In prior work, we adopted a co-design approach that gathered a cohort of university students to participate in the brainstorming and design of sleep health technologies targeted at addressing the sleep health needs of this population more accurately [4]. A total of 51 university students participated in 3 co-design workshops. Thematic analysis of the qualitative data generated 9 themes, including health tracking, sleep environment optimization, sleep literacy, community, recommendation, gamification, generative artificial intelligence, materializing sleep with learning, and personalization. For more details on the workshop and resulting initial design insights, refer to the study by Liang et al [4]. Building on these design insights and evidence-based sleep health knowledge, we developed 3 low-fidelity prototypes of smartwatch sleep games that addressed various themes identified as important by the participants of the initial workshop (Figure 1). The game features for each prototype and their designs are described in more detail in the subsequent sections.

Hero’s Sleep Journey (HSJ) focuses on leveraging the powerful motivational effects of role playing and avatars [27-29]. It features a hero avatar as the main visual theme and is centered on adventure and mission completion. Users play as a hero who grows stronger through good sleep hygiene and fights monsters to protect their village. Some of the core game features are summarized in Textbox 1.

Sleep Tamagotchi (ST) leverages the charm of a virtual pet to engage players in improving their sleep hygiene [30,31]. The central theme involves a cute digital pet that guides users in developing and maintaining healthy sleeping habits. Some of the core game features are summarized in Textbox 2.

Sleepland focuses on using the motivational power of construction games, that is, games where players combine or configure objects to organize and build things [32]. Specifically, it features building construction as its main visual theme, allowing players to create and develop their own land or town. Users are tasked with the challenge of restoring a Japanese village using their sleep power, which is accumulated through healthy sleep hygiene activities. Some of the core game features are summarized in Textbox 3.

We distributed flyers and posters around the campus of the Kyoto University of Advanced Science (KUAS). A total of 45 participants (self-identified as female participants: n=7, 16%; and self-identified as male participants: n=38, 84%) attended the workshop. The age of the participants ranged between 18 and 35 years. In total, 9 (20%) participants had prior experience using a smartwatch or wristband for self-tracking. All participants were undergraduate or graduate students enrolled in the Faculty of Engineering at KUAS during the time of the workshop.

Before beginning the main activities of the workshop, we provided an introduction to the workshop goal as well as background information on serious games, wearable sleep-tracking technologies, and sleep hygiene. Participants were given an icebreaker activity to encourage collaboration and discussion among one another. Following the icebreaker, participants were briefed on the main findings from the initial workshop and given background information about the sleep game. They were then presented with storyboard cards depicting features from all 3 prototypes. In total, 2 examples of these storyboard cards are shown in Figures 2 and 3. Each card included 4 questions on the backside: “How useful is this feature for improving sleep?”; “Would you use this feature?”; “Could anything be improved?”; and “Free comments.” The first 2 questions required participants to rate the feature on a Likert scale ranging from 1 to 4, with higher ratings indicating a more positive appraisal of the feature. The last 2 questions are open-ended. Participants formed groups of 4 to 6 people to evaluate each feature of the 3 prototypes and were encouraged to openly share their feedback within their groups. Once all features were evaluated, the cards were collected.

At the conclusion of the workshop, participants completed the 5-factor player traits questionnaire [33], a validated instrument with 25 items, administered in paper format. This questionnaire assessed participants’ preferences for different game elements and playing styles. Using this tool provided a deeper insight into the participants’ inclinations toward each sleep game. Alongside the questionnaire, participants were also surveyed about the amount of time they spent playing games per week as well as their preferences among the 3 games regarding which one they found most useful for promoting sleep health and which game they would most like to play.

This study used a mixed methods approach, resulting in a comprehensive collection of both quantitative and qualitative data from the workshop. Initially, all data were collected on paper and subsequently digitalized into spreadsheets for analysis. For the quantitative data analysis, histograms were created to illustrate participants’ ratings of each game’s overall appraisal and individual features. The qualitative data, consisting of participants’ free-form comments, were analyzed using thematic analysis. In total, 3 authors independently conducted the coding, and discrepancies were resolved through discussion. Instead of using a predefined coding schema, codes were inductive and identified iteratively by repeatedly reviewing the comments. These codes were then grouped into 8 themes, which will be detailed in the Qualitative Findings section.

