Keywords

1 Introduction

The video game industry, as a fast-growing and expanding industry worldwide, has produced spectacular global revenue of 100 billion dollars each year, amounting to approximately three times the revenue of the movie industry. Storytelling in games, meanwhile, is drawing much attention worldwide, not only because of its fundamental role in communicating emotion, conveying values and enhancing immersion but also due to the growing awareness throughout the global game industry to the strategic reservation of IPs (intellectual property). One can take the recent battle for the adaption right of a Chinese online novel as an example. This novel, titled Wu Xing Tian, with only 1% written, was finally sold to a local game company for 8 million RMB (approximately 1.2 million dollars). Another notable trend is the increasing prevalence of gamification projects in companies outside the game industry, where story creation is also of great significance to the project leaders and developers.

However, no comprehensive model has been found to evaluate the quality of stories for the game adaption, which we refer to as the game-adaptability of stories. Some existing works, which may contribute to such evaluation, lie in three main areas: scriptwriting in the traditional movie industry [1,2,3], game immersion and playability [4,5,6,7,8], and game motivation studies [16, 17, 21]. However, none of these models are specific to the creation, selection and adaption of game stories. A game story, more specifically, is different from a movie script in terms of interactivity and non-linear characteristics [9], and other game features (e.g., mechanism design, difficulty design). It is also different with regard to spiritual impact – the way in which it conveys values, communicates emotion and influences the understanding of life and humanity.

This paper aims to explore the factors of the stories that influence game adaptability. Three steps were taken in this study: (1) Determine variables based on literature and empirical studies, including content analysis and interviews; (2) Survey construction and distribution to collect players’ ratings of the importance of each variable to game playability; and (3) exploratory factor analysis to extract a concise model from the variables. The potential implementation of the study is to provide guidance for game story writing, selection and adaption, not only for game designers but also for the leads of gamification projects in a wider domain.

2 Related Work

2.1 Game Storytelling

The definition of game storytelling is always confused with that of game narrative. Some researchers have referred to game narrative in discussions that are actually about storytelling [10]. To clarify, the core of the story is the description of the process of a character confronting a series of obstacles on the way to obtaining the goal [3]. There is a structure to most stories, which often consists of a beginning, a middle and an ending, according to Aristotle. And the main elements of a story include events (actions and happenings) and existents (characters and setting) [1]. However, narrative is the method and style used to tell the story of the game [9, 11]. The story can be created either by game designers or by players themselves. Therefore, not all games tell stories, but many games do have narratives [12]. For example, fighting games usually do not have storytelling, but may have some narratives. Majewski [10] observed that not all genres of games tell stories, but some genres are mainly dependent on stories, such as first-person shooter games, role-play games and adventure games.

Although previous research regarding screen scriptwriting exists in the film industry [2], these works are not adequate for the analysis of game story writing or selection for two main reasons [9]: film scripts lack interactivity and have a linear structure. According to the famous game designer Chris Crawford, what a computer game really sells is interactivity. A traditional story is entirely created by the author, and the audience rarely interacts with the story. A game story, however, involves users in the story’s creation. Actions performed by players may be “listened to” and “considered” by the game engine and may influence the outcome [13]. Non-linearity is another characteristic of a game story. While the structure of a traditional story is completely fixed the moment when the author finishes the story, game stories have a myriad of possibilities for their structure. Meadows [14] summarized three plot structures in game storytelling. According to these plot structures, each game story may include several main storylines and crossover between the storylines. The elements of storytelling are summarized in Table 1.

Table 1. Variables derived from storytelling models
Full size table

2.2 Models of Motivation for Game Engagement

A simple and intuitive reason for game play is “because games are fun”. However, fun is such an ambiguous expression that it can also be used to describe many other activities, such as watching a movie, singing, or dancing. However, almost none of these forms of entertainment are comparable to games with regard to their appeal, sustained engagement and effect on well-being [15].

