How to Step Up Your VR Game? An Analysis of Structural Properties Related to Psychological Needs as Determinants for the Success of VR Games
Ellina Nazarova (10809864)
Faculty of Social and Behavioral Sciences, University of Amsterdam MSc Communication Science: Entertainment Communication
Dr. Jeroen S. Lemmens June 25, 2020
Abstract
This paper aims to analyze the relationships between the structural properties of VR games and success. Goal metrics, multiplayer options, number of inputs, play areas, or player modes are assumed to relate to psychological needs, a need for competence, relatedness, or
autonomy, in accordance with self-determination theory (SDT). Fulfillment of psychological needs and support of other structural properties are examined on their predicting power of the success for VR games, which is measured in sales and user ratings. A content analysis of 633 VR games from Steam suggests that structural properties that support a need for competence and a casual genre seem to be the only predicting factors for sales of VR games. However, user ratings of VR games are predicted by structural properties supporting the needs for both competence and autonomy, and by some types of platforms, inputs, play areas, and genres. Besides, a higher number of inputs in more recent VR games leads to higher user ratings, but in older VR games, it leads to lower user ratings. In conclusion, focusing on structural properties related to the needs for competence and autonomy, and other structural properties affecting the success of VR games, can facilitate a more risk-averse game design and help developers appeal to their audience.
Keywords: virtual reality games, self-determination theory, Steam, structural properties, success.
How to Step Up Your VR Game? An Analysis of Structural Properties Related to Psychological Needs as Determinants for the Success of VR Games
The beginning of 2020 gave virtual reality (VR) a significant boost. The release of the first high-quality VR game Half-Life: Alyx and social distancing due to the Corona crisis provided perfect circumstances for the growth in VR users (Curtis, 2020; Valve, 2020). In April 2020, the number of VR users on Steam increased by 50%, which brought the total amount of VR gamers close to two million (Heaney, 2020).
The most common definition of VR is that it is a computer-generated three-dimensional environment which relies on the hand and body tracking of a user, who can interact with virtual objects and receive visual and auditory feedback through a head-mounted display, also referred to as a headset (Kardong-Edgren et al., 2019). For most people, VR is associated with gaming. This belief would not be a misconception as most of the investments related to VR are allocated to the gaming industry (Carvalho, 2018; The Goldman Sachs Group, 2016). VR is unique for its high immersion levels and presence, which makes it attractive for gamers seeking flow in their gaming experiences (Cairns et al., 2014; Pallavicini et al., 2019). Two popular platforms to purchase VR games are the Oculus Store, which is exclusive to titles for Oculus’ headsets, and Steam, the largest digital distribution platform for PC games. Steam seems to be the most suitable source for the general overview of VR titles as it has high inclusivity of brands for VR headsets and it hosts more than 2,000 VR games. This makes Steam the leading distributor for VR content as 96% of VR games can be purchased through its platform (Jank et al., 2019).
Compared to traditional gaming platforms, VR gaming is still in its emergence state, hence, competition within the VR industry is relatively low. The development of content and hardware is stuttered by the so-called chicken-and-egg dilemma. While developers wait for a wide consumer base to own VR hardware before creating high-quality content, consumers
wait for content worthy of buying VR headsets (Wijman, 2019). Additionally, VR being a newer technology and developers having less experience in the field of VR content creation results in higher investment risks for the development when compared to traditional PC gaming (Sharma et al., 2018). Therefore, developers need to know which properties
distinctive to VR games appeal to users and how an implementation of these might facilitate a more risk-averse strategy. Although some games can be remastered from their original
platform into a VR version, games that were created originally for VR seem to have higher quality as they utilize all the potentials of the platform through the use of a three-dimensional environment and motion controllers. A heightened sense of presence and perceived emotions can be essential for the gameplay, where such properties are more prominent in VR games compared to games on traditional platforms (e.g. Pallavicini et al., 2019; Tan et al., 2015). Past research examined features related to the success of mobile and PC games. For example, mobile games from the Google Play store were more successful if they included achievements or customization which led to more downloads (Moreira & Ramalho, 2014). PC games with a high number of players within a game were associated with a higher probability of success (Klimmt & Hartmann, 2006). Intuitive controls and perceived freedom to make choices also predicted the success of PC games (Ryan et al., 2006). In the latter study,
researchers referred to self-determination theory (SDT) to investigate motivations for playing PC games and, thus, predict the success. According to SDT, three psychological needs are essential for the facilitation of intrinsic – the needs for competence, relatedness, and
autonomy (Ryan & Deci, 2000). In their research, authors incorporated intuitive controls and freedom to make choices into the needs for competence and autonomy, which indicated the reliability of SDT in the context of PC games (Ryan et al., 2006). However, no research was found that investigated properties related to psychological needs and their relationship to the
success of VR games. VR games are used for the same entertainment purposes as PC games; thus, it is anticipated that SDT will provide a relevant measure to predict VR games’ success.
According to the sub-theory of SDT, cognitive evaluation theory (CET), conditions that support psychological needs will enhance intrinsic motivations (Ryan et al., 2000). This implies that certain structural properties of VR games that satisfy the needs for competence, relatedness, or autonomy might contribute to the game’s success. Therefore, this paper aims to investigate structural properties that relate to the needs for competence, relatedness, or autonomy in the context of VR games and answer the research question: How do structural properties of a VR game influence the success of a game?
Theoretical Background Conceptualization of Success
It is challenging to assess the success of a VR game as there is no public information available on the number of sales or revenue per title. However, the success of a video game is not just in its monetary value for the developers and publishers but also in the players’
enjoyment of a game. Players are a target audience of any video game; thus, their enjoyment is vital for success (Koch & Bierbamer, 2016).
In their research, Lin et al. (2019) used the number of owners from Steam Spy as an estimate for sales. Steam Spy is a website that automatically gathers data from the Steam store and provides a rough estimate of the number of owners per title (SteamSpy, n.d.). This
measure might be the closest estimate for the monetary success of a video game on Steam. On SteamSpy (n.d.), owners are defined as people that purchased a video game on Steam or in retail and then activated it through Steam. Therefore, the estimated number of owners is a relatively accurate measure for sales.
