stute readers of this column have no doubt noticed patterns in the way I've broken down different game situations and the advice I've given. Magic is a game with limitless possibilities and strategic depth, but the same basic principles can guide you through those situations. Your goal is victory, and with that goal in mind each decision comes down to evaluating how each possibility changes the course of the game and your probability of winning in each possible future. There are some staple tricks that reliably help assist that cause, like noting cases in which you always win or always lose no matter what you do. Today, I want to take a step back and take a look at the concept of risk, because I think it is a piece of that puzzle that it helps to spell out. I will be talking about ideas in deck construction because that is an easier frame to put on the issue and make it clear what I'm talking about, then I'll bring things back to a major in-game question that I haven't dealt with properly yet.
In a game of Magic, as in any endeavor, there are a lot of things that can go wrong. For each thing that can go wrong, there are likely things that can be done to minimize or eliminate that risk. If you never want to be mana screwed, you can add more lands. If you want to always make your two-drop, play more of them. If you don't want to be caught without removal for a big creature, you can save it rather than use your removal on a small creature. You choose which cards to play around and what to guard against.
Magic is a card game that makes it clear that you have to take risks. There's no way to win every game or guard against every possibility. The process of choosing what you're going to guard against begins during deck construction, both in Limited and in Constructed. Looking at it during deck construction allows you to see a more "pure form" of the dilemma, but it is no different over the board. To start with the simplest example, everyone must decide how much land to put into their deck and which type of land. Today, choosing a set of lands isn't easy. You have a lot of good choices, but if you're playing more than two colors, then you have more of each of them available than you would be wise to use, especially once Dissension is added to the mix. Suppose you're playing White, Red and Blue. At this point, you can have up to twelve Ravnica dual lands, up to twelve Painlands, up to twelve Karoos and then finish up with basic lands, Signets and other goodies.
Each of these has its advantages and disadvantages. Those disadvantages become bigger as you play more of each type of land. If you play one Ravnica land, chances are good you can find a turn where you can safely put it into play tapped. If you play too many Ravnica lands, you'll end up having to pay two life more often when putting each one into play, and that life adds up fast. Painlands are great in the short term, but if you have no pain-free way to get your color later in the game, that also adds up. Karoos risk slowing you down at the wrong time, and against some decks like Owling Mine they are all but unplayable. Signets risk you not getting to two mana or not having a good time to get them on the table. Basic lands are solid, but don't help with your color. Most decks will end up with a mix of different types of land, to minimize each risk. The big mistake many players make is that they stop using one of those lands when another copy would make its drawback matter rather than when it isn't worth it anymore.
What does that mean? A player will add the first four painlands, and see that they're good, but then when he goes to add the fifth he notices he starts to get draws with too many painlands, so he takes it out. He tries out Ravnica lands, and notices that after about six he starts to get some draws where he has to pay too much life, so he cuts back down. It's easy to add the easy copies of cards, the ones you're always happy to see. Everyone loves the first Meloku, but it took a while before some brave people realized that you could play four of them. In all of these cases, each copy is worth less than the one before and is more expensive to buy. If you stop doing this where you're sure you have a good buy, you probably won't buy enough.
One reason this happens is that players think that they “shouldn't have to” take certain risks. The risk of not getting your colored mana is accepted because it is unavoidable, but the risk of too much pain from your lands is avoidable. They don't seem like they are the same, but they are. Any time you sacrifice something important to shrink a risk to zero or virtually zero, chances are you're overpaying to do that. The reverse of that is also true. Players think of some risks as “acceptable,” mostly because they're used to them, and it doesn't occur to them that it might be worth trying to buy their way out of them.
This phenomenon happens all the time during play as well, but I noticed something very interesting about the examples I came up with when trying to nail down the issue with a game situation: They all seemed to point back to decisions made during deck construction. That isn't to say that it always has to come from there. It certainly doesn't, but that is where you create the choices you'll have to make later over the board. It gives you the pieces you'll use to play your game, and it tells you what the chances are of different things happening later in the game. Knowing your opponents' deck tells you what pieces he'll be playing with. The more complicated decisions get, the more they involve intimate knowledge of both decks, or finding out as much about them as you can. You can't separate one part of the game from the rest of it.
That leads into what is probably the biggest subject that I haven't tackled systematically yet - the use of spot removal, especially early in the game against fast creatures. If you burn everything you have to try and never get hit, you won't have what you need to deal with the bigger threats that come later, but if you hold back, then you risk falling behind and taking needless damage. These come down to probability calculations when the decision is close. What are you likely to draw, and what is likely to be coming at you? Will they give you a good target next turn, or will you simply end up taking damage and then doing what you would have been better off doing last turn? What's the chance that you're dangerously stranding yourself? Knowledge of the decks is vital.
The first rule I have is that whenever I choose not to kill a creature that's doing something useful in a Constructed match (and attacking me is certainly useful), you need a reason. The default is to kill it dead. In Limited, the reverse is true. There the default is to let everything live unless you have a good reason not to. A Constructed deck with very little removal can be treated as midway between the two: Make sure it counts, but don't worry about burning everything you've got. If you're terrified to use your removal, then remember: If brute force doesn't solve your problem, you're not using enough. Why not use more?
Having a reason to not kill something, assuming that reason isn't just that your mana is busy doing something more important, means having a plan for how the game is going to go, at least over the next few turns. You could have big creatures coming that will make the creature you don't kill irrelevant, you could use mass removal. Those and similar reasons are good. The danger comes when overthinking about future threats, so here's how I think about that. The last removal spell you have in your hand should be reserved for something that you find unacceptable to have on the board. If you don't mind taking one damage a turn, and you don't have a use for your mana for the next few turns, don't burn that Lightning Helix
or Last Gasp
on a one-power creature. If you don't mind two damage, and there are better targets, hold out for better as long as you don't mind. If you do mind, go ahead and use just about anything. If he has something in reserve you like even less, you'll have to live with that.
However, the second and subsequent removal spells are different. Here you need to be actively happy with the state of the game to let something live. The more you have of any resource, the more willing you must be to spend it, to balance the risk of running out and the risk of being stuck with too much or not getting the most out of your resources. This is especially true if you are not yet sure that such spells are guaranteed to be good against your current opponent. But it's important not to get greedy either way.
That sounds a lot like the way I said to decide how to use mass removal, doesn't it? Well, that makes a lot of sense, since the two problems are different mostly by a matter of degree. Mass removal's key quality here is that it can have a huge impact when used at the right time and acts as a safety valve. Spot removal is the same, only not as extreme: You can take out a big creature instead of a small one, or kill a creature that wasn't naturally doomed to die, or a creature that has a particularly nasty ability. Again, the two biggest considerations are what you consider an acceptable board position and how much you can afford to "give back" out of your advantage in the long term to stay alive in the short term.
When looked at in general terms, a lot of Magic principles and rules of thumb I've developed over the years start to look a lot like each other – because they are. The essence of my approach to Magic is to streamline my ability to weigh my options by giving myself easier ways to evaluate what is important and what different things are worth to each player. The framework is more important than any individual section of the game, the same way that that element is more important than knowing how to play any individual deck or match.