The Design and Personalization of Aim Assistance Algorithms
Aiming is the act of directing an action toward a target object, location, or direction. Many activities in games require aiming. In order for players to perform these activities successfully, they must be able to aim accurately. When an action is too challenging, the player can become frustrated. Players that are not able to aim accurately enough to perform an action well can be helped through in game systems called aim assistance algorithms. However, it can be challenging to identify how to appropriately assist players. Additionally, it can be difficult to determine how much assistance is required for individual players. We have developed a novel taxonomy to classify aimed actions in games. The taxonomy classifies an aimed action based on its immediacy, target interaction, and embodiment. The classification can be used to inform the design of appropriate aim assistance algorithms. We used this taxonomy to classify aimed actions in three games. Then, based on this classification we designed aim assistance algorithms for each game. We ran two experiments to evaluate these aim assistance algorithms by simulating players of varying skill levels. Our first experiment established a linear relationship between the level of assistance given to a player and their performance of an aimed action. This linear relationship allowed us to personalize our aim assistance algorithms for different skill levels. The second experiment compared the effectiveness of personalized aim assistance algorithms to giving all players the same level of assistance. Personalizing the aim assistance algorithms resulted in players finishing rounds of games with closer scores. This thesis highlights the importance understanding aiming so that players with different abilities can receive personalized support. Additionally, it also shows that personalizing assistance algorithms can result in more balanced game difficulty.
Request an alternative formatIf you require this document in an alternate, accessible format, please contact the Queen's Adaptive Technology Centre
The following license files are associated with this item: