HETZI - Jump and Run: Development and Evaluation of a Gesture Controlled Game Maximilian Schmidt Cornelius Pöpel University of Applied Sciences Ansbach University of Applied Sciences Ansbach Ansbach, 91522 Ansbach, 91522 maximilian.schmidt@hs-ansbach.de cornelius.poepel@hs-ansbach.de Abstract guaranteeing that all other features of the application kept working. In this paper an interactive game is described, The game was tested and evaluated against four that has been developed for a company using hypotheses in order to be able to figure out weaknesses new technologies for raising awareness at job and strengths of the game and to be able to improve it fairs. An already existing framework of a with respect to the use at job fairs. The hypotheses were game terminal for job fairs is used. The defined in a way allowing to figure out whether the corporate identity, the content of the game will contribute to the company’s overall advertised job, but also the attraction of a appearance at fairs. Subsequently final adjustments to young and technologically interested audience further improve the game were made. is combined in a game in order to develop an Especially in the context of a job fair it is important interactive experience that is easy to grasp, for a company to be remembered by the visitors. Past raises awareness for the company and is fun to experience shows that usually exhibitors do not rely on play. Within the gesture controlled Jump and Jump and Run games for marketing purposes. Run game, the audience accompanies the Therefore it should serve the purpose of becoming a character “Hetzi” through the guts of a more recognizable brand and helps to make the colocation centre, avoiding hazards and appearance more memorable. Thus hypothesis one animating the player to move in order to win, focuses the question whether the player will associate the company’s vision of a modern the game with the company in case the game occurs communication strategy became true. After within the company’s working environment and the describing the idea and implementation an player becomes aware of that. evaluation of the game is presented. Due to the Hypothesis two: If the sound matches the events on results and implemented improvements the the screen, the sound might contribute to the overall company is convinced and the game will be gaming experience and enjoyment. used in future job fairs. Hypothesis three: If the controls of the game are simple and intuitive or at least easy to describe and 1 Introduction comprehend, the player is able to play the game instantly and without time consuming explanations. At job fairs companies generally try to catch the This would contribute to making a fast impression on attention of the visitor and want to be remembered by visitors, who usually only spend little time at each potential future apprentices. In order to achieve this single stand of an exhibition. goal for a company providing hosting solutions and Hypothesis four: If the complexity of the game is being an experienced data centre operator a gesture increasing in correlation with the amount of time the controlled Jump and Run game was developed. It was player is investing, it equally suits both new and more implemented on an already existing terminal designed experienced players. by Fliehr [Fli15], which consists of a Windows Computer running Max 7, a Full HD display in portrait orientation and some additional hardware. The Jump 2 Related Work and Run game replaced a version of Tetris in the Fliehr describes how the single components of the original terminal application and had to be integrated underlying terminal are arranged and function. seamlessly into an existing Max 7 environment, Moreover, he suggests improving the integrated camera Copyright © by the paper’s authors. Copying permitted only for private based tracking system and a further enhancement of the and academic purposes. brick-based game [Fli15]. Since the version of Tetris In: W. Aigner, G. Schmiedl, K. Blumenstein, M. Zeppelzauer (eds.): and the camera based system were removed, there was Proceedings of the 9th Forum Media Technology 2016, St. Pölten, no need to consider these improvements. Austria, 24-11-2016, published at http://ceur-ws.org 91 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game Lubitz and Krause [Lub12] proved that even if input Followed by an initial instruction of how to play the methods for controlling a Jump and Run game game by the supervisor the participants were asked to character are considered easy by a very small group of play and explore the game by themselves for a period persons, a large group of players may consider the of at most ten minutes. If a user had any questions, controls too demanding. This emphasizes the during gameplay, they were answered and used to start importance of usability testing and the necessity of a a short conversation in order to gain further very simple interaction. information. Trepte et al. [Tre11] examined the relation of player Once the participants finished playing, they performance, game-related self-efficacy experience and anonymously filled in the questionnaire. the enjoyment of the gaming within an equation. Their study reveals that player performance has a significant 5 Development influence on game enjoyment. This relation has to be balanced accordingly in order to create an enjoyable The following specifications were outlined at the experience for both beginners and advanced players. beginning of the development process: - The game consists of one endless level, 3 Method dynamically and randomly put together. - The speed of the game and of the character The Jump and Run game was developed in close increases continuously throughout the game to collaboration with the company. Each step of the increase the challenge (which is important to development process was shown to the company in maintain motivation [Har10]). order to guarantee the game is tailored according to the - Different kinds of obstacles appear randomly. company’s vision. After the game requirements had - The game ends as soon as the character been specified, the game design, sound and controls collides with an obstacle and can be restarted were developed. Finally the game was implemented on multiple times. the actual terminal and evaluated. - The character is controllable via a simple hand The evaluation practice consisted of a two step gesture recognized by a Leap Motion process. During and after the usability test, the controller. participant took part in a narrative interview. - An 8-bit style melody is playing when the Afterwards a questionnaire was filled in by the game has started. participant, ensuring the systematic survey of the - The game is running in an existing Max 7 overall impression of the game. . environment of the terminal. The questionnaire was comprised of 19 questions, by which the participants were asked to either rate on a The storyline of the game follows an employee who has scale from one to five, fill in text or tick a checkbox. to fix a broken server at the other end of a colocation Further insights were gained by the live feedback of centre. On his way through the colocation centre he the participants while playing the game and the needs to avoid colliding with obstacles in his path. statements they made without prior request. Ten persons participated in the survey. Most of them were employees of the company the game was designed 5.1 Design for. All other participants were been given a short outline of the company. The overall design is kept very basic in a “flat” design in order to match with the existing design of the other terminal application and to create a modern, 4 Implementation suitable and “clean” look. The game is kept two- The usability test was performed with each dimensional, which lowers the barrier for people who participant individually, during which the participants only play occasionally as it creates a feeling of control. were standing roughly one meter in distance from a [Gan16] screen in front of a Leap Motion controller installed at hip height. After a short introduction to the terminal and how the test had to be performed, the start-screen of the game was presented to the participant. 92 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game After designing all the single visual components of the game, a sprite sheet was generated 1. This helped to further reduce the needed resources, improve performances and allowed for quick and small adjustments of the design. 5.1.1 Game character The game character (“Hetzi”) shows a somewhat stereotypical employee of the company. “Hetzi” is Figure 2: Obstacles wearing “nerdy” glasses, has an inconspicuous haircut, is wearing a black shirt with the company’s logo, which The first types are servers that are modelled in two is represented by the red “H” and carries a toolbox with different sizes. Both servers have the same width but the utensils he needs in order to repair a broken server. differ in height. The servers are standing on the floor “Hetzi’s” appearance wilfully plays with the and the character has to avoid a collision with them by widespread notion of the portrayal of an IT-technician jumping. (although many might disagree with this The second types of obstacles are cables, which (in representation). spite of the usual safety regulations) are hanging from the ceiling and creating a danger for “Hetzi”, if he jumps into them. The third type represents a hacker attack pictured symbolically as a virus. The virus obstacle spawns randomly at one of three predefined heights making it more difficult to avoid a collision with them and creates more diversion. 5.1.3 Setting Figure 1: Game character - Hetzi The scene of the game shows a colocation centre which is similar to one of the company’s (most To make him look a little more innovative and probably without the hazardous surrounding). One of futuristic he is surfing on a hoverboard instead of just the special features set in the background is a diagonal ceiling supporting an energy efficient air circulation running through the colocation centre. which is why this feature is represented in the design of the setting. Furthermore there are servers, not too 5.1.2 Obstacles similar to the obstacles, standing in the background in There are three different kinds of obstacles inspired by addition with the company’s logo on the wall. real objects, which can be found in a data centre Figure 3: Setting 1 Sprite sheet: An image containing all images used in the game 93 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game Obstacles are either standing on the grey floor with either the right or left hand of the player within the (servers), hanging from the ceiling (cables) or flying interaction area of the Leap Motion controller. somewhere in between (viruses). If the Leap Motion recognizes a swipe-up the y-axis The character has a fixed x-Position (horizontal) and variable increases. If a certain threshold is reached the a controllable y-Position (vertical) to allow him to jump starts. The threshold guarantees that the jump jump. If the character is not jumping he is hovering in a isn’t triggered by accident, allowing the player to sinusoidal movement above the floor. slightly move their hand without affecting the game. Since the learning curve appears to be very gentle, the game offers entertainment for all possible user groups. 