In the future, video games might measure your emotions to deliver the perfect experience

In the 1989 video game adventure film The Wizard, Jimmy Woods — played by Luke Edwards — finds himself competing in a video game competition after discovering his innate gaming ability. After attaining the high score in Super Mario Bros. 3, Woods leaves the competition with a $50,000 grand prize, all because his ability was underestimated.

When making any new game, designers are tasked with balancing the difficulty to serve a wide range of players. If you make a game too easy or too difficult, you risk alienating a large subset of the community. Often, the question of difficulty is solved by a selection menu at the start of the game. Each individual player can choose a gaming experience which works best for them. Even so, players may struggle with selecting the right gaming experience, one which will land them in the sweet spot where a game is challenging enough to keep them coming back and not so challenging that it puts them off.

Future generations of gamers, however, might look back at the difficulty selection screen the same way modern gamers remember setting the television to channel three: with nostalgic reverence. It seems that manual difficulty selection may have one foot in the digital grave. Scientists from the Gwangju Institute of Science and Technology have developed a new technology which utilizes dynamic difficulty adjustment to tweak the gameplay experience in real-time using data gathered from the player. Their results were published in the journal Expert Systems with Applications.

Today, dynamic difficulty adjustment is employed in some games to crank the difficulty up or down during gameplay. However, it’s less dynamic than it could be because it only takes one factor into consideration. Modern DDA pays attention to a player’s skill level as they navigate through a game. It makes some base assumptions about how skilled an average player is, then bumps it up against the individual player’s performance. If it notices you’re cruising through challenges more easily than it anticipated, a game might crank up the difficulty on your next encounters.

That kind of dynamic adjustment is successful at landing someone at the sweet spot for difficulty, based on skill, but it doesn’t necessarily make for the most enjoyable gaming experience. Just because you can play at a certain level, doesn’t mean you always want to.

Instead of looking only at skill level, scientists gathered information about the player’s experience as it pertains to four game criteria: challenge, competence, valence, and flow. The team’s novel DDA algorithm could adjust gameplay to maximize any one of those criteria based on player’s preferences.

To build their data set, researchers recruited 20 volunteers and set them down with a controller to play a fighting game against an array of artificially intelligent opponents. After each round, volunteers completed a questionnaire about their experience and that information was used to tune an algorithm called Monte-Carlo tree search.

Over time, the algorithm learned how certain gaming experiences affected the players and how to tune the game in real-time with player data. The tuning itself worked in much the same way as standard dynamic difficulty adjustment, with the key difference that it was focused on maximizing player enjoyment or another identified emotional state.

Deploying this type of system into commercial games could result in a more perfect gaming experience for all players. Each of us could spend time in the same gaming universe while being delivered a bespoke experience tuned to our specific emotional states. The researchers also noted that this system has potential applications in other environments, particularly IRL ones. Notably, they could be used in education to level up curriculum or one-to-one support tailored to an individual student.

That would be good for gaming too. The faster you get your homework done, the more time you have for games.