Both empirical literature [16, 17] and theoretical research from a psychological perspective [15, 18] noted that games can provide experiences that satisfy universal human needs. In 1996, Bartle first proposed a Player Type model based on an online debate regarding the question, “What do people want out of a multi-user dungeon (MUD)?” Answers from experienced players of MUDs were summarized, and the players were categorized into four main types: Achievers, Explorers, Socializers and Killers. The main motivation for achievers is to accomplish game-related goals, advance to higher levels and gather treasures and points. Explorers are motivated by the quest for knowledge of the internal machinations of the game. They enjoy exploring the out-of-the-way places and have fun discovering interesting features. Socializers are motivated by a desire to maintain inter-player relationships. Killers are fulfilled when they cause massive distress to other players [16]. Based on Bartle’s model, Yee conducted a factor analysis study on the motivation for game play. The importance of 39 items (such as “level up as fast as possible”) were rated by 3200 respondents. A principal component analysis revealed 10 factors, which fit into 3 categories: Achievement, Social and Immersion. Achievement players seek advancement, want to learn mechanics, and enjoy competition. Social players like to socialize with other players, establish in-game relationships and have fun with teamwork. Immersion players enjoy exploring the virtual world, being immersed in the new world, customizing their characters and escaping from the real world [17].

Other theoretical works [15, 18] on the motivations for game playing are mainly rooted in self-determination theory [18, 19]. According to self-determination theory, both intrinsic motivation (i.e., the fundamental needs of human beings: competence, autonomy and relatedness) and extrinsic motivation (e.g., reward, punishment or self-esteem pressure) are important for understanding what people want out of a game. Intrinsic motivation is especially significant. Competence needs refer to the necessity of challenge and feelings of reflectance. Autonomy needs refer to feelings that the decision making will impact results. Relatedness needs refer to the needs for social interaction. Table 2 lists the variables derived from the motivational models.

Table 2. Variables derived from game motivational models
Full size table

2.3 Game Design Heuristics

In the last three decades, several video game design and evaluation heuristics for enhancing the immersion, playability, enjoyment, or generally speaking fun of game play have been published [4,5,6,7,8].

The earliest heuristics concerning video games were proposed by Malone [4], in which three features were observed to make the system enjoyable: Challenge, Fantasy and Curiosity. Several sub-features such as clear goal, uncertain outcome, emotionally appealing fantasies, and metaphors were also mentioned. Clanton [15] developed a list of usability guidelines for computer games based on hundreds of hours of observation, mainly from three perspectives: game interface, game mechanics and game play. Federoff [6] reviewed the existing heuristics and enriched the list of guidelines based on a case study of a game design team. Desurvire et al. [7] introduced a Heuristic Evaluation for Playability (HEP) model which included game story as an evaluation category in the model. Sweetser and Wyeth [8] developed a new evaluation model, GameFlow, which concentrated on player enjoyment in games rather than usability and validated the model with two games. Eight elements were included in the model: Concentration, challenge, skills, control, clear goals, feedback, immersion, and social interaction. The related variables are listed in Table 3.

Table 3. Variables derived from game design heuristics
Full size table

3 Methodology

This research adopts a three-step methodology: variable derivation, questionnaire construction and exploratory factor analysis. The goal of the first step is to derive related variables from both theories and empirical findings, while the next two steps aim to reduce the dimension of the model. The research framework is shown in Figs. 1 and 2.

Fig. 1.
figure 1

Research framework

Full size image
Fig. 2.
figure 2

Procedure for finding variables

Full size image

3.1 Variable Derivation

Originally, 27 variables were derived from a literature review with the following principles: (1) items that do not change with stories were removed (e.g., the “reward” item from Clanton [5] and the “Use visual and audio effects to arouse interest” item from Federoff [6]); (2) similar items were combined into a single item (e.g., “uncertain outcomes”, “no definite way to win”, and “ending” were combined into “Ending” - “The ending of the story can be changed”); and (3) keep as many variables as possible when ambiguity occurs.

