Leveling Up in Hearthstone

I really enjoy card games of all varieties and for this post, I will talk about a game I play regularly, Hearthstone, and some interesting aspects of what constitutes a good Hearthstone player. Hearthstone is Blizzard Entertainment’s online collectable card video game, similar in concept to Pokemon TCG, Yu-Gi-Oh, or Magic: the Gathering, in which you play against one other person and win by achieving a certain objective, in this case reducing the opponent’s health to 0. There are two game modes: Constructed, in which players choose cards for their decks from the entire card pool, and Arena, in which players are randomly offered a small set of cards to choose from. The gameplay rules are identical for both, and thus many of the basic skills are transferable, but the experiences are vastly different. In Constructed, players can expect to encounter the same decks with the same cards over and over, whereas in Arena, every deck is random and thus, every game is different.

From my description it should not be surprising that I am an avid Arena player who occasionally dabbles in Constructed (mostly for the end-of-month rewards). What I love about Arena is how each game is like a new puzzle or maze to navigate through. There are countless decisions to make throughout each game, and you can never be exactly sure what the best play is, kind of like life. At this point, I have become a machine at Arena and will reflect upon some of the level-up moments that helped me get to this point. Whether or not these lessons extend appropriately into the arena of life (I’m terrible) is left as food for thought.

I am going to assume players know the rules of the game and have some experience with card games in general. Obviously, the more experience one has the faster one will pick up Hearthstone. And some people are more talented than others, etc. But Arena is not terribly difficult to become competent at, and so I will ignore innate differences in ability and speak more generally.

The first step to learning Arena is to simply get a feel for how games play out, what cards are good or bad, what tends to work or not work, etc., and this knowledge is gained mostly through experience. Very seasoned Constructed players will almost always struggle with Arena in their first few attempts because they lack general knowledge about the Arena landscape. Fortunately, this knowledge is accumulated over time, almost without effort, as long as you play when lucid.

Once players have some general knowledge of the game, and thus have some notion of what to do, the next step is to actually become a consistently winning player. I think the single biggest trait that separates the vast majority of the player base from those who have the potential to become really good is the ability to plan ahead. Most players make decisions by only considering what’s in front of them, but great Arena players always plan for at least a turn in advance, if not more. For me, this specifically entails envisioning how the board state will develop on my opponent’s turn, and how the board will appear when it gets back to me for my next turn. Interestingly, this process is not difficult at all; it simply requires imagining several small steps for a couple of different scenarios, nothing that a normal person cannot carry out. I strongly believe that anyone can learn to do this. But interestingly, some people are much more inclined to this than others, and the ability is not necessarily correlated with general intelligence. Rather, it’s a type of game shrewdness that many intelligent people may not be predisposed to, although it can be learned over time.

Consistently planning ahead is a trademark of above average players, but the truly great players take this a step further by always trying to maximize their chances of winning. This sounds obvious, but the application is quite nuanced because humans are not rational decision makers. For example, people tend to take the risk-averse route, even if the riskier choice has a slightly higher expected value. People also tend to be overly optimistic when evaluating their own situations. The way this plays out in Arena is that players often choose to play around the worst-case scenario in order to not lose the most, or evaluate cards based on their highest possible upside, which may require jumping through hoops to actually pull off. These tendencies may serve a useful purpose in life, but in the context of Hearthstone they more often than not decrease our chances of winning. Being able to recognize and then overcome these habits distinguishes the great players from the rest.

I want to provide a practical example of how this works. If you play card games often, then you’ve probably heard the expressions “play to your outs” or “play to win, not to not lose”. These phrases are a manifestation of playing to maximize winning. When faced with a difficult decision, simply try to maximize your chance of winning the game (if you like math, think of it as expected value). If you have two or more choices, you can draw a decision tree, make a table, or do some quick math to figure out which option is best. For example, suppose you are in a winning position and are trying to decide whether to play around a specific card. If the opponent has the card and you don’t play around it, you win 10% of the time, but if you do play around it, that number shoots up to 50%. If the opponent doesn’t have the card and you don’t play around it, you win 80% of the time, and if you do play around it, you only win 60% of the time. Now suppose you can estimate that the opponent has a 30% chance to have that card. If you don’t play around it, you will win (0.3)(0.1) + (0.7)(0.8) = 59% of the time, whereas you will win (0.3)(0.5) + (0.7)(0.6) = 57% of the time if you do play around it. If you knew the probabilities perfectly, then the correct decision would be to not play around the card.

The example above was very specific, and in an actual game, you would never be able to pinpoint probabilities with reasonable accuracy. To complicate matters, each decision has to be made in the 90 seconds allotted for each turn. In reality, I consider more than just one variable for most decisions and never make actual expected value calculations. Instead, I rely on my intuition to roughly determine which option leads to the highest win rate. Oftentimes, good players will have a reasonable idea of what the optimal play is just from their intuition. However, to become a great player it is extremely important that the idea of playing to win is at the core of each decision. Intuition is developed from past experiences and habits, and those on their own do not improve over time without proper critical assessment. For example, I may prefer to make a certain play because I am comfortable with it and know from experience that it works most of the time. However, that play may not lead to the highest win rate, nor has it ever, but because my intuition was developed without a framework, I completely overlook the actual optimal play. In this case, I have developed a sub-optimal intuition for making decisions because I did not properly evaluate how each decision would impact my chances of winning from the outset.

Life is oftentimes the same way, albeit more complex. Humans tend to flock to the familiar road and not necessarily to the best path. We prefer comfort, and change can be extremely difficult. But to improve at a game we are familiar with requires constant willingness to evaluate and critique our own decisions. And the same can be said about life.