Another measure used in research to assess the success of a game is its rating (Koch et al., 2016; Joeckel, 2007). For example, a user rating on Steam is calculated by dividing the
number of positive reviews by the total number of reviews of a game (Lin et al., 2019). Higher ratings are given when a user enjoyed a game (Oliver et al., 2016). Ratings indirectly measure players’ enjoyment which influences the game’s success (Siegfried et al., 2015; Tellis & Johnson, 2007). Thus, in this paper, the success of a VR game is measured with a combination of two measures – game sales and user ratings.
Structural Properties Related to Psychological Needs and Success
Intrinsic motivations associated with cognitive and social developments were found to be the core of playing games, where more intrinsically satisfying games increased user
enjoyment (Csikszentmihalyi & Rathunde, 1993; Malone & Lepper, 1987).
Self-determination theory (SDT) refers to intrinsic motivations for actions and their relation to well-being (Deci & Ryan, 1985). According to SDT, the needs for competence, relatedness, and autonomy are essential for the support of intrinsic motivations (Ryan et al., 2000). Multiple research papers applied SDT to gaming motivations in the context of sports, puzzle games, and PC games (Deci, 1975; Hagger et al, 2003; Ryan et al., 2006). All three
psychological needs were correlated with enjoyment (Oliver et al., 2016). A sub-theory, cognitive evaluation theory (CET), aims to specify contextual factors that result in the variability of intrinsic motivations (Ryan et al., 2000). According to CET, conditions supporting psychological needs will incentivize intrinsic motivations. Multiple studies focused on attributing different properties of a video game to one of the three psychological needs to predict a game’s success and user enjoyment (Johnson et al., 2014; Oliver et al., 2016; Ryan et al., 2006). This implies that the structural properties of a VR game that relate to psychological needs will influence intrinsic motivations and, therefore, might affect the game’s success.
A need for competence is defined as a need for mastery and effectiveness (White, 1959). This implies that when playing a VR game, a user seeks a challenge and feedback from
the actions performed in a virtual environment. Goal metrics are basic game elements that provide feedback on the player’s performance in a video game and fulfill a need for a
challenge (Zagal et al., 2005). Therefore, a VR game that includes goal metrics might support a need for competence. There are three properties of a VR game on Steam that fall under this category – leaderboards, achievements, and trading cards. A leaderboard displays scores of users for each game depending on associated goals, which provides feedback in the form of scores and challenges players to get higher on the leaderboard (Steamworks documentation, n.d.-a). Achievements are types of badges that a player receives when completing objectives or executing certain actions (Steamworks documentation, n.d.-b). Trading cards on Steam are virtual cards that are earned by playing video games; they can be traded for in-game badges or exchanged between Steam users (Steam, n.d.). In the research on mobile games, it was found that the presence of achievements and leaderboards in a game led to more downloads
(Moreira et al., 2014). This implies that goal metrics supporting a need for competence might affect higher game sales.
Another psychological need is a need for relatedness, which is defined as a desire to be social and connect with others (Baumeister & Leary, 1995). Video games can be a social activity when played with other people. Most VR games are categorized as single-player games, meaning they can only be played by one user. However, there are also multiplayer VR games that can be played concurrently by more than one user. Multiplayer games amplify social connections and, thus, fulfill a need to connect (Molyneux et al., 2015). This implies that multiplayer games support a need for relatedness. Research regarding PC games illustrates that a higher number of players within a game was associated with a higher probability of success (Klimmt et al., 2006). Similar to the effects of structural properties supporting a need for competence, a multiplayer option relates to a need for relatedness, which increases users’ intrinsic motivation to play a game and their enjoyment. Therefore, a
VR game that supports multiplayer as a feature related to a need for relatedness might affect game sales and its user rating.
Lastly, a need for autonomy is a need to have control and volition, where for video games it implies that users seek flexibility and an ability to choose (Ryan et al., 2000). In a VR game, these needs can be satisfied by a higher number of play areas, types of inputs, or player modes to choose from. A play area is a feature unique to VR games as it defines a user’s physical position and a possibility for a movement during gameplay. Steam categorizes play area into three modes – seated, standing, and room-scale – which can also be chosen according to the preferences of a user. A higher number of included play areas in a VR game is expected to support a need for autonomy.
The term input relates to game mechanics which defines how a user can interact with virtual content (Windleharth et al., 2016). For a traditional PC game, a choice of inputs is usually between a keyboard with a mouse and a gamepad, whereas for a VR game there is also a possibility of using tracked motion controllers. Tracked motion controllers are two wireless controllers that are placed in the left and right hands (Webb et al., 2017). Such controllers are designed to work by responding to their position in a configurated area that tracks a user’s movements and position in space (Pallavicini et al., 2019). VR games are not exclusive to one type of input; some games provide a user with the choice to use a preferred type of input. It is implied that a higher number of inputs in a VR game might support a need for autonomy. A choice of a player mode refers to a choice between a single-player or a multiplayer mode. Therefore, a VR game that includes both of the player modes might support a need for autonomy in comparison with VR games that support only one of the player modes. In conclusion, a higher number of options for each of these structural properties of a VR game is assumed to support a need for autonomy and, thus, positively affect the game’s success.
