Faster is Beetr : The Speed of Player Character Growth aefcts Enjoyment and Perceived Competence

Abstract

The Self-Determination Theory (SDT) predicts that people tend to be proactive and engaged in activities that can satisfy three specific intrinsic needs: 1) the need for competence: the innate desire to grow our abilities, 2) the need for autonomy: the innate desire to be the causal agent of our own life, 3) the need for relatedness, or for meaningful interactions with others. The SDT turned out to be a highly successful tool for investigating video games under several perspectives. In more than 15 years of research, it has been constantly found that the most successful, engaging, and fun video games are the ones that satisfy the intrinsic needs of competence, autonomy, and relatedness. Here, we put to the test the SDT predictions by manipulating the amount of reward given to the player within a commercial video game, Torchlight II - in opposition to applied games and gamified application used in previous studies. Game metrics, video recordings, and self-reported feedback have been collected and analyzed from two groups of video game players that participated in a 60-minutes play session: a) the control group played the standard version of the game, while b) the experimental group played a version of the game that provided five times the amount of rewards. Results showed that the speed of player character growth affected the participants' perceived competence and their enjoyment of the game, although the game metrics indicated that the two gameplay sessions were objectively almost identical. Considerations for games researchers and game designers are presented.

Keywords

• eol>Self-determination theory
• Game Design
• Rewards
• Video Games

CCS CONCEPTS

• Human-centered computing → User studies; Empirical studies in HCI; • Applied computing → Computer games; †Corresponding author

INTRODUCTION

Video games have been catching the atention of academics since their infancy as an industry during the eighties [ 12 ]. One of the most researched topics is the motivational power of this media and why video games are so successful at engaging their users [ 1, 2, 17, 26 ] – as also demonstrated by the emergence of the “gamicfiation ” phenomenon [ 5 ], see [ 11 ] for a recent review. The Self determination Theory, a macro theory of human motivation, proposes that people behaviors are determined by three specicfi intrinsic needs: 1) the need for competence or the innate desire to take part in activities which allows us to feel capable and eefctive , 2) the need for autonomy or the need to experience freedom in the activities we choose, 3) the need for relatedness or the need to feel a sense of meaningful connection to others [ 3, 4, 20, 21 ]. In more than 15 years of research, a number of researchers applied SDT specicfially on video games and constantly found that the most successful, engaging, and fun video games are the ones that satisfy the intrinsic needs of competence, autonomy, and relatedness [ 17, 18, 22 ]. However, those studies focused on testing SDT predictions using diefrent commercial video games thought to difer a priori in needs satisfaction [ 9, 22 ], or some specicfially created ones [ 10, 15 ]. In the present study, SDT has been put to test by manipulating one single feature of a commercial video game – thus maintaining a high level of control on the experimental conditions, and at the same time retaining a high degree of ecological validity. Our hypothesis was that by improving the amount of reward given to the player, the need for competence would be satisfied to a greater extent, with a positive influence on enjoyment.

RELATED WORK

Ryan, Rigby, and Przybylski [ 22 ] tested the validity of SDT by employing several commercial video games. For example, in study number 2 (ibid. p. 353), they pre-selected two titles according to their position in a game ranking survey, and they anticipated that participants would rate the amount of their perceived competence and autonomy satisfaction diefrently . The h igh-rated title would be beter at satisfying the psychological needs identiefid by the SDT , while the low-rated game would be lacking at doing so. In a following work, Rigby and Ryan [ 18 ] went beyond the general diefrences between games (i.e., good or bad) and they described in greater detail the role of game design elements in afecting the satisfaction of psychological needs. Th ey reviewed a great variety of video games genres: music games, shooters, sports/driving games, platformers, role-playing games, etc.

A similar approach (but this time empirical) has been taken by Johnson et al. [ 10 ] and Phillips et al. [ 16 ] in the study of a specicfi game mechanic: the impact of virtual rewards on the player experience. They created a videogame with three levels of rewards and found that eofrt, enjoyment , and presence signicfiantly increased when all rewards were present in comparison to the conditions in which only some of them were presented to the players.

Peng et al. [ 15 ] created an exergame where specicfi SDT needs supporting features were manipulated: for example, the need for competence was satisefid with the support of 1) a dynamic dificulty adjustment mechanism , b) a “heroism meter”, i.e., a cumulative score, c) achievements badges for the player. In the context of gamicfiation studies, similar empirical researches have been performed by Sailer et al. [ 23 ], and Meckler et al. [ 14 ]: specicfi game design features have been manipulated in two repetitive and not particularly stimulating nor motivating tasks: a “handling of materials and supplies” simulation and an image annotation (i.e., tagging) task.

In summary, SDT predictions have been empirically tested using either a) commercial video games in a general context [ 19 ], or b) non-commercial video games [e.g., 10, 16, 17] or gamiefid activities in a controlled context [e.g., 14, 24]. Here, we wanted to close the gap by employing a commercial video game with the manipulation of a specicfi game design feature, i.e., the amount of reward given to the player.

EXPERIMENT

According to SDT, a video game player will appreciate to a greater extent the video game that is beter at satisfying their needs for competence, autonomy and relatedness. We hypothesized that the same video game can be preferred over itself if the amount of reward for any accomplishment within the game is enhanced.

A comparison was performed between two versions of a commercial videogame: the “standard” one, and a second version with an enhanced XP reward curve (cf. later) Please note that everything else has been kept constant across the two experimental groups we compared. The “enhanced reward” version of the game should obtain more success because of its ability to satisfy to a greater extent the need for competence, thus eliciting a higher level of enjoyment [ 18, 22 ].

Among the various game design features, the reward curve has been chosen because it is considered one of the most impactful and ubiquitous features in game design and gamicfiation [ 11, 22 ].

THE VIDEO GAME

hTe game employed in this study was Torchlight II [ 19 ], an action-role playing video game. A role-playing game (or RPG) is a game genre where the player controls the actions of one character (or more characters within a group, called party) that embarks on a series of dificulty -increasing quests in a narratively well-denfied world. This player character has quantiafible features such as a level number and several artibutes expressed by a score. Character progression is used as an essential measurement of achievement: successful advancements in the game translate into experience points (XP points) through which the player character can “level up” to new powers and skills [ 26 ]. In Torchlight II, every new level provides the player with atribute points that they can use to improve their player character in: Strength (increases weapon damage), Dexterity (increases the chance to dodge enemy atacks and perform critical strikes), Focus (increases magic damage), and Vitality (increases health). At the core of the game, there is a loop between playing, collecting, and improving as in gfiure 1.

hTe additional term “action” in “action -RPG video game” speciefis that the gameplay is focused more on dynamic aspects (such as combat and exploration), and less on logic (such as puzzles).

In summary, genre distinguishing features of action–RPG games are: 1) a strong emphasis on character development, 2) a considerable freedom for the player to choose their own way to play the game – thanks to a great number of game contents (i.e. missions, enemies, equipment, etc.), 3) a higher accessibility in comparison to traditional RPGs. hTese three features justified the choice of using Torchlight II in the experiment: a game with a gentle learning curve that could guarantee adequate support of the needs of competence, autonomy, and relatedness.

THE VIDEO GAME: MANIPULATION hTe core of the game implies gathering experience points in order to level up, and thus improve the player character’s atributes. In turn, these improvements allow the player to undertake even more dificult quests, unlock game areas with more dangerous enemies, and dedicate their atention to more rewarding activities – in turn, this speeds up the improvement process.

Obtaining XP is a central part of the gameplay of Torchlight II. Each kill of an enemy will give the player some of it, while nfiished quests give even more. The amount of XP points that each activity releases to the player (the XP reward curve) is carefully determined by game designers who set their vision of what is the “right” way of playing. Also, the relationship between the number of XP points and the thresholds for leveling up (the XP thresholds curve) has to be carefully set by game designers according to their vision. The common strategy is to build a progressive curve : players quickly level up during the early stages of the game (i.e., the XP points needed to level up are in the tens) but they have to collect more points in order to reach higher levels (i.e., the XP points needed to level up are in the tens of thousands and more).

However, more than a game design strategy, this is the direct application of the Weber-Fechner law, which describes the relationship between the actual change in a physical stimulus and the perceived change, established by the founders of psychophysics in 1860 [ 6 ]. In the context of role-playing games, players need a logarithmic increase of XP to feel the same progress pace when the leveling up curve is represented by a linear relationship [ 7 ]. In Torchlight II, designers used, to a similar result, a linear function for the XP reward curve and a quadratic function to represent the XP thresholds curve.

Here, a new experimental condition has been introduced by multiplying by a factor of 5 the standard XP reward curve, which serves as a control condition. The manipulation has been made using the embedded game editor which allowed to change the game under several aspects. In gfiure 2, the standard condition is represented with a doted black line (1x XP reward curve); the experimental condition is represented with a black line (5x XP reward curve); the XP threshold curve (doted red line) remained the same.

MEASURES

Two types of feedback have been collected during the experiment: 1) objective and 2) subjective feedback. Objective measures concerned all the events that happened during the gameplay session, with a focus on participants’ actions in the game. The objective measures were collected in three moments during the session: 1) after 20, 2) 40, and 3) 60 minutes of play. To simplify the analysis of the data, they have been grouped into three sets: 1) Interactions with the Player Character, 2) Interactions with the Environment, 3) Interaction with the Enemies. The rfist group consisted of the actions related to the player character improvements; the second group included the result of the exploration and interactions with the environment; the third group summarized the actions taken during combat.

Subjective measures concerned participants’ thoughts and atitudes towards the video game. Five scales of the Intrinsic Motivation Inventory (IMI) [ 13, 24 ] have been employed to assess participants’ 1) enjoyment, 2) perceived competence, 3) eofrt, 4) felt pressure and tension, 5) perceived choice. 30 items on a 7-point Likert-type scale have been translated to Italian and proposed to participants in random order at the end of the session. In addition, a survey with four questions was submited during the session in three intervals: after 20, 40 and 60 minutes of play. These questions assessed the perceived 1) speed of leveling up, 2) progression in the game, 3) dificulty of the gameplay, and 4) atachment to the game .

PARTICIPANTS

A prolfiing questionnaire was employed among university students to identify gamers with the following characteristics: 1) be familiar with the PC gaming control layout (i.e., keyboard and mouse), 2) be familiar with the action-RPG video game genre, 3) have not played the commercial video game Torchlight II. From a total of 23 respondents (22 males, 1 female), 14 met the above requirements and were subsequently contacted and invited to participate in the study. All accepted and took part in the experiment. Participants were all males, aged between 21 to 37 years old (mean age = 26.35; SD = 3.9).

PROCEDURE

Upon arrival, participants were informed about their gameplay sessions and spontaneous voice comments being recorded. Aeftr providing consent, they were instructed on their task: they had to play Torchlight II, as they were at home, for a total of 60 minutes, and to compile two questionnaires at specicfi times during the session. The player character type was kept constant: all used the Embermage class with identical statistics at the start – but they were free to assign points to the four atributes at will during the session. Every 20 minutes, they were asked to compile a brief survey while the experimenter was retrieving the objective data relative to the previous 20 minutes of gameplay. Aeftr 60 minutes, they were asked to compile the IMI questionnaire. During the test, the experimenter sat behind the participant for giving support if requested, and for observing the PC monitor and taking notes about relevant in-game behaviors. At the end of the session, participants received a debriefing detailing the goal of the research. All participants played on an ASUS ZenBook flip s UX370UA notebook with a 13.3 inches monitor; a Tecknet cordless optical mouse and keyboard have been employed for controlling the pointer instead of the standard touchpad in order to faithfully reproduce the standard PC gamers set-up with this type of games.

ANALYSIS

Objective and subjective data collected within the sessions were analyzed using a mixed Analysis of Variance (ANOVA) with two independent variables: 1) betweensubjects “XP reward condition” and 2) within-subjects “Time”. The subjective data collected with the IMI scale at the end of session was analyzed with a univariate ANOVA.

RESULTS

Interactions with the Player Character. As expected, participants assigned to the experimental condition saw more signicfiant growth in the level of their player character in comparison to the control condition (F(1,12)=45.194 p<.001). Post-hoc comparisons showed that the most substantial diference happened within the rfist 20 minutes of play, although the diefrence met statistical signicfiance throughout the session – see gfiure 3.

Consequently, the number of points assigned to the four atributes is higher in the experimental condition. However, when considering the percentage of points assigned to the four atributes, we found that only Strength is close to signicfiance (p=.056): participants in the 5x XP condition invested more points on improving their atack power - see gfiure 4.

XP=4.2; 5x XP=5.3; F(1,12)=5.368 p<.05). The graph in gfiure 5 depicts the scores in the five scales.

Interactions with the environment. No signicfiant diferences were found for the following game metrics: collected gold, steps taken, completed missions, traps activated, portals used, collected loot. The only signicfiant result concern the number of destroyed objects: participants in the control conditions destroyed more objects (1x XP: 60; 5x XP: 44; F(1, 12)=10.265 p<.01).

Interactions with the enemies. Again, no signicfiant diefrences were found between the 1x and 5x XP conditions for the number of enemies killed (normal and bosses), health potions used, player character deaths, maximum damage taken The only signicfiant result concern the maximum damage inflicted with one hit: 5x XP participants inflicted more damage (1x XP: 99; 5x XP: 135; F(1, 12)=11.24 p<.01).

Subjective experience. Participants assigned to the experimental group rated the speed of leveling up faster than the control condition (1x XP: 4.1; 5x XP: 5; F(1,12)=5.562 p<.05). The signicfiant interaction (F( 1.59, 19)=4.709 p<.05) and post hoc comparisons showed that the ratings diefred only in the rfist 20 minutes. On the other hand, t here is no diefrence between how participants rated progression ( 1x XP=4.9; 5x XP=4.7; p=0.708), dificulty (1x XP =2.8; 5x XP=3.1; p=.46), and atachmen t to the game (1x XP=3,3; 5x XP=4.8; p=.09).

IMI scales. Reliability of the vfie scales has been tested using Chronbach’s Alpha: the range of values were between .946 for the Interest/Enjoyment scale and .604 for the Eofrt/Importance scale. Signicfiant di efrences have been found in two scales: Interest/Enjoyment (1x XP=3.5; 5x XP=5; F(1,12)=4.759 p=.05) and Perceived Competence (1x

DISCUSSION

hTe purpose of this research was to test whether increasing the XP points given to the players for any in-game accomplishments in an action-RPG game could lead to a higher level of competence satisfaction and consequently to a higher enjoyment of the game, as predicted by the Self Determination Theory .

Two groups of video game players were asked to play two diefrent versions of the ga me Torchlight II: the commercially available version, and a modified version that granted players with five times more XP points. Objective data were compared between the two groups in order to control for any diefrence in the gameplay (i.e., what they did). Subjective data ware compared to search for diefrences in their experience of the game (i.e., what they think), and to test our hypothesis.

Our results indicate that perceived competence ratings were indeed higher for the participants assigned to the enhanced XP reward condition. Also, the Interest Enjoyment subscale received higher ratings in this condition. It is interesting to note that participants were aware of this boost in XP because they rated the speed of leveling up as faster in comparison to the control condition. However, they did not experience the gameplay diefrently from the control condition: both groups rated progression as rather adequate, dificulty as low, and atachment to the game as average. The few objective diferences foun d concern expected and derivative results: the number of levels gained, atribute points assigned, maximum damage inflicted .

Taking these results together, it is reasonable to conclude that the XP reward manipulation was able to elicit a higher degree of perceived competence and enjoyment without altering the original designers’ vision since no substantial diefrences have been found in the way participants played the game. Designers may take advantage of this result by balancing their titles accordingly. hTis research is our small contribution to bridging the gap between the academic and business cultures [ 8 ].

ACKNOWLEDGMENTS

hTe authors would like to thank the anonymous referees for their valuable comments and helpful suggestions.

References

[1] Bostan , B. ( 2009 ). Player Motivations: A Psychological Perspective . ACM Computers in Entertainment, Vol. 7 , No. 2, Article 22.

[2] Boyle , E. A. , Connolly , T. M. , Hainey , T. , & Boyle , M. B. ( 2012 ). Engagement in digital entertainment games: a systematic review . Computers in Human Behavior 28 , 771 - 780 .

[3] Deci , E. L. , & Ryan , R. M. ( 1985 ). Intrinsic motivation and selfdetermination in human behavior . New York: Plenum.

[4] Deci , E. L. , & Ryan , R. M. ( 2010 ). Self-determination . In Corsini encyclopedia of Psychology.

[5] Deterding , S. , O 'Hara , K. , Sicart , M. , Dixon , D. , & Nacke , L. ( 2011 ). Gamification: Using Game design elements in Non-gaming Contexts . Proceedings of ACM CHI 2011 , May 7- 12 , 2011 , Vancouver, BC, Canada.

[6] Fechner , G. T. ( 1966 ). Elements of psychophysics . New York: Holt, Rinehart & Winston. Original work published in 1860 .

[7] Hodent , C. ( 2018 ). The Gamer's Brain. How neuroscience and UX can impact video game design . CRC Press, Taylor & Francis Group, Boca Raton, FL, U.S.

