New gaming model changes level of difficulty based on emotions expressed by players

The researchers behind the model believe it could help balance the difficulty of games and make them more appealing to all types of players.

Difficulty is a key aspect of any videogame and can make or break the player’s experience. Finding the right balance between challenging and fun can be challenging for game developers. To address this issue, a team of scientists from Gwangju Institute of Science and Technology in Korea has developed a novel approach to dynamic difficulty adjustment (DDA) that considers the players’ emotions. This new model estimates players’ emotions using in-game data and adjusts the difficulty level accordingly to maximize player satisfaction. The scientists believe their efforts could help balance the difficulty of games and make them more appealing to all types of players.

Dynamic Difficulty Adjustment (DDA)

Most game developers use DDA to adjust the difficulty level of a game in real-time based on player performance. If the player’s performance exceeds the expected level for a given difficulty, the game’s DDA agent can automatically increase the difficulty to provide a more challenging experience. However, this approach has limitations since it only considers player performance and not their emotional state.

The new approach

The researchers took a different approach and developed DDA agents that adjust the difficulty level to maximize one of four different aspects related to a player’s satisfaction: challenge, competence, flow, and valence. The DDA agents were trained using machine learning on data collected from human players who played a fighting game against different artificial intelligence (AI) opponents and answered a questionnaire about their experience.

Using Monte-Carlo tree search, each DDA agent used actual game data and simulated data to adjust the opposing AI’s fighting style to maximize a specific emotion or ‘affective state.’ The scientists noted that this approach does not rely on external sensors, such as electroencephalography, which is an advantage over other emotion-centered methods.

The results

The scientists tested their approach using an experiment involving 20 volunteers. They found that the proposed DDA agents produced AIs that improved the players’ overall experience, regardless of their preference. This is the first time affective states have been incorporated directly into DDA agents, and the scientists believe this approach could be useful for commercial games. They noted that game companies already have huge amounts of player data that they could use to model players and address game-balancing issues using their approach.

Potential applications

The scientists believe that their approach could have potential applications beyond gaming, such as healthcare, exercise, and education. These fields could benefit from gamification techniques that could improve patient engagement and motivation. For example, healthcare providers could use gamification to encourage patients to follow a healthy diet and exercise regimen. Education providers could use gamification to engage students and make learning more fun and enjoyable.


The Korean scientists’ approach to dynamic difficulty adjustment that considers the players’ emotions is a significant step forward in video game and casino game development. Balancing a game’s difficulty is crucial to providing a pleasant player experience. By taking into account the players’ emotions, game developers can create games that appeal to a broader audience and provide a more satisfying experience. The scientists’ approach could also have applications beyond gaming and could improve patient engagement and motivation in healthcare, exercise, and education. Overall, the approach could help make the world a more enjoyable and engaging place.

Zoran Markovic

Zoran Markovic is a reporter at Breakthrough.

Latest from Blog