5.4 Software development The software development process started with the analysis and the adjustment of the pre-existing terminal application. The new game logic was then implemented Figure 4: Character and obstacles in the setting in JavaScript and integrated in the Max 7 environment, followed by the integration of the Leap Motion 5.1.4 Score Controller into Max 7. The current score and the high score are displayed close to the top-right corner of the game’s canvas. The 5.4.1 Initial situation high score is indicated by “HI” and the current score updates in real-time while playing. The initial application consisted of: - The start screen, - A menu to choose one of two games, - The Tetris-clone, - Numberquiz (a game where you have to convert decimal into binary numbers and plug the number with network-cables in a switch with the upper row representing ones and the Figure 5: Score and high score lower row zeros as fast as possible), - A photograph of the player, taken by an 5.2 Sound attached webcam and printed together with the Three events result in a sound aiming to create a time from the Numberquiz more interactive gaming experience: The navigation of the application was achieved by - The start of a jump of the character. holding physical Tetris-bricks with a fiducial mark into - Reaching a multiple of 100 points. the field of view of the installed camera. The camera - A collision of the character and an obstacle. then tracked the fiducial mark and translated the Moreover a background soundtrack starts every time information into the desired output. the game begins. The soundtrack is an 8-bit sound file, which is supposed to be reminiscent of the sound 5.4.2 Programming Environment Max 7 design of older 8-bit-platform games such as Super Mario Bros. ®. The Jump and Run game was added to the pre-existing software whilst all other features had to remain 5.3 Controls working. The navigation links were adjusted accordingly. In order to keep the control as easy as possible only To navigate through the application the Leap Motion one action can be triggered by the player: The jump. was introduced. As there is a maximum of two This jump of the character is triggered by a swipe-up navigational options on each screen one option is triggered by clenching a fist with the left hand for a 94 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game short amount of time while the other option is triggered receives a new x-position which is the second by holding a fist with the right hand. background’s x-position plus its width. The layout of all the screens has been adjusted to The current score is constantly increasing. If the take advantage of the full resolution of the terminal’s current score modulo 100 is zero the achievement screen (1080*1920px). sound is played. If the game state is set to game over and the current score is higher than the high score, the 5.4.3 JavaScript in Max 7 high score gets updated. If the game state is set to game over, the game For a Jump and Run game the “JSUI-object” fits frames and the update cycles stop. If the game is then best which is the equivalent within the Max framework restarted everything besides the high score is set to the to the Canvas-element in HTML (Cyc16). The initial values and the game begins again. JavaScript code updates the position of the background, If you are using “this.” in JavaScript as you usually the character and the obstacles every frame. Then all would, problems can occur as Max sometimes falsely the elements on the canvas are cleared and redrawn seems to relate “this.” to the JSUI-object in your patch. with their new positions. This can lead to errors in the code, shown by the The character has a fixed x-Position creating the console or to unexpected behaviours of the game. illusion of a forward movement by decreasing the x- Therefore you should avoid using “this.” or you would Position of all other elements. The hovering of the have to double-check each occurrence of “this.” to character is created by updating its y-Position. It make sure it is working the way you need it to. increases until it reaches an upper threshold and then decreases until it reaches the lower threshold. When a 5.4.4 Leap Motion integration in Max 7 jump is triggered the y-Position is updated in regard to the existing thresholds. Using the “leapmotion-object” by Jules Francoise Initially six random obstacles are created. If an [Fran14] the integration of the Leap Motion Controller obstacle’s x-Position is lower than zero minus the is straightforward. The object extracts all the tracking obstacle’s width, it is removed and a new obstacle is data from the Leap Motion and utilizes it in the relevant created. The x-Position becomes then the x-Position of Max-Patch. the prior obstacle plus its width and an additional To navigate through the application the grab strength random gap. The virus obstacle can only be created if of both hands is evaluated. If it is found to be higher the speed, which increases throughout the game, has than a threshold, a variable is increased; if it