An ability to make interesting choices in a VR game was found to be a necessary feature for users which increases their intrinsic motivation to play a game (Shelstad et al., 2017). An enhancement of intrinsic motivations will motivate users to play a VR game
resulting in a higher number of sales. In the case of PC games, this relationship was supported where games that included structural properties related to the needs for competence,
relatedness, and autonomy were found to have higher game sales (Ryan et al., 2006). Therefore, the first hypothesis is:
Hypothesis 1: A higher amount of structural properties related to the needs for competence, relatedness, or autonomy will positively affect sales of VR games. The second measure of success is user rating. For a VR game to get high user ratings, players have to enjoy the gameplay. Intrinsic motivation to act influences a person’s
enjoyment, where the supported needs for autonomy and competence in a PC game were found to be associated with higher levels of enjoyment (Oliver et al., 2016; Ryan et al., 2006). Thus, enjoyment from playing a VR game is expected to result in higher user ratings. In the case of traditional video games, higher ratings were found for games that included properties supporting the needs for competence and autonomy (Johnson et al., 2014). However,
following CET, conditions supporting any of the three psychological needs will lead to enhanced intrinsic motivations. Therefore, it is expected that the structural properties of a VR game supporting psychological needs will increase user enjoyment and, thus, result in higher user ratings. The second hypothesis is:
Hypothesis 2: A higher amount of structural properties related to the needs for
competence, relatedness, or autonomy will positively affect user ratings of VR games. Moderation Effects of Recency
Some studies remarked on the effects of recency of a game when examining success factors (Cox, 2014; Koch et al., 2016). VR is a rapidly changing and developing technology
that might affect the quality of experience. For example, playing a VR game three years ago might have been less immersive and less fulfilling for the psychological needs due to
inadequate structural properties, whereas a VR game released just a month ago is of a higher quality and, thus, will have more chances to enhance intrinsic motivations. Additionally, recency effects were found to influence motivational processes (Earley & Erez, 1991), which might imply that there is also a difference in the effects of psychological needs on intrinsic motivations. Therefore, the relationship between the structural properties supporting psychological needs and the success of a VR game might be moderated by the game’s recency (see Figure 1). Such a moderated relationship would imply that users have different preferences regarding newer and older VR games. Thus, the sub-question is:
Sub-question 1: Does the recency of a VR game increase the effects of structural properties related to psychological needs on the game’s success?
Psychological needs Success Leaderboards Achievements Trading cards Multiplayer n of play areas n of inputs n of player modes Recency User rating Sales C om pe te nc e R el at edne ss A uto no my Figure 1
A Conceptual Model for the Relationship Between Structural Properties Related to Psychological Needs and Success Moderated by Recency
Other Structural Properties and Success
Besides structural properties related to psychological needs, there are other features of a VR game that might affect the game’s success. Such structural properties include genre, the type of platform on which a VR game can be played, the type of input, and the type of play area. Contrary to the measures of the number of inputs and the number of play areas for a need for autonomy, the types of inputs and play area are not about the number of options, but the support of the specific input or play area.
A genre of a video game defines a thematic nature of content and gameplay (Arsenault, 2009). Genre is not unique to VR games; any video game or movie can be
categorized into genres. Some basic genres of video games are action, adventure, educational, racing, role-playing (RPG), simulation, sports, strategy, casual, and indie. However, the latter genre – indie – does not refer to the thematic nature or gameplay of a video game. A concept of an indie game comes from the word “independent” and it implies that a game was
developed by one developer or a small team without the help of a publisher or licensor (Lin et al., 2019). Regardless of its irrelevancy to a thematic nature or gameplay, indie seems to be a common category when defining the genre of a video game. Multiple studies found that various genres of mobile and PC games influenced the profitability and the success of a game (Cox, 2014; Lee & Raghu, 2014; Marchand et al., 2013). Even though there are conflicting findings, it is evident to expect a difference in the success of VR games for various genres.
In the case of VR games, platforms are headsets that run a game. There are four types of headsets available as a gaming platform for VR games on Steam: HTC Vive, Valve Index, Oculus Rift, and Windows Mixed Reality. Just like traditional video games, VR games can be exclusive to one platform or be available on multiple ones. It is common for developers to struggle to choose which platform to focus on when developing a VR game as there is
apparent segregation of the target audience (Dredge, 2016). It might be interesting to see if VR users appeal to a specific platform.
Similar to platforms, various inputs might also have unique effects on the success of a VR game. The inclusion of a specific type of input could affect success, while another type of input would not. For example, tracked motion controllers seem to work best for VR games as they make use of full immersion in a three-dimensional virtual environment (Anthes et al., 2016). This might imply that the inclusion of tracked motion controllers could be an appealing structural property of a VR game for users.
Next, there are three types of play areas. First, is a seated mode which is the closest to PC games and VR’s least interactive gaming technique. In a seated VR game, a user remains seated on a chair and has a 360-degree view, which enables a user to look around. During a seated VR game, a user cannot move away from the point of view (Shewaga et al., 2017). Second, a standing option is a play area that has the same properties as a seated type but a user is standing. In a standing VR game, a user can perform small lateral movements and thus experience the depth of a virtual environment (Shanklin, 2017). Third and last, is a room-scale play area. In a VR game that allows a room-room-scale, a user can physically walk in a configurated area, which mirrors movements in the real world into the virtual environment (Shewaga et al., 2017). The inclusion of each of the play area modes might have different effects on the success of a VR game. It is interesting to see which of the play areas is preferred by VR users. Therefore, the next sub-question is:
Sub-question 2: Which other structural properties influence the success of VR games? Method
Sample
The total number of VR games collected for the analysis was 3,647. On April 5th, 2016 HTC Vive headset was released, which was the first high-quality headset bringing VR
gaming to a wider consumer public (Dempsey, 2016). Therefore, only VR games published after 2016 were chosen for the analysis as any VR games developed before the release of the HTC Vive cannot be compared to titles on the present VR platforms. The data was collected on April 16th, 2020; thus, it does not include VR games or information updated after this date. Furthermore, VR games not included under the category “VR only” were excluded from the list as these titles imply that they are also playable on PC or other platforms (N = 523). Such VR games usually differ in their gameplay and structural properties from games exclusive to a VR platform. Next, free-to-play and early access VR games were excluded (N = 1,180). VR games in these categories have different monetization strategies and, thus, different measures for success (Lin et al., 2019; Marchand & Hennig-Thurau, 2013). To exclude personal projects and VR games of low-quality, titles with less than 25 reviews were excluded (N = 1,311; Lin et al., 2019). This left 633 VR titles included in the sample.
The sample included 179 VR games released in 2016, 225 were released in 2017, 128 in 2018, and only 85 and 16 VR games were released in 2019 and 2020 respectively. The highest number of reviews for a VR game was 123,625 where the lowest was 26, due to filtering out VR games with less than 25 reviews (M = 639.58, SD = 5,322.71).
Procedure
A methodological approach to answer the research questions and to test the
hypotheses was to gather data from multiple sources that include the necessary information for the analysis. First, an application program interface (API) for Steam Spy was used to collect a list of video games from Steam that were tagged as “VR game”. The list contained such information as title, developer, publisher, consumer rating, number of positive reviews, number of negative reviews, the number of owners, and price.
Second, a code for the web scraper was manually written in Python to collect the data from the Steam store (see Appendix 1 for the example of a code). The data were collected
only on the VR games that were already on the list collected from Steam Spy. The information from the Steam store included the type of application, supported languages, suitable
platforms, categories, genres, achievements, and release dates. Both datasets were merged. Third, as previously described under the sample, the data was cleaned by type, categories, and the number of reviews. Fourth and last, the data was imported into Statistical Package for the Social Sciences (SPSS) for further statistical analysis.
Measures Success
The success of a VR game is measured by two variables – game sales and user rating. An initial measure for sales is presented in eight ranges. The range is coded between 1 and 8 where 1 = 0 to 20,000, 2 = 20,000 to 50,000, 3 = 50,000 to 100,000, 4 = 100,000 to 200,000, 5 = 200,000 to 500,000, 6 = 500,000 to 1,000,000, 7 = 1,000,000 to 2,000,000, and 8
= 2,000,000 to 5,000,000. However, game sales were unevenly distributed within these ranges, where 76.5% of VR games had between 0 and 20 thousand sales (M = 1.49, SD = 1.08). Therefore, a new measure was created by computing a mean of existing ranges, where the minimum number of sales was 10,000 and the maximum 3,500,000 (M = 44,620.85, SD = 180,154.69).
Steam uses a scale in percentages for its user rating, thus, between 0% (bad) and 100% (good). The minimum user rating from the list of 633 VR games was 22.41% with a
maximum of 100% (M = 78.28, SD = 14.52). There was a significant positive correlation between sales and user ratings; VR games with high sales also had high user ratings and vice-versa, r = 0.14, p < .001. However, it is not possible to combine the measures for sales and user ratings into one variable due to their different scales. Thus, the effects of independent variables on game sales and user ratings were analyzed separately.
Competence
A need for competence was measured by the inclusion of goal metrics in a VR game – leaderboards, achievements, and trading cards. Each of the goal metrics was measured with 0 = not supported or 1 = supported. Leaderboards were supported in 19.6% of VR games (SD = 0.40), 47.9% supported achievements (SD = 0.50), and 20.7% included trading cards as a feature (SD = 0.41). An aggregated competence was measured by the number of goal metrics supported in a VR game that ranged between 0 and 3, where 0 implied that there were no goal metrics and 3 that all of the goal metrics were supported (M = 0.88, SD = 0.92). There was a significant, moderately strong positive correlation between the support of achievements and of trading cards, which implies that if achievements are supported, trading cards are likely to be included too, r = .36, p < .001. The support of achievements was also positively correlated with the support of leaderboards; if a VR game supported achievements, it was also likely to have leaderboards, r = .25, p < .001. However, there was no significant correlation between the support of trading cards and leaderboards, r = .07, p = .070.
Relatedness
A need for relatedness was measured by one structural property of a VR game – support of a multiplayer. In this paper, multiplayer includes games that feature cooperative and player-versus-player game modes. The variable was measured with 0 = not supported or 1 = supported, where 21.8% of VR games included a multiplayer mode and, thus, supported a need for relatedness (SD = 0.41).
Autonomy
A need for autonomy was measured by the number of play areas, inputs, and player modes. The number of play areas was measured on a range between 1 and 3, which indicated the number of available play areas in a VR game (M = 2.08, SD = 0.76). Thirty-three percent of VR games included all three play areas, 41.9% included two play areas, and 25.1%
supported only one type of play area. Similarly, the number of inputs was measured on a range between 1 and 3, where only 6.6% included all three inputs, 14.2% included two types of inputs, and 79.1% supported only one type of input (M = 1.27, SD = 0.58). Next, the number of player modes was measured with 0 = no choice or 1 = choice. A choice between a multiplayer and single-player was coded as 1, and VR games with only one type of player modes, such as only single-player or only multiplayer, were coded as 0 (M = 0.19, SD = 0.40). An aggregated autonomy was measured by the sum of the number of all the supported play areas, inputs, and player modes in a VR game. The minimum number of structural properties in a VR game related to a need for autonomy was 2 and the maximum was 7 (M =3.55, SD = 1.03).
Structural Properties
There is a moderator, recency, and four other structural properties as independent variables – platforms, inputs, play areas, and genres. The recency of a VR game was
measured with the difference in days between its release date and the release date of the oldest VR game from the list, April 28th, 2016. Thus, a VR game published more recently had a higher value for recency, where for the older VR game the value was lower. The highest value for recency was 1,473 and the lowest was 0 (M = 565.40, SD = 387.23).
Steam hosts games that run on five headsets: HTC Vive, Valve Index, Oculus Rift, and Windows Mixed Reality. Each of the platforms was coded as a separate variable between 0 = not supported or 1 = supported. From a total of 633, the Valve Index headset was
supported for the lowest amount of 111 VR games (M = 0.18, SD = 0.38), whereas the highest amount of 630 supported the HTC Vive headset (M = 1.00, SD = 0.07). Oculus Rift also had a relatively high number of 541 VR games that supported the platform (M = 0.85, SD = 0.35). For the Windows Mixed Reality headset, the number was only 207 (M = 0.33, SD = 0.47). Three types of inputs were coded into separate variables with 0 = not supported or 1 =
supported. The least number of VR games supported a keyboard with a mouse with a total of 100 titles (M = 0.16, SD = 0.37). Next, 121 VR games supported a gamepad as an input (M = 0.19, SD = 0.39). The highest proportion of 586 VR games supported tracked motion
controllers (M = 0.93, SD = 0.26). Similarly, three play area modes were coded 0 = not supported or 1 = supported. A standing play area was supported by 491 VR games (M = 0.78, SD = 0.42), whereas a room-scale option by 463 (M = 0.73, SD = 0.44), and a seated play area by only 362 VR games (M = 0.57, SD = 0.50). VR games are not exclusive to one of the genres, which implies that they can include multiple categories. There were nine variables for different genres, which were also measured with 0 = no or 1 = yes: action (53.1%),
adventure (38.2%), indie (64.3%), casual (35.1%), simulation (32.4%), sports (10.6%), strategy (8.1%), RPG (7.0%), and racing (3.6%).
