Why Players Struggle at Card Games and What Actually Drives Winning

mean, really awful. But is there any way to be good, or am I doomed by my bad luck? Hello, internet. Welcome to Game Theory, the show that reminds us of when cards were for playing and not for collecting and selling on eBay. Kenneth Slime, there's an offer. Seriously? Already? Yes. Okay, wait. What's the offer? 620,000. Fortunately, we're recapturing that childhood nostalgia with a new show coming soon to Theory Burst. Are you bored yet? Where we will play board games and card games like Magic the Gathering. But here's the thing: Santi and Lee didn't invite me to that episode, and it's all because I'm terrible at card games. I know I made the joke before that cards were for playing and not for collecting, but as a kid

I collected a lot and I didn't have any friends to play with. So imagine the childhood trauma that resurfaced when I discovered that the friends I have now didn't want to play with me. And I missed all this exciting action. Okay. I'm just kidding. I promise the show is much more fun than that clip suggests. But of course, after this betrayal, the first thing I did was rewatch the video Matt made 6 years ago, where he broke down the math of card games to become a better player. But, despite that, I'm still having difficulties. So when our friends at Marvel Snap asked us to sponsor a portion of today's video, I thought it was the perfect time to take Matt's video and take it a step further. If your video was card games 101, this one is

card games 201. We're going to analyze the probability, psychology, and decision-making that separate a lucky player from a truly great one. So, start opening those packs, loyal theorists, because we're going to gather information on exactly what makes a great CCG player. As I mentioned before, Matt made a video about how to improve at card games several years ago. No, I'm not going to mention the number again. It makes me feel old. That video basically boils down to one thing: math. When you play card games, you constantly have to think about probabilities. If you play a card game with a regular deck, the odds remain the same, whether you play Go Fish or Blackjack. In addition, there are many books that teach you how to count cards, so you can become a

poker pro in no time. But when you play card games like Pokémon, Yu-Gi-Oh, Magic, or Marvel Snap, things get a little weird. Each game allows you, as a player, to build your own deck, which means you choose the probability of a specific card appearing. If you put four of your favorite cards in a deck, they are more likely to come up, and so on. But, to take this card course to the next level, we must delve deeper into the psychology beyond the player's gestures. Real scientific theories that can take us from being a beginner to a Pokémon master. And to do that, I want to talk about

Bayesian inference. This concept is vital to being good at any card game. It combines two things that video game nerds love: prior knowledge and facts. The core of this idea is Bayes' theorem, which I will display on screen right now. At first glance it seems complicated, but it can be broken down in less mathematical ways. Essentially, you start with a prior belief. Then, you take the new data into account. So, that new data makes you update your beliefs. Frankly, it's less a statistical theorem and more an advanced learning technique. All you have to do is apply it when playing card games. Remember what I said about how in each deck, some cards are more likely to appear depending on how often they appear. But, as you draw cards, those

probabilities are updated and change. They do not remain static. So now you need to update your own mental dataset. The longer the card game lasts, the more accurate your predictions will be. For example, let's say you're playing Pokémon. Before the game, you'll want to study the current meta to find out what 's popular and what the latest releases are. There are usually very popular decks in the current meta that many players tend to use. You might even know your opponent and what decks they're using. Whatever the case, that knowledge serves as your prior knowledge when entering the game. Then, when they play their first card, that's your first clue about the deck they 're actually using. Imagine

this: you and your Pokémon opponent have placed your starter Pokémon cards. Then you turn over those initial cards and see that your opponent has revealed Charmander. That means your opponent is now using a deck with a Charmander, which, yes, seems like a pretty obvious observation. But that means all popular decks that don't use Charmander are ruled out. This greatly reduces their possible next moves. The game progresses, they show you more and more of their deck, which helps you update your perception. Are you going to use a Charizard X Entei deck or a Charizard Y Moltres deck? However, I do n't have those letters. I never had them as a child. I know. I'm sorry. The dream has been shattered. I don't have all the

Pokémon cards. But eventually, you'll make very well-founded guesses about what their next move will be. Not because you have a special psychic ability, but because you constantly incorporate new evidence and think one step ahead, hoping to outmaneuver your opponent at every turn and helping you achieve victory. Once you have mastered all that probability, you can start to delve into psychology. This is what separates the statistician from the professional. Sure, you can play with fractions all you want, but being good at card games is more than that because, well, I know how to do math and I'm still terrible at card games. So, one thing we can observe is the signal detection theory.

Signal detection theory helps to measure decisions made under uncertainty. Specifically, decisions with two possible options. In normal situations, it would be easy to respond to a signal you receive. Either you see the sign or you do n't. But maybe it's a foggy day. Perhaps you can't see clearly or there's heavy rain. That's where uncertainty comes in, expanding those two options to four possible outcomes. You respond correctly to the signal, that's a success. You don't respond to the signal, but the signal is there; that's an omission. If you don't respond to the signal and the signal isn't there, that's a correct rejection. And finally, you respond to the signal and the signal isn't there, that's a false alarm. And just like with

Bayesian inference, signal detection may seem obvious at first, but it has many real-world applications. For example, of course, all decisions have consequences; that's simply the reality of making decisions. But you can use it to analyze your own personal decision-making. The theory itself was developed in the 1940s during World War II. The researchers were trying to better understand how radar technology worked. Those decisions, whether or not to respond to a point on the radar, could be a matter of life or death. But sometimes those points were just random noise. So signal detection theory was invented to help make those systems more accurate. And you can use it too to get better at Yu-Gi-Oh! Yes, that problem seems a little small

compared to World War II, but bear with me. Let's say you 're playing Yu-Gi-Oh! Your opponent has placed a card face down, which could be the dreaded Mirror Force. No, I know this isn't Mirror Force. It's the closest thing I have. I never got it because it's such a good and devastating card that it's hard to obtain. In any case, it's a card that, when activated, reflects your attack and destroys all monsters on your side of the field. It is a devastating letter, and therefore, signal detection theory shows us four results. They put Mirror Force on and you don't attack. So, you didn't overextend yourself and now you can figure out how to attack safely later. They do have Mirror Force, but you attack anyway, which ends up eliminating

your side of the board. They don't have Mirror Force and you attack, correctly identifying that it was a bluff and subtracting life points. Or they don't have Mirror Force, but you don't attack, which means you wasted your turn and gave your opponent more of an opportunity to gain the upper hand. This decision is obviously made much more difficult because there are many other things that affect the field. Are you attacking a monster in attack or defense position? Or are you going straight for life points? How many monsters would you lose if you guess wrong? Are any of them particularly high-value summons? There are also thousands and thousands of trap cards, so it may not be Mirror Force, but it could be something

just as bad or worse if it's triggered. In addition, there are other things that could influence the decision beyond strategy. Does your opponent have any telltale signs? Do they seem overconfident or annoyed with their first hand? If you play in public, you might be distracted by the loud noise of the place. If you're at home, maybe your dad comes in halfway through the game and asks, "Are you winning, son?" All these things influence the uncertainty of every decision you make, making the game more difficult than it already is. And in these questionable circumstances, you might be inclined to make one decision or another without even realizing it. If you end up tracking your data, you might realize that you tend to use more cards

in an attacking position, which means you have a liberal bias. If you tend to put more cards in a defensive position, you might have a conservative bias. None of these are bad in themselves, but you should try to change your decision-making to optimize utility. Expert players don't try to eliminate mistakes completely, but rather choose the least costly option. And that really relates to another decision-based theory I want to talk about: prospect theory. Essentially, you have to ask yourself a question. Am I making the mathematically best decision or the one that is emotionally more comfortable? Yes, we 're getting all therapist-like around here. The theory basically analyzes in more detail how people make

decisions. It assumes that people really dislike losing. They avoid losses like the plague, and their hatred of losing is actually stronger than their love of winning. This is also known as loss aversion, and in general, humans are less likely to take risks to gain additional profits, but are willing to take risks to avoid further losses, even if the odds indicate otherwise. So, every day you have to actively work against that theoretical bias. If you're playing the new draft mode from our sponsor Marvel Snap, you'll probably encounter this phenomenon. In case you didn't know, Marvel Snap is a digital collectible card game where you can build a deck with all your favorite Marvel characters using cubes as victory currency. And at any

time you can use the "Snap" to double the number of cubes you bet on your win. But now, they're introducing a new draft mode to the game. Instead of bringing a pre-built deck, you start with a smaller one that you use during the game. After each encounter, you select an additional card and apply modifiers that improve your existing cards before fighting again. Your goal is to build an increasingly stronger deck over time. And you need to get five wins without losing twice. If you lose twice, you're out. Since you are only allowed two losses, every decision you make carries considerable weight. And you might take more risks than normal just to avoid losing. You have to ask yourself personally: is the next move the

mathematically correct decision, or are you just afraid of suffering another defeat? If you really want to win, you have to work against loss aversion and make decisions you would n't normally make. Your opponent will likely try to minimize those losses, which opens the door for you to take more risks to gain and move forward. It might seem like the wrong thing to do given the low risk of losing, but psychology says it's what will end up working. And in fact, you can check it out for yourselves right now by clicking on the link in the description or scanning the QR code on screen to download the game. Remember, card games are based on resource management, and information is a resource.

Constantly update your data and draw new conclusions. Change your decision-making process to optimize utility and don't be afraid to make uncomfortable choices. Maybe Santi will invite me to his next magic game now, but hey, that's just a theory. A game theory. Thanks for watching the video. World premiere on game theory. Will I become the new world champion of magic?