To include the opinions of game experts or professionals, we chose to conduct a content analysis based on online discussions on Zhihu, a Chinese community question answering system known for professional answers, over interviews after evaluating the content quantity as well as quality and time efficiency. Altogether 11 questions were found related to our topic and up to 83 replies and 18,538 Chinese characters were collected and analyzed. The conventional approach to content analysis was employed without making pre-assumptions on the category of results. The purpose of this is to allow the categories to flow from the data. Following the process described in Hsieh and Shannon [20], 9 clusters were found and added to the item pool.

Finally, 6 students from the HCI laboratory of Tsinghua University with experience in questionnaire design and more than 10 years of game-playing experience were invited to the pretest. The interview has three purposes: (1) further explore potential factors; (2) check if they can understand the statements of each variable (The statements were carefully phrased to describe the factors. They are composed of original sentences from the literature, which are translated into Chinese, and brief quotes from the interview and answers from zhihu.com; and (3) minimize redundant questions. Nine items were added, of which 5 items were expanded from the original 27 items. The final 45 variables are listed in Table 4.

Table 4. Items list
Full size table

3.2 Construct Questionnaire

The questionnaire contained the following parts: demographic information (age, gender, and education), game preference (type), game experience (maximum game-playing frequency, current game-playing frequency, subjective ratings on expertise), overall rating on the importance of storytelling in game play, and ratings on the importance of each variable with a statement. All of the statements were checked by interviewees to ensure correct understanding and minimal redundancy. For each question requiring a rating of attitude, a 7-point Likert scale was adopted in which 7 represented strongly agree and 1 represented strongly disagree.

3.3 Procedure

The questionnaire was posted on websites in December 2014. The links were spread over social media platforms (WeChat and Renren.com) and a game forum (Baidu Tieba “Steam” - the forum of a large game distribution platform). Approximately four-fifths of the participants were from the game forum. Experienced game players were targeted as participants. To prevent inexperienced game players from participating, the following pre-screening question was added: Do you satisfy the following conditions: Love games; Spend much time playing games; Have played at least one game with a story. Participants who failed to meet the requirements were filtered out. The questionnaire was introduced as an “investigation of game preferences in China”, without informing participants of the actual purpose of the study. We rewarded each participant with a 10-yuan mobile phone bill credit.

3.4 Participants

Overall, 516 questionnaires were collected. 84% of the participants were male. The ages of over 90% participants ranged from 15 to 3. Additionally, 74% of the participants had a bachelor’s degree or higher. 94% had no working experience in the game industry. On average, the participants had played computer games for 9.48 years (SD = 4.11). The maximum duration of their game-playing career is on average 1.22 h per day (SD = 5.127). However, currently, over 80% of them play games for less than 5 h per day. The top five game types preferred by the participants were shooting games (51.68%), action games (40%), role-playing games (39.6%), strategy games (29.31%) and adventure games (2.99%). Finally, 72% of the participants claimed that they played many types of games, and 56% of the participants regarded themselves as experts in game playing.

4 Results

Among the 516 answers, 11 unqualified participants (who did not satisfy the pre-screen conditions) and 51 invalid answers (answers with nulls, all-7 answers, 1 vs 7 in highly correlated questions) were eliminated, resulting in 454 valid answers.

An exploratory factor analysis (EFA) was conducted to find the structural characteristics of the questionnaire. The Kaiser-Meyer-Olkin test (KMO = .876) revealed significant correlation between items, which means that the items have enough common information. The EFA enabled the reduction of the items into factors that are comparatively less correlated.

The procedure of EFA is shown in Fig. 3. First, factors with an eigenvalue larger than 1 were extracted. A component matrix was calculated and rotated relative to orthogonal rotation for further interpretation. Quartimax rotation was adopted in this study to get factors with lower correlation. After the calculation, the following conditions were imposed: (1) the communality should be larger than 1 for most variables; (2) the total variance should be larger than .5; (3) The factor loading of each item should be larger than .45; and (4) there should be no items with two factor loadings larger than .45. If these conditions were not satisfied, the items violating the conditions were deleted one by one until all of the conditions were met.