Results
The Effects of Structural Properties Related to Psychological Needs on Success
Multiple linear regressions were performed to analyze the relationships between game sales and structural properties related to the needs for competence, relatedness, or autonomy. Table 1 shows that only achievements related to a need for competence had a significant positive effect on game sales. The inclusion of a multiplayer option that supports a need for relatedness did not have a significant effect on sales. Similarly, support of a need for
autonomy did not affect sales. This implies that the first hypothesis was partially approved as only support of competence (achievements) increased game sales.
To analyze the relationship between structural properties related to psychological needs and user rating, multiple linear regressions were performed (see Table 1). The measure for aggregated competence significantly increased user ratings of VR games. Support of achievements increased user ratings, whereas support of trading cards in VR games decreased user ratings. Similar to sales, relatedness (multiplayer) did not influence user ratings. A higher
number of play areas increased user ratings, whereas a higher number of inputs decreased them. Therefore, hypothesis 2 is partially approved as only support of the needs for
competence (achievements and aggregated competence) and autonomy (number of play areas) increased user ratings of VR games. Additionally, some measures of the needs for
the opposite direction of the effect that was predicted. The relationships between structural properties related to psychological needs and the success of a VR game were analyzed for moderation effects by the recency of a game using the model 1 in PROCESS macro (see Table 2). Only the relationship between the number of inputs and user ratings was moderated by recency. For newer VR games there was a positive relationship between the number of inputs and user ratings, but a negative relationship for older VR games. Thus, recency of a VR game did not have a strong moderating effect on the relationship between structural
properties related to the needs for competence or relatedness and game's success, R2 < .04, as recency only moderated the relationship between a need for autonomy (number of inputs) and user ratings.
The Effects of Other Structural Properties on Success
The independent sample t-tests were performed to analyze the effects of other
structural properties on game sales (see Table 3). None of the platforms, inputs, or play areas, except casual genre, influenced game sales. Casual VR games had fewer sales when
and genres on user ratings were examined. Table 4 shows that VR games received higher user ratings when Valve Index or Windows Mixed Reality were supported. However, the support of a gamepad as an input decreased user ratings of VR games. A seated play area had a significant negative effect on user ratings, whereas standing and room-scale increased user ratings. Adventure and simulation genres received lower user ratings than non-adventure and non-simulation VR games, respectively. However, VR games categorized under the indie genre received higher user ratings in comparison with non-indie VR games.
Discussion
This paper aimed to examine how the structural properties of VR games influence success. First, hypothesis 1 stated that a higher amount of structural properties related to psychological needs will positively affect sales of VR games. This hypothesis was partially approved as only one of the structural properties, achievements, related to a need for competence had a positive effect on game sales. This finding is in line with the results of Moreira et al. (2014), who examined mobile games and found that the inclusion of achievements led to more downloads. However, other properties related to a need for competence did not influence sales. Therefore, supporting a need for competence by providing VR users with challenges and feedback did not have a strong influence on their intrinsic motivations to purchase a VR game. Such an outcome could be explained by a very low percentage of 26.6% VR games having more than one game metric in gameplay.
Accordingly, developers of VR games from Steam seem to not focus on game metrics such as leaderboards, achievements, or trading cards, which might indicate non-essentiality of these structural properties in VR games.
A multiplayer option related to a need for relatedness did not affect sales of VR games, where other research papers examining PC games found a significant effect on this relationship (Klimmt et al., 2006; Ryan et al., 2006). The difference in the results between VR games and PC games could be due to PC games being more popular among gamers, which makes a population in multiplayer games denser than for VR games with a relatively low user base. Therefore, a need for relatedness might be a significant predictor for sales of PC games but not for VR games.
There was also a significant positive relationship between structural properties related to a need for autonomy and sales for PC games (Ryan et al., 2006). However, for VR games, the number of play areas, inputs, and player modes related to autonomy did not influence
game sales. This is an opposite result from the outcomes of a study, which found that an ability to make interesting choices in a VR game positively affected intrinsic motivations to play a game (Shelstad et al., 2017). This difference in the results could be due to the use of different features for PC and VR games related to a need for autonomy, as for VR games such structural properties are unique to VR. Besides, other research papers investigating support of a need for autonomy for traditional platforms measured flexibility and an ability to choose within the gameplay (Ryan et al., 2006; Shelstad et al., 2017), where, in this paper, structural properties related to a need for autonomy measured a general structure of a game. Thus, flexibility and ability to choose does not appeal to VR users in terms of buying decisions for VR games. According to the results of this paper, structural properties related to the needs for relatedness and autonomy do not predict the sales of VR games, where a need for competence has a weak predicting effect.
Second, hypothesis 2 predicted that a higher amount of structural properties related to psychological needs will positively affect user ratings of VR games. Similar to the first hypothesis, hypothesis 2 was partially approved as only an aggregated measure for competence, support of achievements, and the number of play areas related to a need for autonomy increased user ratings. An aggregated competence was found to have a significant positive effect on user ratings of VR games. This implies that a higher number of goal metrics – achievements, leaderboards, and trading cards – related to a need for competence increases user ratings of a VR game. Furthermore, as in the case with game sales, the inclusion of achievements related to a need for competence had an individual positive effect on user ratings of VR games. These findings are in line with the results from other research papers, which found that a supported need for competence in traditional and PC games increased ratings (Johnson et al., 2014; Oliver et al., 2016). However, trading cards decreased user ratings, which could be due to the reasoning behind the use of trading cards. Even though the
inclusion of trading cards is part of goal metrics, some people use them to get financial profit from earning and then reselling them or for other tangible benefits such as “unique Steam chat emoticons, backgrounds, and sometimes game discounts” (Roberts, 2016). Accordingly, the negative effects of the inclusion of trading cards on user ratings could be attributed to other motives and, thus, be related to diverse types of needs rather than to only a need for
competence.
The support of a need for relatedness did not influence user ratings of VR games. As mentioned above, the user base for VR is less dense than for PC games, which implies that a multiplayer option might be less relevant to a VR user. Therefore, no effect of a multiplayer on user ratings could be explained with a low amount of VR games that include a multiplayer option, which then makes this variable less reliable for comparison.
