An upcoming podcast by Alexey “Flazm” Davydov includes a discussion in which I touched on cognitive biases, but the subject is so deep that I want to discuss it separately.
To begin with, a simple example from childhood. When we hit someone, it is a fair and justified act; when someone hits us, it is a cosmic injustice and the offender is a complete jerk. The event seems to be the same — someone hit someone — but we perceive it differently. Over time, this simple example develops into dangerous phenomena such as belief in a just world and victim blaming. People develop their own view of the world, which may be far removed from reality, and explain any discrepancies in ways that support that view, sometimes to the point of absurdity.
As unfortunate as it may sound, cognitive biases cannot really be called a psychological disorder. They are more of a protective mechanism. Less doubt. Fewer resources spent on perceiving our surroundings. Fewer negative feelings about ourselves. Faster decisions. A useful thing, after all. Evolution did not create it for nothing. Perhaps without it we would be dreary robots or a single organism, perceiving reality in the same way and feeling no need to prove someone wrong on the internet.
How can we put this knowledge into practice? The clearest example is shaping a customer’s perception of a price to suit the seller. There is an excellent discussion of this on Oleg Yakubenkov’s blog; make sure to read it. This knowledge applies to traditional retail and can equally be adapted to a game’s monetization model.
Of particular interest for game design is choice-supportive bias. One example is when someone buys a product and starts praising it to everyone while overlooking its obvious flaws. Similarly, if a negative event in a game happens because of the player’s decision rather than automatically, the player will explain to themselves that the action cannot be so bad, since they chose it. For instance, after a battle, players choose whether to repair their tank (or enable automatic repairs). Of course they will repair it, because refusing to do so would harm their subsequent play. But because they made that choice, they will tell themselves that the repair cost is small, that everyone does it, and that paying for repairs is nothing to worry about. Naturally, do not push this too far, or the player will reject the choice altogether and leave the game.
Or take another bias — zero-risk bias. A key costing 5 gold opens a chest with a guaranteed reward, while a key costing 3 gold opens a chest that might be empty. Which key would you prefer?
Loss aversion — players will feel a loss much more strongly than they enjoy new rewards. It is therefore better to combine negative moments into a single event and, conversely, spread positive moments across several events so they do not escape the player’s attention and create a positive background. For example, when the player’s character dies, the loss of progress is compounded by repair costs and the time needed to travel from the last checkpoint. Meanwhile, the income allocated to the level by the game’s balance is distributed a little at a time among all the barrels, so a coin drops at every turn.
The classic obedience to authority — players will pay more attention to tutorial tips if they come from a character who has authority within the game’s universe. Of course, that authority must first be communicated to the player.
The Hawthorne effect can work both ways. The results of focus testing with players invited to the studio will differ dramatically from statistics collected indirectly. An invited player who receives everyone’s attention will start to feel involved in the creative process and stop behaving like an ordinary player. At the same time, this effect is used in community work. Players become more loyal when they are given a chance to contribute to the project’s development.
This is only the tip of the iceberg. A full list of these effects can be found here (bookmark it). I hope these examples help you understand how to take other cognitive biases into account in game design.
To begin with, a simple example from childhood. When we hit someone, it is a fair and justified act; when someone hits us, it is a cosmic injustice and the offender is a complete jerk. The event seems to be the same — someone hit someone — but we perceive it differently. Over time, this simple example develops into dangerous phenomena such as belief in a just world and victim blaming. People develop their own view of the world, which may be far removed from reality, and explain any discrepancies in ways that support that view, sometimes to the point of absurdity.
As unfortunate as it may sound, cognitive biases cannot really be called a psychological disorder. They are more of a protective mechanism. Less doubt. Fewer resources spent on perceiving our surroundings. Fewer negative feelings about ourselves. Faster decisions. A useful thing, after all. Evolution did not create it for nothing. Perhaps without it we would be dreary robots or a single organism, perceiving reality in the same way and feeling no need to prove someone wrong on the internet.
How can we put this knowledge into practice? The clearest example is shaping a customer’s perception of a price to suit the seller. There is an excellent discussion of this on Oleg Yakubenkov’s blog; make sure to read it. This knowledge applies to traditional retail and can equally be adapted to a game’s monetization model.
Of particular interest for game design is choice-supportive bias. One example is when someone buys a product and starts praising it to everyone while overlooking its obvious flaws. Similarly, if a negative event in a game happens because of the player’s decision rather than automatically, the player will explain to themselves that the action cannot be so bad, since they chose it. For instance, after a battle, players choose whether to repair their tank (or enable automatic repairs). Of course they will repair it, because refusing to do so would harm their subsequent play. But because they made that choice, they will tell themselves that the repair cost is small, that everyone does it, and that paying for repairs is nothing to worry about. Naturally, do not push this too far, or the player will reject the choice altogether and leave the game.
Or take another bias — zero-risk bias. A key costing 5 gold opens a chest with a guaranteed reward, while a key costing 3 gold opens a chest that might be empty. Which key would you prefer?
Loss aversion — players will feel a loss much more strongly than they enjoy new rewards. It is therefore better to combine negative moments into a single event and, conversely, spread positive moments across several events so they do not escape the player’s attention and create a positive background. For example, when the player’s character dies, the loss of progress is compounded by repair costs and the time needed to travel from the last checkpoint. Meanwhile, the income allocated to the level by the game’s balance is distributed a little at a time among all the barrels, so a coin drops at every turn.
The classic obedience to authority — players will pay more attention to tutorial tips if they come from a character who has authority within the game’s universe. Of course, that authority must first be communicated to the player.
The Hawthorne effect can work both ways. The results of focus testing with players invited to the studio will differ dramatically from statistics collected indirectly. An invited player who receives everyone’s attention will start to feel involved in the creative process and stop behaving like an ordinary player. At the same time, this effect is used in community work. Players become more loyal when they are given a chance to contribute to the project’s development.
This is only the tip of the iceberg. A full list of these effects can be found here (bookmark it). I hope these examples help you understand how to take other cognitive biases into account in game design.