is lower exceeded a certain limit, since it is the most difficult the value of the variable is set to zero. The navigational obstacle to avoid a collision with. option is triggered if the variable reaches a threshold. The collision detection validates if: Additionally, the confidence (a probability value of the - the character’s x-position is smaller than the placement of an object) of each hand must be obstacle’s x-position plus its width evaluated, since the “leapmotion-object” keeps the last and known value of the grab strength stored and does not - the character’s x-position plus its width is reset it when the tracked object is moved out of the greater than the obstacle’s x-position interaction area. While the user moves his hand to the and edges of the interaction area, the confidence decreases - the character’s y-position is smaller than the allowing the program to detect the possible leave of the obstacle’s y-position plus its height users hand and therefore is able to set the value of the and relevant variable to zero. - the character’s y-position plus its height is To provide a visual feedback to the user a loading greater than the obstacle’s y-position bar beneath the navigational option continues to fill If all conditions are met, it returns a collision and the while the user is holding a fist. By that way the user is game state is set to game over. able to experience how long to hold the fist and when The background consists of two parts. If the first to open the hand again. part’s x-position is smaller than zero minus its width, it 95 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game 6 Evaluation Table 1: Design The results of the evaluation were split into four 5 parts: 4 - Design. - Gameplay. 3 - Overall. 2 - Improvements. 1 The participants surveyed helped in identifying 0 major bugs as well as receiving an overall first Character company overall Design company Setting impression from the users. The study was not designed suits suits with the goal to generate statistically exact measurement results on an overall user experience but to see tendencies in the user experience with respect to the hypotheses and the aims of the development. Thus the graphs in the following represent the results The overall design was rated 4.1 out of 5 points. according to the answers of the participants and are Some of the participants were not sure if the servers used here as a means of users tendencies in the in the background of the game were obstacles. At first perception of the game. they had trouble distinguishing the obstacle servers Eight out of the ten participants were male. The from the background servers. However, after a short average age throughout the group was 24.4 years. Two period of playing the game this problem disappeared. of them had prior experience with a Leap Motion Nevertheless, the background was adjusted in order to controller. Only one person had never played a Jump address this problem. and Run game before. At the time of the survey, eight of the ten participants were working at the company the 6.2 Gameplay game was designed for. 6.2.1 Sound 6.1 Design The sound was perceived as either supporting or Being asked what the character represented, 90% disturbing. The participants, who perceived the sound recognized the character as a stereotypical employee of as disturbing, agreed on the fact, that the game should the company. 60% identified his means of contain sound, but they were displeased with the transportation as a hoverboard. Even though 40% did chosen song. not, only one participant wondered why the character was hovering over the floor and thought it to be a All participants agreed that the sounds for jumping “carpet-like thing” out of Aladdin’s tales. All and a collision were fitting and enhanced the quality of participants identified the background as a colocation feedback. centre. With all above mentioned characteristics the participants were asked to fill in the blank text. Rating 6.2.2 Controls on a scale from 0 (not at all) to 5 (fully agreed) the On a scale where 3 represented “just right” the swipe participants agreed that the character as well as the up was considered neither as very simple nor as setting suited the company’s conception and image. complicated. Most participants said it was rather intuitive and fast to learn. 96 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game Table 2: Controls Table 3: Obstacles 5 5 4 4 3 3 2 2 1 1 0 0 Too few / Too small Too easy obstacles too much difficult complicated surprising gameplay intuitive / / too big / too simple / gaps We found that it is rewarding to conduct small At first the results were quite surprising, because a adjustments to the individual proportions of the visual swipe up did not seem to be challenging at all. The elements. The ratio between the character’s height, the problem most participants experienced resulted from possible jump height, the obstacle’s width and height as the narrow tolerance range of the Leap Motion well as the period of appearance of the obstacles and controller, which often failed to recognize the the speed of the game have to be treated carefully. movement as a swipe up. Several reasons could be identified. In case the movement was carried out too 6.3 Overall slow, the participants were able to quickly adapt the movement. However, in case the participants hand had The participants of the study expressed that the game suited the company very well. Despite some left the interaction area, the participants were not able weaknesses they rated the overall game 4.2 out of 5 to realise this and therefore could not adjust their points. motions accordingly. For that reason further adjustments had to be made to the game (6.4 Table 4: Overall Improvements). Despite the difficulties, the participants rated the 5 gameplay 3.8 out of 5 points. This result was used to 4 slightly adjust the game-controls in order to strengthen the gameplay. 