Our observations suggest that the genre of the game influences participants’ perceptions of its features. Many participants who indicated their preference for HSJ were originally fans of role-playing games (RPGs), drawn to the fantasy environment embodied by the theme and storyline of the HSJ game. For instance, the participants said the following:

Similarly, participants who favored Sleepland were originally fond of building games, while those who chose ST felt connected to animals and pets. They said the following:

Another primary factor influencing participants’ preferences for different sleep game prototypes was their orientation toward competitive or noncompetitive gameplay. The HSJ game distinguished itself as the most interactive game among the 3, featuring level-up mechanics and customization that appealed to competitive players. P34 noted the following:

In addition, a few participants found HSJ useful for improving sleep due to its motivating and engaging nature. In total, 4 participants who voted for HSJ believed that the daily challenges would motivate them to “be more mindful of sleep health” and they would “try to sleep well so that the hero can win.” P13 mentioned that “progress can be made much faster (in HSJ) compared to the other two (games) due to the element of competition.”

Conversely, not all participants identified as competitive players. Some preferred Sleepland and ST precisely for their passive nature. Compared with HSJ, Sleepland was considered less complicated because, as P3 said, “the tasks can be automated, and it takes minimum input from the users.” The pet theme of ST was also perceived as more lighthearted compared with the other 2 games, as P5 noted, “I don’t think I would be a competitive sleeper, which is the requirement for Hero’s Sleep Journey.” Unlike HSJ, which excites players, ST causes less stress as, according to P5, “it seems like a fun app in an easygoing manner. Not too competitive. And feels like a very gently encouraging app.”

Participant feedback highlights the emotional impact of sleep games. Interacting with cute pets in ST seems to create an affection effect that reinforces positive behavior change by fostering a sense of companionship and mirroring behavior. Participants reported feeling connected to virtual pets, with some expressing willingness to “do anything for a cute virtual pet, even sleeping early” or “try to complete any tasks that a virtual pet gives me.” Furthermore, the friendly nudge feature in ST was praised for its companionship effect, potentially enhancing users’ motivation. P5 commented as follows:

Virtual pets may also have a mirroring effect, as stated by P40:

A similar effect was also noted in the HSJ game. One participant associated the healing during sleep with the healing process of the hero character, thus considering an RPG the most suited for sleep health intervention.

Nevertheless, we also observed potential negative emotions that sleep games may elicit, particularly in relation to reward features. While rewards can serve as external motivation for some participants, others expressed concerns that earning health points or completing tasks in HSJ may induce unnecessary stress. In addition, rewards based on sleep quality may inadvertently trigger rumination, since factors like the amount of deep sleep are often beyond an individual’s control.

Surprisingly, the most popular game is not necessarily perceived as the most useful. While HSJ seemed to be the most popular, ST was perceived as the most useful. The primary reason for ST’s perceived utility lies in its features, which are more focused on sleep health and less distractive from the sleep goal.

More than one-third (13/38, 34%) of the participants considered the ST features highly pertinent to sleep health, stating that ST “has the most sleep and helpful driven features out of the three” (P12). They acknowledged that “there are lots of features that could help me improve my sleep health even if the features or tasks might not be interesting” (P22). Sleep report and sleep education were 2 features that participants highly appraised for their potential positive impact on users’ awareness of sleep health. They considered the sleep education contents useful and informative, and the sleep reports would allow users to “look back in the long run” to “analyze sleep patterns” so that “users can be in touch with this for a longer period without getting bored” (P43). These features help users stay focused on improving sleep health without overengaging with features that are not critical for sleep health. As P30 stated, “the other games will make you get distracted and prioritize your game more than your sleep.”

Similarly, due to the perceived direct link to sleep, the sleep reward features in all 3 prototypes, which manifested differently as a building feature in Sleepland, a sleep health points feature in HSJ, and a sleep reward feature in ST, were highly praised by participants. In contrast, the step reward feature in ST, while also a form of reward, garnered less interest. Given the unclear relationship between step count and sleep quality improvement, participants were less inclined to engage with this feature.

