Are you feeling good and in the zone? Or are you hot and bothered? Irritable and frustrated? Or maybe sad and melancholic? While all kinds of games exist for many varieties of mood, it can be a good idea for a video game to adjust its difficulty based on how the player is feeling. Because continually feeling mad at a game might not be so fun or so good for you.
Scientists from South Korea, at the Gwangju Institute of Science and Technology, have come up with a rather intriguing method for such a thing. The researchers developed a dynamic difficulty model that would adjust based on player emotions and change accordingly to ensure maximum player satisfaction. Because who doesn’t love maximum satisfaction?
Game developers have long known the required balance of game difficulty and player progression, trying to find a sweet spot that is neither too difficult nor too easy to ensure the gaming experience is pleasant. Although the settings can usually be changed, this often requires the player to manually adjust the setting. Korean scientists are coming up with something much more dynamic.
Their model involves training Dynamic Difficulty Adjustment Agents (DDAs), using machine learning that gathered data from human players, who then adjust the game’s difficulty to maximize the one of four different aspects related to a player’s satisfaction: challenge, skill, flow, and valence.
The scientists used a fighting game for their model and to train their DDA agents, as human players played the fighting game against AI opponents, generating data for the agents, and the humans also had to respond to a questionnaire about their experience. Using an algorithm called Monte Carlo Tree Search, each DDA agent uses real game data and simulated data to tune and adjust the opposing AI’s fighting style in ways that maximize an emotion or a specific “affective state”.
Associate Professor Kyung-Joong Kim, who led the study, said one of the advantages of their approach was that the player did not need to be monitored with external sensors to detect their emotions. “Once trained, our model can estimate player states using only game features,” he said.
The study was small, using just 20 volunteers, but the team said DDA agents produced AIs that improved the overall player experience. However, fighting games offer the most direct feedback, which begs the question of how it could be used for other types of games, but the professor had an answer for that.
“Commercial game companies already have massive amounts of player data. They can leverage this data to model players and solve various game balancing issues using our approach,” Prof Kim said.
Their paper documenting the model, “Diversifying Dynamic Difficult Adjustment Agent by Integration Player State Models into Monte-Carlo Tree Search”, will be published in the journal Expert Systems With Applications on November 1. But for those interested, it is already available online and can be found here.
Next: Tekken Bloodline Needed To Be More Forgiving To Itself