forced to pass the turn with Forest, Island, and get hit by a Cub (or face a 4-drop ahead of schedule), I’m unlikely to come back.

Planning for the Worst

We encounter many situations like this in MTG. It’s unclear what decision is best, and the possibility tree is too complex for a human being to work through and find an answer. So what should you do? You can’t simply be paralyzed with indecision, and you can’t concede the game. You have to pick something.

Paulo Vitor Damo da Rosa wrote an excellent article where he described the use of heuristics to aid your game play. A heuristic is a tool for problem solving.

I hesitate to use the same term as Paulo since his article focused more on “default” or “go-to” plays for certain situations, whereas what I’m about to describe is even more nebulous and unscientific. But whether you call it a heuristic, a guideline, or starting point, the concept is the same.

When I’m analyzing a complicated situation, one of the first things I consider is the worst-case scenario. I start by thinking about how badly things can go, and I progress to thinking about how I can prevent things from going wrong. Following those lines, I can work towards a satisfactory play that avoids catastrophe (whenever possible). Also important, it leads to being surprised less often because I’m not blindsided by something that I didn’t consider.

This philosophy is closely in line with “Play Around Everything,” which is a phrase that was used in Magic strategy in the days when Faeries and Caw-Blade reigned supreme. But planning for the worst focuses not just on what the opponent might have, but also on combating the elements of randomness in the game.

An example for the Legacy players: You’ll see me shuffle with my Ponders a lot less frequently than most. I only want to shuffle with Ponder if it’s clear that I can’t win the game if the three cards remain on top, or if they’re close to the worst three cards I could draw. The reason is that I prefer an average situation (or even a slightly-below-average situation) that’s controlled over one that’s random. What can happen when you shuffle your library? You can draw the worst card. You can even draw the three worst cards in a row! If I Ponder in a topdeck situation, I’m going to happily draw into a spell-followed-by-a-fetchland rather than shuffle and try to get lucky.

Keep in mind that planning for the worst is a starting point for decision making, not a necessary conclusion. In many situations, a more aggressive line of play, or one requiring a little bit of luck, will clearly give you a higher chance of winning the game. And it’s important to be able to smell blood when the situation calls for it. Planning for the worst will always have you consider the possibility of your opponent topdecking Entrancing Melody in Ixalan Limited, but if that leads you to the conclusion that you should never cast your creatures, then you’re clearly lowering your win rate. Treat it as a tool, but not as gospel.

Planning for the Worst in Deckbuilding

A line you’ll often hear from me is, “I’d rather cast my spells a turn late than not cast my spells at all.” (Indeed, you can see this philosophy in the way I played my Attune with Aether in the above example.)

I always lean toward conservative mana bases, and when I face a close decision such as whether or not to include Evolving Wilds in my 2-color Booster Draft deck, I love to get the extra colored source in there. I’m more likely to cut a color from a deck than to add one, even when the cost is relatively low, as in the case of Temur Black in current Standard. Even though “5-Color Green” is one of my favorite Limited archetypes, I’m the guy building G/w/b with 10 basic Forests, and not the guy splashing Rags // Riches in his Naya deck.

When building a control deck, my first priority is to pack enough card draw into the deck that it’s virtually impossible to lose to mana flood. If I ever get to turn 7 on a stable board and lose from that position, I’m either making a dramatic change for next time, or scrapping the deck.

When building midrange, I choose creatures that can either provide card advantage or neutralize opposing threats. When I draw Grim Lavamancer or Courser of Kruphix in Modern Jund, I can use them as tools to dictate the course of the game. When I draw a Goblin Rabblemaster, all I can do is try to race.

For every card I consider adding to a deck, I run it through two tests: The first is, “How good is this card when I’m losing?” The second is, “How good is this card when I’ve drawn too many lands?” In other words, Pious Interdiction is a high impact card that can keep me in the game if I’ve drawn six lands and three spells by turn 4 of the game. I can even feel okay keeping an opening hand with five lands, Pious Interdiction, and another reasonable spell. On the other hand, Vampire’s Zeal and Cobbled Wings are low impact cards that are going to look fairly awful if I’ve drawn them as one of only three spells in a game where I’m flooded.

Play and Draw

Choosing to play first is the default in Magic. When both players hit their lands drops and curve out well, the player who goes first starts with an advantage and never gives up that advantage. Speaking generally, the better both players’ hands are, the larger an advantage it is to play first.

But we all know that games where both players have good hands only make up one slice of the pie. People mulligan, they miss land drops, and they get flooded. When these things happen to one or both players, it can be to your advantage to be on the draw.

One way of looking at things is that a player doesn’t start playing until they make a few land drops and cast a spell. In the games where that player’s opening hand doesn’t have enough lands or an early play, they get to start playing half a turn sooner when they’re on the draw.

It’s still usually correct to play first, but when I’m not specifically worried about the game ending with a fast curve-out, I do consider choosing to draw.

To put things in perspective, I choose to draw less than 1/3 of the time in Sealed deck, less than 1/5 of the time in Booster Draft, and less than 1% of the time in Constructed.


One spot where planning for the worst is particularly tricky is in mulligan decisions. Many hands, like ones that don’t have enough or the right kind of lands, have a failure rate, and sometimes I’ll keep them and hope for a little luck.

The point that I want to emphasize is that you take a risk both ways, either by keeping a speculative hand or by mulliganing. You can either keep and hope that the top of your library is good to you, or you can mulligan and hope that the top of your library is good to you after you shuffle up. Whenever you consider mulliganing to six, remember that you’re taking on the risk of possibly getting something even worse.

The type of hand that planning for the worst leads you to keep would be something like five lands and two spells, or something that’s missing a creature drop on the curve.

Conclusion and Disclaimer

You may have noticed that I’ve framed all this as, “I do it this way,” and “I think about it like this.” I haven’t written this article to be an instruction manual or to tell anyone that they have to approach the game the same way I do.

Planning for the best is becoming an increasingly useful skill in Magic. As the games get faster and the cards get more powerful, it becomes more likely that the player with an early advantage is the one who goes on to win the game. Aggressive mulligans, drafting decks that like to curve out, and knowing when to throw an Aura on your creature are all useful skills as well.

It’s to your advantage to know when the pendulum is swinging too far in one direction. If you only know how to plan for the best, you’re putting a ceiling on your win rate because might not have the tools to win the games where your draw is below average, or when the top of your deck doesn’t deliver right away.

All I recommend is that you use planning for the worst as one of the tools in your toolkit. When you face a difficult situation, it should be one perspective from which you look at the problem. You can still choose to play first, you can still mulligan looking for a 2-drop, and you can still cast your Glorybringer into a possible removal spell. Just don’t miss the spots where playing safe is the best way to play.