We observed that some interactive features, despite being fun and engaging, were considered distracting from the goal of improving sleep health. Participants noted that the diversity of tasks in HSJ is likely to keep users engaged. P1 stated the following:

However, participants also acknowledged that these interactive features are not necessarily tied to improvement in sleep quality, just as one participant mentioned:

Similar complaints were noted about Sleepland by P12, “it (Sleepland) feels more like Clash of Clans than a sleep improving app.”

Participants’ preferences for game features were influenced by their perception of autonomy while engaging with those features. They highly valued the flexibility of tasks in Sleepland and ST, emphasizing the importance of being able to “choose their tasks” and “do them at any time” to avoid interference with their work or class schedule. In contrast, the hero selection feature in HSJ received criticism for requiring frequent user choices. Excessive customization imposed a sense of forced action that compromises autonomy. Similarly, the sleep checklist feature imposes a sense of forced action. Participants stated that they would be “too tired at the end of the day to check the list” (P12) and “people have different strategies in winding down for sleep” (P22).

The daily challenge features, including the friendly nudge feature in ST, a daily challenge feature in HSJ, and the Guardian’s task feature in Sleepland, may also compromise autonomy. For example, participants found that the 10,000-step goal was too laborious and challenging to achieve, preferring apps that do not dictate their actions. P45 preferred to do things by herself, and she did not “need the app to tell me what to do.” In addition, concerns were raised regarding the potential frequent notifications resulting from the daily challenge features. Participants expressed worries that these notifications could be disruptive and cause annoyance rather than motivation.

Participants offered valuable feedback for improving the sleep games, as summarized in Textbox 4. The suggestions largely fall into 2 categories: enhancing the game experience and aligning the game with sleep health goals and outcomes. Suggestions for improvement included customization options, diverse reward structures, and integration of community features. In addition, participants emphasized the importance of aligning game features with sleep health goals, such as discouraging oversleeping and providing personalized sleep recommendations.

Participants highlighted the need to maintain personal space and avoid excessive notifications, which can become intrusive and bothersome. Allowing users to customize their notification preferences, including whether and when to receive notifications, was considered essential. Customization in other areas of the game was also valued. Along the same line, daily challenges should be tailored to match a user’s capability and daily routine. As P1 stated, “a nice middle ground must be found regarding the difficulty of the daily challenges so as to not demotivate the player.” Diversifying game resources, such as introducing various hero and pet types and various forms of in-game currency such as gold and mana, was suggested to enhance the gaming experience. Participants also proposed the idea of assigning different heroes with unique skills that tackle a set of “daily challenges tailored to specific aspects of sleep hygiene” (P27). For example, 1 hero could “focus on exercise for better sleep,” while another could “focus on sleep preparation” (P27).

Participants highly valued the sleep reward features in all 3 games but suggested that rewards should not follow a linear structure. They proposed awarding bonus points for maintaining streaks and offering higher rewards to those with a low sleep baseline as strategies to encourage consistent, healthy habits and recognize users’ progress. In addition, some participants expressed interest in incorporating penalties for not adhering to good sleep schedules or neglecting daily challenges. They also appreciated the concept of progressive difficulty in the game, with the challenge levels increasing as users advanced. This approach may help maintain engagement and interest while fostering a sense of accomplishment over time.

They also advocated for more community and social features in the games, such as fostering collaboration, competition, team play, and enabling users to share or trade items with other players. For instance, 1 participant suggested allowing users to “visit and interact with other users’ villages” in the Sleepland game.

Furthermore, participants valued the use of multisensory interaction. They proposed the integration of pet sounds, ringtones, and vibration patterns as effective ways to boost user engagement and immersion in the ST game.

Participants also offered a variety of suggestions to align game features with sleep health, that is, using more endogenous designs [35,36]. These proposals aim to integrate gaming experiences with healthy sleep habits and encourage users to develop better sleep practices grounded in scientific evidence in sleep research.