Traditional and PC games that included properties related to a need for autonomy were also found to affect higher ratings (Johnson et al., 2014; Oliver et al., 2016; Ryan et al., 2006). This paper supported this relationship for VR games, where the number of play areas related to a need for autonomy increased the game’s user ratings. Therefore, VR games that support a need for autonomy by providing flexibility in choosing the type of play areas, appeals to VR users. This is an important finding, as the choice of play areas is a unique property for VR platforms. On the other hand, it was found that a higher number of inputs decreased user ratings. VR games that use other than a native VR input – tracked motion controllers – possibly use different in-game mechanics for interacting with a virtual environment, which might feel less natural for VR users, thus affect in lower user ratings (Anthes et al., 2016).
Third, one of the sub-questions examined the effects of the recency on the
relationships between structural properties related to psychological needs and the success of VR games. Recency did not moderate the relationships between any of the psychological
needs and game sales. The only relationship moderated by recency was between the number of inputs and user ratings, where more recent VR games with a higher number of inputs received higher user ratings, but older VR games with a higher number of inputs received lower user ratings. Besides, newer VR games had generally higher user ratings than older VR games despite the number of supported inputs. These findings could be explained by more advanced techniques used with time in the development of VR games (Carvalho, 2018), where for older VR games multiple inputs might have disturbed an experience due to lower quality, but for newer VR games, more inputs might work better, thus enhancing user experience.
Fourth and last, the second sub-question aimed to look at the effects of other structural properties on the game’s success. Casual VR games had fewer sales when compared to non-casual games. This implies that VR users prefer a non-non-casual genre when deciding to purchase a VR game. Next, VR games that support the Valve Index headset received higher user ratings. Similarly, the support of the Windows Mixed Reality headset increased user ratings. These findings suggest that VR users enjoy the most playing VR games that support Valve Index or Windows Mixed Reality platforms. Support of a gamepad decreased user ratings, which shows that VR games supporting this input do not appeal to VR users. This could be due to a gamepad not being an intuitive interaction mechanic for VR, which then affects the gameplay and decreases user enjoyment and thus user ratings (Ryan et al., 2006). Support of a seated play area did not appeal to VR users, where the support of a standing or room-scale play area increased user ratings. These findings are in line with the concept of immersion in VR, which defines VR games with a seated play area as low immersion and room-scale as high (Shewaga et al., 2017). VR games under adventure or simulation genres received lower user ratings than non-adventure and non-simulation VR games, respectively. Therefore, adventure or simulation genres seem to be not enjoyable for VR users. However,
VR games under an indie genre received higher user ratings in comparison with non-indie games, which means that VR games created by independent developers appeal to gamers more than other VR games. This is an interesting finding as usually indie games can be of lower quality when compared to games developed by big studios.
Limitations
There are some limitations to this paper. First, the measure for sales was unevenly spread which might have affected the outcomes of the analyses. Additionally, some consumers are unable to always properly evaluate structural properties related to
psychological needs at a pre-purchase stage. Therefore, the effects of structural properties related to psychological needs on game sales are obscured by the measures used in this paper. Second, there was only one structural property related to a need for relatedness and three properties for each of the needs for competence and autonomy, which reduced the validity of measures for psychological needs. For this research, these were the only available measures from the data collected on Steam and Steam Spy. Third, only VR games purchased
through Steam were used in the analyses for sales, which excludes VR games sold on other platforms. Thus, the number of sales used in this paper might be underrated as it does not include users that purchased a game but did not activate it through Steam. Fourth, other variables can affect the success of VR games, for example, the marketing strategy and content of the gameplay. Such features could potentially influence the success of VR games or the relationships between structural properties and success. However, these variables were not accounted for in this research.
Directions for Future Research
For future research, it might be interesting to examine other variables prominent in VR games. For example, to look into the relationship between the content of a VR game and the game’s success. It could help to better understand what appeals to VR users. Also, it is
recommended to use another measure for success rather than sales, as in the context of VR games it is too unevenly spread. Additionally, when applying STD, future studies should make use of more structural properties in a VR game related to psychological needs, which would help investigate their effects on the success of VR games in more detail. Lastly, examining other platforms that sell VR games, such as the Oculus Store, could show if there are differences regarding different sources for purchasing VR games. That being said, much research remains to be done before understanding the scope of variables affecting the success of VR games.
Conclusion
As the success of VR games was measured in sales and user ratings, the effects of structural properties differ per measure. A need for competence and casual genre appear to be the only predicting factors for sales of VR games. Thus, the structural properties of VR games do not have strong predicting power for game sales. However, user ratings of VR games are more predictable as they are influenced by structural properties supporting the needs for competence and autonomy, and by some types of platforms, inputs, play areas, and genres. It is also important to note that even though VR technology rapidly advances, the recency of VR games only moderated the relationship between the number of inputs related to a need for autonomy and game’s user ratings, which implies that, in context of psychological needs, VR users do not perceive the difference between older and newer VR games.
Due to the method of using the whole population of VR games from Steam in the analyses, the results of this paper can be applied to most of the VR games that are exclusive to a VR platform. This paper shows that determinants of the success of VR games moderately differ from PC and other traditional games. Also, Steam is used by a wide range of gamers with different backgrounds which implies broad generalizability for the population of VR users. According to the results of this research, focusing on structural properties related to the
needs for competence and autonomy might help developers create a more risk-averse game design. Furthermore, creating games that support Valve Index, Windows Mixed Reality, standing play area, and room-scale play area will aid in appealing to their audience. In the case of the first high-quality VR game Half-Life: Alyx, which has a user rating of 98,7% and sold around 750,000 copies on Steam, all the structural properties – except its genre – are in accordance with the results of this study. Even though this is a one-case example, it
demonstrates practical implications for the outcomes of this paper.
In conclusion, the findings of this study are important for VR professionals as high generalizability of results imply that the effects apply to a broad range of VR games and audiences. Despite the limitations of this study, it is one of the first ones to examine
determinants of the success of VR games. Therefore, one of the aims of this study is to aspire more research into investigating unique features that appeal to the success of VR and its audience.
