25.02.2021

Rethinking our emotional connection with games

People love games. Currently, we have an approxiamately 170 billion US dollar industry, which is forecasted to double in size within the next ten years. I have played pretty much all my life and was hooked since my parents allowed me to watch them play The Legend of Zelda.

Now working in Games Research, I often think about the appeal of games. People do love them, but... why exactly? I'm sure everyone would give a slightly different answer to that question, although research provides us with great frameworks about the motivational pull and the rewards of games that feed into our psychological desires (e.g. Ryan, Rigby & Przybylski, 2006). In the end, we play to feel - maybe accomplishment, maybe calmness; but we play to feel and we feel something when we play.

mysterious person holding game controller

Of course, that is not exactly news to anybody in the industry or to researchers. But still, there is very little consequence following this simple train of thought. In fact, game design frameworks often don't even have many assumptions about the actual players, but rather guide the process from a personal perspective of the designer (Sotamaa, 2007). There are efforts to change this to a more user-centered design process that regards players' emotions as a key principle to design around. Often, this is called Affective Design or Emotional Design and it tries to accomplish two things: (a) understand emotions of players and (b) utilize user-centered desings to evoke target emotions. The aim is to create games with a greater emotional impact, games based on the emotional needs of players, simply put: better games.

This is great - a promising perspective to make more exciting games. But... how do we do this? This sounds like a straightforward process, we're all emotional beings, we understand our feelings and can just design our games around them, right? Well, unfortunately, it's not that easy. To illustrate some of the problems, let's consider the relation between game and player:

model of emotional feedback loop A simple model on an emotional feedback loop, based on the affective loop by Sundström (2005).

This simple loop illustrates the emotional interaction between game system and player. Ideally, the game system has a constant model of player emotions, fed through affective input from the player. The game then adapts its content in a very user-centered way to evoke target experiences, which we can call affective output. Of course, emotions are not static - and so the loop goes on. Now that we have a model of the process, let's look at some issues by trying to design a horror game:

Let's make our game dark and moody, let's add some unsettling sound frequencies. We put the player in a deserted hallway, nowhere to go but straight into the darkness. With every step we build the tension - and at the peak of the players' fear, we turn the sound off and the light out. That should be scary enough, right? Or so I hope, as I just imagine myself being scared. There is no way to check if my player actually is in fear, so it should be fine. Of course, there might be different reactions from different players, but you can't make everybody happy.

Maybe we can let our players wear a biofeedback device. We have access to players' heart rates, that changes everything, right? We can just measure it and feed it to our system - high heart rate means fear and low heart rate means calmness. Now we can make our game emotion-adaptive and amplify the horror when people aren't scared enough. But, how do we know what's enough? How do we know what is fun? Some players are confused about why there is no content in their game, others are overwhelmed. It's hard to say why, we tried our best to be adaptive.

Now, let's ignore that this is all a bit over the top and let's look at what's wrong and what we can do.

Affective Output:
First, when designing the game as "scary", we need an understanding of player emotions, or at the very least some assumptions about our players. We need data-backed and user-centered principles that can target specific aspects of (problematic) interactions between player and game, so we have to connect these principles with our affective feedback loop.

Affective Input:
Then, we need to understand emotions. Physiological reactions measured via biofeedback devices are one way, but right now they offer only very specific solutions, like building a game that uses facial expressions to control gameplay. What is with other components of emotion - like the whole psychological component? What is with games that don't want to use biofeedback systems? There is a lack of solutions to connect an emotional design vision to a working system that can adapt in a meaningful way to emotions - and therefore to players.

There is a lot to do and a lot to gain. When these problems are solved, we will be able to have games that are holistically made for the emotions of a player, a system that changes with these emotions. We will have a system that is not just for a certain ideal player - but for specifically you and your feelings. Games would be more immersive, more accessible, more rewarding. While this might still take some time, my research will hopefully bring us closer in this direction.

Written by Maximilian Croissant
25.02.2021