While recognizing the value of checking sleep statistics, participants emphasized the importance of providing personalized recommendations based on users’ current sleep patterns and predicted future sleep quality. Several participants emphasized the importance of aligning daily challenges and rewards with the day-night cycle. For the Sleepland game, P35 suggested implementing a timeline for tasks, such as “sleeping only at night to gain points and exercise in the morning to earn points.” P3 recommended that “sleeping during specific hours of the day should allow the hero to heal faster.” To minimize distractions from the app during sleep hours, P10 proposed that “if the village is viewed during designated sleep time” in the Sleepland game, “some repairs should be undone.”

Oversleeping appears to be a common issue among university students. Participants underscored the importance of discouraging oversleeping, even suggesting penalties for doing so. As P35 noted, “sleeping too much (more than 7-8 hours) should result in point deduction.”

By contrast, a few participants expressed concerns about rewarding players based on sleep quality, arguing that rewards should instead focus on sleep habits and discipline, such as maintaining a consistent sleep schedule. They noted that sleep quality can be beyond a person’s control, whereas habits and discipline are more manageable.

Participants’ inclination to engage with sleep games is heavily influenced by their prior game experience and personal preferences, highlighting the importance of creating psychological relevance in the design of serious sleep games. Our findings indicate that participants preferred sleep games that aligned with their favorite game genre. For example, those fond of RPGs favored HSJ, while others attracted to pet-related themes were more inclined toward ST. Participants’ preferences for sleep games were also shaped by their orientation toward competitive or noncompetitive gameplay: competitive players preferred HSJ, and noncompetitive players leaned toward ST and Sleepland.

Several participants emphasized their sense of connection to the game characters as a main reason for their willingness to engage with the game, further highlighting the importance of sleep hygiene games to create psychological relevance. In HSJ, for example, the hero healing feature serves as a metaphor for the restorative effects of sleep, framing the player’s actions as instrumental in restoring the hero’s vitality. Similarly, users care for virtual pets in the ST game, prompting them to reflect on their own schedules as they maintain a regular schedule for the virtual pet. This creates a digital twin effect [4], aligning with the strategy of similarity used in other health-interventional games [26]. The sense of psychological relevance reduces psychological barriers and enhances the sense of autonomy for behavior change. In contrast, the guardian’s task feature in Sleepland, which involves repairing damaged villages, did not seem to foster the same association with the healing benefits of sleep for humans, possibly due to a lack of similarity and relevance between repairing a lifeless object and human recovery. The importance of establishing an emotional connection in serious games has been previously highlighted in [45-47]. Our findings contribute further evidence within the context of sleep health. Game design elements that foster psychological relevance to players’ prior game experiences and personal lives, particularly those that elicit feelings of autonomy and relatedness, may be crucial to the initial adoption of sleep game technologies.

Sleep games need not adhere to a one-size-fits-all design. Instead, offering a variety of game genres allows users to choose those that resonate most with their preferences and personal experiences, thereby transforming the external demand for healthy sleep behavior change into a more self-determined, personally endorsed, and volitionally initiated process. Furthermore, creating game characters that users feel connected to will enhance perceptions of self-relevance and promote a sense of relatedness [48-50]. This, in turn, boosts intrinsic motivation to engage with the game and healthy behavior change.

The second dimension of relevance for sleep games revolves around whether the game features logically contribute to improving sleep. As a form of serious game, sleep games use game design elements to influence players’ behavior toward better sleep health. Aligning game design with the serious context poses a well-acknowledged challenge in serious game research [51,52], and sleep games are no exception.

Our findings indicate that perceived relevance to sleep health significantly influences the utility value of game features, which in turn leads players to feel more competent in the act of improving sleep health. Features that directly focus on sleep, such as sleep education and sleep rewards in ST, were highly regarded for their relevance and potential to improve sleep. This relevance was considered a major reason why ST could be more effective than the other 2 games. Conversely, features, such as fighting monsters in HSJ, while enjoyable, were not directly linked to sleep health and were consequently rated as the least popular among all HSJ features. Furthermore, several participants expressed concern that features irrelevant to sleep health might distract users from their sleep goals, potentially harming rather than improving sleep. This concern reflects the tension between the health care context and the entertainment nature of games highlighted in previous studies [53]. Addressing this tension requires careful alignment between the serious context and the game features following established design frameworks [41,53]. Indeed, participants made pertinent suggestions on how to better align the game design with sleep health goals, including aligning daily challenges and rewards with the day-night cycle and discouraging oversleeping.