References
Anthes, C., García-Hernández, R. J., Wiedemann, M., & Kranzlmüller, D. (2016, March). State of the art of virtual reality technology. In 2016 IEEE Aerospace Conference,
1-19. https://doi.org/10.1109/AERO.2016.7500674
Arsenault, D. (2009). Video game genre, evolution and innovation. Journal for computer game culture, 3(2), 149-176.
https://www.eludamos.org/index.php/eludamos/article/view/vol3no2-3/125
Baumeister, R. F., & Leary, M. R. (1995). The need to belong: desire for interpersonal attachments as a fundamental human motivation. Psychological bulletin, 117(3), 497.
https://doi.org/10.1037/0033-2909.117.3.497
Cairns, P., Cox, A., & Nordin, A. I. (2014). Immersion in digital games: review of gaming experience research. Handbook of digital games, 337–361.
https://doi.org/10.1002/9781118796443.ch12
Carvalho, P. (2018, July). Following the white rabbit: The virtual reality for games. In International Conference on Virtual, Augmented and Mixed Reality, 172-183.
https://doi.org/10.1007/978-3-319-91584-5_14
Cox, J. (2014). What makes a blockbuster video game? An empirical analysis of US sales data. Managerial and Decision Economics, 35(3), 189-198.
https://doi.org/10.1002/mde.2608
Curtis, T. (2020, April 16). Half Life: Alyx and how the menial becomes the magnificent in VR. VRfocus. https://www.vrfocus.com/2020/04/half-life-alyx-and-how-the-menial-becomes-the-magnificent-in-vr/
Csikszentmihalyi, M., & Rathunde, K. (1993). The measurement of flow in everyday life: Toward a theory of emergent motivation. In Jacobs, J. E. (Ed.), Current theory and
research in motivation, 40. Nebraska Symposium on Motivation, 1992: Developmental perspectives on motivation, 57–97. University of Nebraska Press.
Deci, E. L. (1975). Intrinsic motivation. Springer. https://doi.org/10.1007/978-1-4613-4446-9 Deci, E. L., & Ryan, R. M. (1985). Conceptualizations of intrinsic motivation and
self-determination. Intrinsic motivation and self-determination in human behavior, 11–40.
https://doi.org/10.1007/978-1-4899-2271-7_2
Dempsey, P. (2016, August). The teardown: HTC Vive VR headset. Engineering &
Technology, 11 (7-8), 80-81. https://doi.org/10.1049/et.2016.0731
Dredge, S. (2016, November 10). The complete guide to virtual reality – everything you need to get started. The Guardian.
https://www.theguardian.com/technology/2016/nov/10/virtual-reality-guide-headsets-apps-games-vr
Earley, P. C., & Erez, M. (1991). Time-dependency effects of goals and norms: The role of cognitive processing on motivational models. Journal of Applied Psychology, 76(5), 717–724. https://doi.org/10.1037/0021-9010.76.5.717
Hagger, M. S., Chatzisarantis, N. L. D., Culverhouse, T., & Biddle, S. J. H. (2003). The processes by which perceived autonomy support in physical education promotes leisure-time physical activity intentions and behavior: A trans-contextual
model. Journal of Educational Psychology, 95(4), 784-795.
https://doi.org/10.1037/0022-0663.95.4.784
Heaney, D. (2020, May 3). Half-Life: Alyx catapults VR headsets on Steam to 2 million users. UploadVR. https://uploadvr.com/steam-hardware-survey-2-million/
Jank, A., Kozich, F., Fröhlich, F., & Erenli, K. (2019, June). VR fact sheet 2019 - An overview of VR movies & games. In 2019 IEEE Games, Entertainment, Media
Joeckel, S. (2007). The impact of experience: The influences of user and online review ratings on the performance of video games in the US market. DiGRA Conference.
Johnson, D., Watling, C., Gardner, J., & Nacke, L. E. (2014, October). The edge of glory: the relationship between metacritic scores and player experience. In Proceedings of the first ACM SIGCHI annual symposium on Computer-human interaction in play, 141-150. https://doi.org/10.1145/2658537.2658694
Kardong-Edgren, S. S., Farra, S. L., Alinier, G., & Young, H. M. (2019). A call to unify definitions of Virtual Reality. Clinical Simulation in Nursing, 31, 28-34.
https://doi.org/10.1016/j.ecns.2019.02.006
Klimmt, C., & Hartmann, T. (2006). Effectance, self-efficacy, and the motivation to play video games. Playing video games: Motives, responses, and consequences, 133-145. Koch, S., & Bierbamer, M. (2016). Opening your product: impact of user innovations and
their distribution platform on video game success. Electronic Markets, 26(4), 357-368.
https://doi.org/10.1007/s12525-016-0230-5
Lee, G., & Raghu, T. S. (2014). Determinants of mobile apps' success: Evidence from the app store market. Journal of Management Information Systems, 31(2), 133-170.
https://doi.org/10.2753/MIS0742-1222310206
Lin, D., Bezemer, C. P., Zou, Y., & Hassan, A. E. (2019). An empirical study of game reviews on the Steam platform. Empirical Software Engineering, 24(1), 170-207.
https://doi.org/10.1007/s10664-018-9627-4
Malone, T. & Lepper, M. (1987). Making learning fun: A taxonomy of intrinsic motivations in learning. In R. E. Snow & M. J. Farr (Eds.), Aptitude, learning and instruction: vol. 3. Conative and affective process analyses, 223–253. Lawrence Erlbaum.
