Adjustment of Difficulty Level on Wobble Board-Based Game Using Monte Carlo Tree Search Algorithm

There are many training methods applied using video game as a medium to improve user motivation in training. Besides its game design, the setting of difficulty level also affects user motivation. If a game is too difficult, its player will be stressful. And if it's too easy, its player will be bored quickly. A game must balance player's skill and challenge provided in it. Dynamic Difficulty Adjustment (DDA) is a technique used to adjust difficulty level in a game with its player's skill, using Artificial Intelligence (AI) or Algorithm. Monte Carlo Tree Search can be applied by using AI DDA agent to convert option policy and playout evaluation heuristically. It is applied to balance the difficulty level with player's skill. A test has been carried out by testing AI DDA agent's accuracy and comparing the effects of every difficulty level strategy in a balance training game with wooble board-based. Its result shows that AI DDA agent is able to adjust difficulty level with 82% accuracy. However, the strategy comparison of difficulty level has no significant difference, but one of the parameters, i.e. Health Point, shows that the game can adjust difficulty level with player's skill.

[1]  E. Wikstrom,et al.  Understanding and Treating Lateral Ankle Sprains and their Consequences , 2013, Sports Medicine.

[2]  Albert A. Rizzo,et al.  Interactive game-based rehabilitation using the Microsoft Kinect , 2012, 2012 IEEE Virtual Reality Workshops (VRW).

[3]  Stefania Serafin,et al.  Gameplay as a Source of Intrinsic Motivation for Individuals in Need of Ankle Training or Rehabilitation , 2012, PRESENCE: Teleoperators and Virtual Environments.

[4]  Guillaume Maurice Jean-Bernard Chaslot Chaslot,et al.  Monte-Carlo Tree Search , 2010 .

[5]  Benjamin Ultan Cowley,et al.  Adaptive Artificial Intelligence in Games , 2016 .

[6]  L Neumann,et al.  Wobble board training after partial sprains of the lateral ligaments of the ankle: a prospective randomized study. , 1996, The Journal of orthopaedic and sports physical therapy.

[7]  Xiaodong Li,et al.  Monte Carlo tree search based algorithms for dynamic difficulty adjustment , 2017, 2017 IEEE Conference on Computational Intelligence and Games (CIG).

[8]  Nazzun Hanif Ahsani.  Penerapan Algoritma Monte Carlo Tree Search Pada Permainan Komputer Maze Treasure , 2017 .

[9]  Matthew K. Miller,et al.  Effects of balancing for physical abilities on player performance, experience and self-esteem in exergames , 2014, CHI.

[10]  Glauco A. P. Caurin,et al.  Dynamic difficulty adjustment with Evolutionary Algorithm in games for rehabilitation robotics , 2016, 2016 IEEE International Conference on Serious Games and Applications for Health (SeGAH).

[11]  Peta Wyeth,et al.  A framework of Dynamic Difficulty Adjustment in competitive multiplayer video games , 2013, 2013 IEEE International Games Innovation Conference (IGIC).

[12]  P. Gribble,et al.  Functional rehabilitation interventions for chronic ankle instability: a systematic review. , 2010, Journal of sport rehabilitation.

[13]  M. Pierrynowski.  Twelve-week biomechanical ankle platform system training on postural stability and ankle proprioception in subjects with unilateral functional ankle instability , 2009 .

[14]  Simon M. Lucas,et al.  A Survey of Monte Carlo Tree Search Methods , 2012, IEEE Transactions on Computational Intelligence and AI in Games.

[15]  Cameron McGuinness.  Monte Carlo Tree Search: Analysis and Applications , 2016 .

[16]  Nadia Hocine,et al.  Adaptation in serious games for upper-limb rehabilitation: an approach to improve training outcomes , 2015, User Modeling and User-Adapted Interaction.

[17]  M Tamassia,et al.  Artificial intelligence techniques towards adaptive digital games , 2017 .

[18]  Xiao Liu,et al.  Dynamic Difficulty Adjustment of Game AI by MCTS for the game Pac-Man , 2010, 2010 Sixth International Conference on Natural Computation.