Interestingly, we observed that the perceived utility of sleep game features does not always match their actual utility. Some features, despite having a direct positive impact on nighttime sleep supported by scientific evidence, were not perceived as relevant by many participants. Features such as the daily challenges in HSJ and friendly nudges in ST encourage users to engage in daytime activities conducive to better sleep. These activities represent meaningful behaviors for improving sleep and empower users to achieve their sleep goals. However, not all participants immediately recognized the relevance of these activities and were hesitant to engage with them. For instance, while some participants understood the connection between daytime physical activity and nighttime sleep, others did not. While goal mechanisms have been proven effective in serious games, particularly in educational settings [54], our study shows that it is equally, if not more important, to communicate the rationale behind these goals to users, particularly regarding their logical relevance to improving sleep.

We also found that perceived relevance alone is not sufficient for users to engage with a feature. Despite its perceived relevance, the sleep education feature in ST was the least popular. Instead of presenting educational content in long text, the sleep education feature could be gamified and presented in smaller chunks through gameplay, as exemplified in [55].

Educational content of sleep knowledge can be embedded within the game’s narrative and tasks; for example, introducing a character that provides sleep tips as part of the storyline or using interactive elements, such as quizzes, minigames, and challenges that reward users with in-game benefits for learning about sleep health [55].

The third dimension of relevance for sleep games involves the situational or circumstantial context of users, akin to the notion of situational relevance discussed in prior research [56]. This notion of “life context” has also become an important issue identified in recent behavior change literature [57]. A key finding of this study is that frequent game interactions during the daytime are not always welcomed, as participants’ willingness to engage with those features varied across the day. While sleep games do not necessarily need to be played at night [14], it is important to maximize the situational relevance of daytime interactions to avoid compromising the sense of autonomy.

All 3 games implemented features that address daytime activities. There is a large body of evidence showing that numerous daytime activities can influence nighttime sleep [4,5], making daytime interactions a natural design target for sleep games. However, many participants expressed concerns about being disrupted by notifications irrelevant to the activities at hand, especially during classes or other busy times. Consequently, features likely to generate excessive notifications, such as friendly nudges and sleep education in ST, were rated as the least popular features among the workshop participants. Several participants mentioned that they do not need the sleep game app to dictate their actions, indicating a compromised sense of autonomy due to potentially excessive notifications.

One way to address the potential interruptions caused by excessive notifications is through boundary management [58,59], which ensures that notifications are pushed at appropriate timing and are relevant to the users’ situational context. This can be achieved through direct approaches, such as user customization, or indirect approaches, such as leveraging contextual information and users’ behavior patterns to determine optimal notification timing [60]. For example, many participants were reluctant to engage in sleep games immediately upon waking, as they preferred to prioritize their daily routines over gaming. This observation aligns with previous research indicating that people’s acceptance of notifications varies throughout the day [61] and that they are less likely to attend to notifications in the morning compared with other times of the day [62].

Ensuring the relevance of sleep game features to users’ situational context addresses their psychological needs for autonomy. Besides customizing notifications, participants expressed a desire for autonomy in completing daily challenges to gain a sense of competence. Users prefer daily challenges to be achievable rather than feeling frustrated about not being able to accomplish them. For example, instead of setting a default goal of 10,000 steps, step goals should be customizable because users may have busy schedules, prefer biking over walking, or prefer indoor exercise over outdoor walking on some days. On the basis of these findings, we propose the third design recommendation mentioned subsequently.

Allow users to choose when and how often they receive notifications. Make it possible for users to opt to receive notifications only in the morning, evening, or at specific times that suit their daily routines. Contextual information from other apps, such as calendar, GPS, and sleep profile, can be used to customize notification settings automatically, avoiding sending notifications during busy periods, meetings, or events. Providing a variety of choices and allowing users to configure a daily challenge list according to their daily schedule is another way to enhance the situational relevance of sleep game features.