Marchand, A., & Hennig-Thurau, T. (2013). Value creation in the video game industry: Industry economics, consumer benefits, and research opportunities. Journal of
Interactive Marketing, 27(3), 141-157. https://doi.org/10.1016/j.intmar.2013.05.001
Molyneux, L., Vasudevan, K., & Gil de Zúñiga, H. (2015). Gaming social capital: Exploring civic value in multiplayer video games. Journal of Computer-Mediated
Communication, 20(4), 381-399. https://doi.org/10.1111/jcc4.12123
Moreira, Á. V. M., & Ramalho, G. (2014). Which features matter to make a successful mobile game? SBC Journal on Interactive Systems, 5(2), 2-13.
https://seer.ufrgs.br/jis/article/view/45696/31858
Oliver, M. B., Bowman, N. D., Woolley, J. K., Rogers, R., Sherrick, B. I., & Chung, M. Y. (2016). Video games as meaningful entertainment experiences. Psychology of Popular
Media Culture, 5(4), 390. https://doi.org/10.1037/ppm0000066
Pallavicini, F., Pepe, A., & Minissi, M. E. (2019). Gaming in virtual reality: What changes in terms of usability, emotional response and sense of presence compared to
non-immersive video games? Simulation & Gaming, 50(2), 136-159.
https://doi.org/10.1177/1046878119831420
Roberts, K. (2016, June 26). Re: Why do people buy Steam trading cards? [Discussion post]. Quora. https://www.quora.com/Why-do-people-buy-Steam-Trading-Cards
Ryan, R. M., & Deci, E. L. (2000). Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist, 55(1), 68–
78. https://doi.org/10.1037/0003-066X.55.1.68
Ryan, R. M., Rigby, C. S., & Przybylski, A. (2006). The motivational pull of video games: A self-determination theory approach. Motivation and emotion, 30(4), 344-360.
The Goldman Sachs Group (2016). Virtual and augmented reality: understanding the race for the next computing platform.
https://www.goldmansachs.com/insights/pages/technology-driving-innovation-folder/virtual-and-augmented-reality/report.pdf
Shanklin, W. (2017, March 2). Room-scale or standing VR? Why all that walking around may
be overrated. New Atlas.
https://newatlas.com/room-scale-vs-standing-vr-oculus-rift-htc-vive/48219/
Sharma, A., Bajpai, P., Singh, S., & Khatter, K. (2018). Virtual reality: Blessings and risk assessment. Indian Journal of Science and Technology, 11, 1-20.
https://doi.org/10.17485/ijst/2018/v11i20/112577
Shelstad, W. J., Smith, D. C., & Chaparro, B. S. (2017, September). Gaming on the rift: How virtual reality affects game user satisfaction. In Proceedings of the Human Factors and Ergonomics Society Annual Meeting, 61(1), 2072-2076. SAGE Publications.
https://doi.org/10.1177/1541931213602001
Shewaga, R., Uribe-Quevedo, A., Kapralos, B., & Alam, F. (2017). A comparison of seated and room-scale virtual reality in a serious game for epidural preparation. IEEE transactions on emerging topics in computing.
https://doi.org/10.1109/TETC.2017.2746085
Siegfried, N., Koch, O., & Benlian, A. (2015). Drivers of app installation likelihood–a conjoint analysis of quality signals in mobile ecosystems. International Conference on Information Systems (ICIS 2015), Fort Worth, Texas.
Steam. (n.d.). Steam trading cards. https://steamcommunity.com/tradingcards/
SteamSpy. (n.d). Learn about Steam Spy. https://steamspy.com/about
Steamworks documentation. (n.d.-a). Steam leaderboards.
Steamworks documentation. (n.d.-b). Stats and achievements.
https://partner.steamgames.com/doc/features/leaderboards
Tan, C. T., Leong, T. W., Shen, S., Dubravs, C., & Si, C. (2015). Exploring gameplay experiences on the Oculus Rift. In Nacke, L. E., & Graham, T. C. N. (Eds.),
Proceedings of the 20 Annual Symposium on Computer-Human Interaction in Play -
CHI PLAY ’15, 253-263. ACM Press. https://doi.org/10.1145/2793107.2793117
Tellis, G. J., & Johnson, J. (2007). The value of quality. Marketing Science, 26(6), 758-773.
https://doi.org/10.1287/ mksc.1070.0286
Valve (2020). Half-life: Alyx. Retrieved May 5, 2020, from https://www.half-life.com/en/alyx
Webb, J. S., Rogoza, B. E. T., Bristol, P. W., Higgins, J. A., Talati, S. S., Chen, Y. Y., & Konzen, N. W. (2017). U.S. Patent No. 9,678,566. U.S. Patent and Trademark Office. White, R. W. (1959). Motivation reconsidered: The concept of competence.Psychological
review,66(5), 297. https://doi.org/10.1037/14156-005
Wijman, T. (2019, December 12). Newzoo’s games trends to watch in 2020. Newzoo.
https://newzoo.com/insights/articles/newzoos-games-trends-to-watch-in-2020/
Windleharth, T. W., Jett, J., Schmalz, M., & Lee, J. H. (2016). Full steam ahead: A conceptual analysis of user-supplied tags on Steam. Cataloging & Classification
Quarterly, 54(7), 418-441. https://doi.org/10.1080/01639374.2016.1190951
Zagal, J., Mateas, M., Fernandez-Vara, C., Hochhaleter, B., & Lichti, N. (2005). Towards and ontological language for game analysis. In Castell, S., & Jennifer, J. (Eds.), Selected papers of the Digital Interactive Games Research Association’s second international conference (DiGRA 205), 3-14. Digital Interactive Games Research Association.
Appendix A
Example of a Code from the Manually Developed Web-Scraper for Steam in Python
def parse_steam_request(appid, name):
url = "http://store.steampowered.com/api/appdetails/" parameters = {"appids": appid}
json_data = get_request(url, parameters=parameters) json_app_data = json_data[str(appid)]
if json_app_data['success']: data = json_app_data['data'] else:
data = {'name': name, 'steam_appid': appid} return data
download_path = ''
steam_app_data = 'app_list_.csv' steam_index = 'steam_index.txt' steam_columns = [
'type', 'name', 'steam_appid', 'is_free', 'controller_support', 'short_description', 'fullgame',
'supported_languages', 'developers', 'publishers',
'platforms', 'metacritic', 'reviews', 'categories', 'genres', 'achievements', 'release_date',
]
reset_index(download_path, steam_index)
index = get_index(download_path, steam_index)
prepare_data_file(download_path, steam_app_data, index, steam_columns) process_batches( parser=parse_steam_request, app_list=app_list, download_path=download_path, data_filename=steam_app_data, index_filename=steam_index, columns=steam_columns, begin=index,
end=4000, batchsize=5 )