[8] Hopson , J. ( 2006 ). We're not listening: An open letter to academic game researchers . Gamasutra Feature . Retrieved here: https://www.gamasutra.com/view/feature/130200/were_not_ listeni ng_an_open_ .php

[9] Inchamnan , W. , Wyeth , P. ( 2013 ). Motivation during videogame play: analyzing player experience in terms of cognitive action . Proceedings of ACM IE'2013, September 30 - October 1 2013 , Melbourne, VIC , Australia.

[10] Johnson , D. , Klarkowski , M. , Vella , K. , Phillips , C. , McEwan , M. , & Watling , C. N. ( 2018 ). Greater rewards in videogames lead to more presence, enjoyment and effort . Computers in Human Behavior 87 , 66 - 74 .

[11] Koivisto , J. , Hamari , J. ( 2019 ). The rise of motivational information system: a review of gamification research . International Journal of Information Management , 45 , 191 - 210 .

[12] Loftus , G. R. , Loftus , E. F. ( 1983 ). Mind at Play. The Psychology of Video Games. Basic Books , New York.

[13] McAuley , B. , Duncan , T. , & Tammen , V. V. ( 1988 ). Psychometric properties of the Intrinsic motivation inventory in a competitive sport setting: A confirmatory factor analysis . Research Quarterly for Exercise and Sport , 60 , 48 - 58 .

[14] Mekler , E. D. , Bruehlmann , F. , Touch , A. N. , & Opwis , K. ( 2017 ). Towards understanding the effects of individual gamification elements on intrinsic motivation and performance . Computers in Human Behavior 71 , 525 - 534 .

[15] Peng , W. , Lin , J.H. , Pfeiffer , K.A. , & Winn , B. ( 2012 ). Need satisfaction supportive game features as motivational determinants: an experimental study of a self-determination theory guided exergame . Media Psychology , 15 : 175 - 196 .

[16] Phillips , C. , Johnson , D. , Klarkowski , M. , White , M. , & Hides , L. ( 2018 ). The impact of rewards and trait reward responsiveness on player motivation . In: Symposium on Computer Human Interaction in Play, Melbourne, VIC , Australia, October 28 - 31 , 2018 : CHI PLAY '18 Proceedings of the 2018 Annual Symposium on ComputerHuman Interaction in Play . New York, NY: ACM Press, ss . 393 - 404 . doi: 10 .1145/3242671.3242713

[17] Przybylski , A. K. , Rigby , C. S. , & Ryan , R. M. ( 2010 ). A motivational model of video game engagement . Review of General Psychology , Vol. 14 , No. 2 , 154 - 166 .

[18] Rigby , S. & Ryan , R. M. ( 2011 ). Glued to Games: How video games draw us in and hold us spellbound . Santa Barbara, U.S.: Praeger.

[19] Runic Games ( 2012 ). Torchlight II. Retrieved from https://www .torchlight2.com

[20] Ryan , R. M , Deci , E. L. ( 2000a ). Self-determination theory and the facilitation of intrinsic motivation, social development, and wellbeing . American Psychologist 55 , 68 - 78 .

[21] Ryan , R. M , Deci , E. L. ( 2000b ). Intrinsic and extrinsic motivations: Classic definitions and new directions . Contemporary Educational Psychology 25 , 54 - 57 .

[22] Ryan , R. M. , Rigby , S. C , & Przybylski , A. ( 2006 ). The motivational pull of video games: A self-determination theory approach . Motivation and Emotion 30 , 4 , 347 - 364 .

[23] Sailer , M. , Hence , J. U. , Mayr , S. K. , & Mandl , H. ( 2017 ). How gamification motivates: An experimental study of the effects of specific game design elements on psychological need satisfaction . Computers in Human Behaviors 69 , 371 - 380 .

[24] Tsigilis , N. , Theodosiou , A. ( 2003 ). Temporal stability of the intrinsic motivation inventory . Perceptual and Motor Skills , 97 , 271 - 280 .

[25] Wiebe , E. N. , Lamb , A. , Hardy , M. , & Sharek , D. ( 2014 ). Measuring engagement in video game-based environments: Investigation of the User Engagement Scale . Computers in Human Behavior , 32 , 123 - 132 .

[26] Wolf , M. J. P. ( 2012 ). Encyclopedia of Video Games: The culture, technology and art of gaming . Greenwood.