Fig. 3.
figure 3

EFA procedure

Full size image

Following the procedure, 2 items were eliminated: Teamwork and Concentration. The communality of these items was less than .5, indicating that less than half of the variance in measured variables is reproduced by the latent constructs. All of the conditions were met after the elimination. As a result, 11 factors were extracted from 43 items, explaining 63.24% of the total variance. The results of EFA were shown in Tables 5 and 6.

Table 5. Total variance explained
Full size table
Table 6. Rotation matrix
Full size table

5 Discussion

To measure the game-adaptability of stories, this study proposes an evaluation model with 11 factors (as shown in Fig. 4) based on players’ ratings: engage & explore, scriptwriting, distance from reality, autonomy, empathy, values, competition, power, multi-challenge, physical attractiveness, and familiarity. Here is the further explanation of these factors: (Detailed items see Table 7).

Fig. 4.
figure 4

Final model of game-adaptability evaluation

Full size image
Table 7. Important factors for creating, choosing and adapting high game-adaptable stories
Full size table

Engage and explore refer to the attraction on the players that arouses their desire to explore the world in games. Game design is not just about the design of the gameplay and stories. Game design is also about the creation of a lively world, a world that contains many interesting places and stories. Scriptwriting is about how to tell a story in a game, which includes design of a main plot, design of the story structure, depiction of characters and logical correctness. Most of the games are very far from reality, which calls into full play the players’ imagination. Dungeons & Dragons, a famous game category originating from a tabletop role-playing game, is a good example of distance from reality. This type of game creates a world with history and culture which is quite different from the real world. The species in this world are quite different and have their own unique abilities. And the world is mysterious for the player, which arouses their curiosity. Power is one of the most important concepts in most games. Players solve problems and accomplish tasks in the game to improve their abilities. And the more power they acquire, the more tasks they are able to fulfill. This positive feedback will motivate players to move forward in the games. Many games require players to accomplish various challenges and achievements. Attractive characters contribute a great deal to the story telling of a game.

The Witcher 3, the winner of The Game Awards 2015, provides players with great autonomy. It adapted from a famous Polish novel, which promises high quality and high completion of the game story. The plot is well organized, and the characters are well portrayed. Those who have read the story are likely to be familiar with the plot and characters when playing this game. Players can slaughter monsters, collect materials and make their own armor and sword. Additionally, the storyline has various choices. Players are able to change the fate of some other characters by choosing some action or conversations. And due to the different choice, the ending will be different. The main character of The Witcher, Geralt of Rivia, has many friends and enemies. The relationships between characters are complex and touching, which offer profound insights. The presence of many difficult choices in The Witcher 3 arouses players’ empathy. No decision is easy for players because they have to consider morals and interests. Players sometimes have to judge what is right and what is wrong. Sometimes, helping some people in the game means you have to damage interests of other people. Most of these choices are ambiguous, depending on players’ own values, which is a reflection of the real world.

6 Conclusion

In this research, a model for measuring the game-adaptability of stories was established with 11 factors (sorted by significance): engage & explore, scriptwriting, distance from reality, autonomy, empathy, values, competition, power, multi-challenge, physical attractiveness, and familiarity. Among these factors, engage & explore and scriptwriting are the most important factors in a game storytelling. A story with higher potential in these factors is more likely to be welcomed by players when adapted to a game. If a story was extraordinary in some respects, it is also likely to succeed when adapted to a game.

There are three main limitations of this study. Firstly, the study was based on the views of players. And players tend to focus on their own experience of the games. They are not concerned about commercial prospects and technological difficulty, which also play important parts in game development for developers or investors. Secondly, most of the participants of the survey were steam players, which is different from the players who play console games. Thirdly, the survey was conducted in China, and the results of this study might be influenced by culture differences. In addition, the game industry in China is quite different from the game industries in other countries, which might also have some influence on the result. Thus, it will be very interesting to explore the cultural differences in game storytelling in future work.