“A sunk cost is a cost that has already been incurred and cannot be recovered.” -Wikipedia

Sunk cost is a term that’s used by businesses and economists, and applies to your everyday life. As the definition says, it’s a cost that you’ve already paid, and that, no matter what happens, you will not be getting back. For all intents and purposes, that cost is gone—it’s sunk.

In theory, the concept of sunk cost is easy to understand. The problem arises when you start considering sunk costs as a way of making future decisions when they should, in fact, be completely irrelevant at this point. This is called the sunk cost fallacy.

In the real world, few people intuitively understand the sunk cost fallacy, and I frequently find myself explaining it in discussions with muggles (people who do not play Magic). Here is a common example on how the sunk cost fallacy manifests:

Imagine you buy a ticket to a rock concert—it costs a nonrefundable, nontransferable $100, and you’re planning on going. Then a couple days before it happens, you get invited to a special Disney concert that is free and happening at the same time. You’re a smart person, so you know Disney songs are always going to be better than rock, and the Disney concert is therefore superior. Here’s most people’s train of thought:

“The Disney concert would be better, but I paid $100 for the rock concert, and I’m not going to waste that.”

When, in fact, the thought process should be:

“The Disney concert is going to be better, so I’m going to go to that one.”

Whether you go to the rock concert or the Disney concert, that $100 is already lost—it’s a sunk cost. Therefore, it should have no bearing on your future decision to go to either one. The outcomes for the rock concert are, “you pay $100, and you attend the rock concert.” The outcomes for the Disney concert are, “you pay $100, and you attend the Disney concert.” In both cases, the $100 is paid. The only difference is which concert you’re going to attend, and that should solely be based on which one you think you’re going to enjoy more.

Though we Magic players understand those concepts more than muggles, we still fall prey to the sunk cost fallacy during our games all the time.

Magic is a  complicated game, and you’re always going to make mistakes, much like buying that $100 rock concert ticket clearly was. That part simply can’t be avoided, no matter how hard you practice or how talented you are. What can be avoided, however, is doubling down on your mistakes—making more mistakes because you made an original mistake.

Imagine you have a 2/2 and your opponent has a 4/4. You have a Lightning Bolt in hand, and if you attack and your opponent blocks, you’re going to cast Lightning Bolt to kill your opponent’s creature.

You attack, and your opponent blocks. At this point, you realize that maybe you didn’t want to trade Lightning Bolt for your opponent’s creature right now. Perhaps you had a 4-drop that you wanted to play instead this turn. Perhaps you realize that your best out to win the game is topdecking another burn spell and flat out killing your opponent with that plus your Lightning Bolt.

In this spot, 99% of the players would simply Lightning Bolt the creature because they already attacked, falling prey to the sunk cost fallacy. The creature was blocked—it’s already gone, the cost is already paid, and it’s dying no matter what. Since it’s a given of the situation now, it should not impact your future decisions.

In scenario A, the cost is a creature + Lightning Bolt + a mana, and the outcome is that the opponent’s 4/4 creature is dead.
In scenario B, the cost is a creature and the outcome is nothing.

Since the creature is the common variable, you can remove it from the equation. The cost for scenario A is just Lightning Bolt + a mana, and the cost for scenario B is nothing. The decision, therefore, is, “do I want to trade 1 mana this turn plus my Lightning Bolt, for my opponent’s 4/4?” The fact that you already attacked with the 2/2 should not factor into your consideration in any way. You should consider the fact that, if you want to trade Lightning Bolt for the 4/4, then it probably has to be now (you cannot simply make that trade in the future at will), but you still need to consider whether it’s a trade you want to make to begin with and, if you decide that you don’t, you shouldn’t do it.

The key here is that you made a mistake that obviously would have been better to think about before attacking. But that’s already in the past, and it cannot be avoided. The future mistake of trading the Lightning Bolt for the creature, however, can be. All it’s going to cost you is a little bit of pride.

A lot of people subject themselves to the sunk cost fallacy because they do not want to look bad. Attacking a 2/2 into a 4/4 is a bad play—it’s embarrassing. If you Bolt the creature, then suddenly your play makes sense—you might be wrong, but at least you had a reason to do it, and chances are no one is ever going to know you were wrong. If you attack and then do not Bolt it, then it’s going to be crystal clear to everyone—including your opponent—that you made a big mistake.

This is a tough spot to be in, and your inclination is to not flaunt your mistake. As such, we tend to go with the original plan—the one for which we already paid the costs—because it’ll look much better. But if the play is worse, you have to suck it up, admit the mistake, and make the best play you can moving forward, even if that makes you look bad to everyone else. You already made one mistake —that can’t be helped—but you can stop yourself from making another one by doubling down on it. If your goal is to win, then it’s better to make one mistake that everyone knows about than to make two that stay hidden.

This situation is more common than you think. Imagine, for example, that you’re playing against a blue deck in Standard. It’s turn 6, and you play a Glorybringer. Your opponent plays Censor. You stare at your land in hand like the idiot you are.

After this, you have a decision to make—either you play your land, or you don’t. The Glorybringer is already gone, though—you’re never getting it back. In this spot, those are the two possible situations:

A: You play your land. The cost is the Glorybringer, plus you don’t have that land in your hand—the outcome is that you have it in play.
B; You don’t play your land. The cost is the Glorybringer and there is no outcome.

Since the Glorybringer is the common variable, it should not factor into your calculations at all. In scenario A, the cost is the land in your hand the the outcome is the land in play, and in scenario B there is no cost and no outcome. This is the only decision that you should be making. A lot of players would think, “given that I’ve already not played the land…,” but that’s irrelevant.

In this spot, perhaps not playing the land is better. You can bluff something, or discard it to an effect later on. But perhaps playing the land is better, and the fact that you didn’t play it before (and already paid the cost for your mistake) should not factor in at all. Is your opponent going to internally laugh at you for playing the land post-Censor? Yes. Are they going to loudly laugh at you in your face for playing the land post-Censor? Quite possibly. Should that ultimately matter to you? No.

Some players will take the land to the grave with them—no matter what happens, they will not humiliate themselves in this manner, and they will never give the opponent the satisfaction of knowing they could have paid that 1 mana. They will pass without playing the land, and then the following turn they will be unable to pay for Supreme Will because they missed a land drop. In their haste to hide their mistake, they will end up committing another. If your goal is to win the game, “humiliation” should not be a factor.

It’s the same when you consider attacking or blocking with creaturelands. Imagine that you animate a Mutavault, and before attacking you realize that perhaps the attack is not so great after all. In this spot, most people would attack, because they animated it—they would think, “given that I already animated it instead of playing my creature, I’m going to attack.” This would be wrong.

In this spot, you spent a mana to animate the Mutavault, and that mana is already gone—it’s a sunk cost and you’re never getting it back (well, you are next turn, but you get the gist). The only decision left to make is, “Do I want to attack or not?” and the cost that is the common factor in both scenarios (the animation cost) should not even be a part of the decision. I have personally performed the “animate Treetop Village, pass” maneuver into a tapped out opponent, and while I looked ridiculous, I felt it was better to make one mistake that everyone saw than two mistakes that only I knew about.

The sunk cost fallacy also manifests itself in deck selection. Imagine a spot where I am a proficient player with one deck, Faeries. I’ve played Faeries a lot, I can pilot it at a 90% confidence level, and I already own all the cards.

Now for a specific tournament, I decide that Faeries is bad and that I’m going to learn another deck: Storm. Storm is a hard deck, and I spend a month practicing with it to the point where I can also play it with a 90% confidence level. On top of that, I spend a lot of money buying Storm cards.

Then fast-forward to the day before the tournament, when I figure out that the metagame will be different than what I expected, and Storm isn’t actually a great choice. At this point, though, I’m locked into it, because I already spent all my time practicing it and all the money to buy the cards, right? Wrong!

In this spot, if you think Faeries is a better deck, then you should still switch. Sure, if you hadn’t tried Storm then you would have perfected Faeries more, so you’d be playing it at, say, 95% confidence. If you hadn’t bought a Storm deck, you could have foiled out your Faeries deck. But you didn’t, and that mistake is already made. Playing Storm at this point would just be adding another mistake to the list.

In scenario A, the cost is a month of your time + the cards for Storm, and the outcome is that you play Storm at a high level.
In scenario B, the cost is a month of your time + the cards for Storm, and the outcome is that you play Faeries at a high level.

The cost in both of those situations is the exact same because it’s already sunk. Given that it’s the same, then it should not be a factor at all in your decision. It’s better to play the better deck and have “wasted” this month and this much money than to make the inferior deck choice to feel justified, because the month is already used up and the money already spent.

In the end, of course, everyone is going to do what they think is best. For some people, it’s more important that no one knows that they made a mistake than to actually win the game, For some, it’s more important to use the new cards they bought than to play a better deck, and so on, and that’s fine. I write things such as, “you shouldn’t” and “you should,” but that’s under the assumption that winning is your main priority. My goal here isn’t that you do exactly as I do, but that you understand that those biases exist and that they are affecting you, so that when you make a choice in this regard, it’s a completely conscious one.