3 2 6.2.3 Obstacles 1 Regarding the obstacles, the results of the survey 0 the game company showed, that the participants were content with the suits the rating of Overall Game amount of obstacles in the game and the positioning of these obstacles as well as the level of challenge for beginners and experienced players. They were asked to rate on a scale from 0 (too few / small / easy) to 3 (just right) to 5 (too many / big / difficult). Considering the difficulties some of the users experienced, this appeared to be a surprisingly high rating, which indicates a possible potential of the game. 97 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game 6.4 Improvements 8 Prospect After the evaluation, the following improvements were As mentioned in section 6.4. a retest within the made. The difference between foreground and framework of an exhibition or ideally a job fair background was emphasized by slightly desaturating including a bigger number of participants of the target and blurring the background. group is a future project to be undertaken. The threshold for triggering a jump was lowered. According to our experience the most important part Therefore a jump is triggered faster and easier. of making a gesture-controlled Jump and Run game A small green box was included to indicate if the enjoyable is to give the user constant feedback about left, right or both hands are within the interaction area the precision of the applied gesture. This should be of the Leap Motion controller. This should serve as a done through the interface the user sees on the screen permanent feedback for the user and makes it easier to rather than a person who constantly has to correct the avoid mistakes. player’s movement. Not yet implemented but possible improvements: Especially building a physical boundary to indicate - Building a physical boundary for the the interaction area of the Leap Motion controller interaction area of the Leap Motion. would be a great benefit for any gesture controlled application. - Replacing the 8-bit song with a more popular As the Jump and Run game is mostly written in melody. JavaScript it could be adapted to be used on the For further adjustments the game would need to be company’s website or implemented as a smartphone retested by a broader audience at an exhibition where app – which three participants of the usability test also the terminal shall be displayed. suggested. Developing a lightweight casual game with easy controls can be a great benefit for a company. It 7 Conclusion increases the attractiveness of an exhibition stand and can be used as a giveaway in the form of a smartphone Hypothesis one: The participants did associate the app enhancing a company’s impact on future applicants game with the company and responded that it was a and raising memorability. “very good” suit for the company. We believe that this A possible future improvement on the overall is mostly the result of the design of the character, the experience would be the implementation of a persistent obstacles and the environment the game plays in. high score list. Hypothesis two: For the participants of the study Furthermore it would be interesting to implement sound played an important role in making the game different input methods for the jump like a Wii balance enjoyable. Even if the sound does not match with the board or a Kinect camera recognising a jumping user or player’s idea of a suitable music, participants indicated just a simple enter key. that it contributes to the overall gaming experience and enjoyment. Hypothesis three: If the controls of the game are References simple and intuitive, participants were able to instantly [Cyc16] Cycling '74. Cycling '74 Forum. 2016. 28. 07 2016. play the game. Two of the participants came back after . conclude that once an intuitive gesture is performed a [Fli15] Fliehr, Andreas; Müller, Anna; Ehrenfeld, Franziska; few times, it will not be forgotten quickly and the game Pöpel, Cornelius. „Haptische Interaktion beim can be enjoyed more easily. Messeauftritt: Ein audiovisuelles Gameterminal zur Hypothesis four: As the game becomes more spielebasierten Attraktivitätssteigerung einer difficult the longer it runs, newbies as well as more Firmenpräsentation.“ Editor G. Schmiedl. experienced participants enjoyed the game. The only Glückstadt: Verlag Hülsbusch, 2015. difference seemed to be that more experienced [Fran14] Francoise, Jules. Leap Motion skeletal tracking in participants made faster progress in the game and Max. November 2014. Leap Motion skeletal reached a higher speed in fewer attempts. tracking in Max. 28. 07 2016. . 98 HETZI – Jump ’n Run: Development and Evaluation of a Gesture Controlled Game [Gan16] Gandolfi, Enrico. „The two dimensions as a metaphor of control in gaming landscapes.“ 2013. Game - the Italian journal of game studies. 28. 07 2016. . [Har10] Harrigan, Kevin A.Dixon, Michael J.; Fugelsang, Jonathan A. „Addictive Gameplay: What Casual Game Designers CanLearn from Slot Machine Research.“ Vancouver Digital Week. Vancouver: ACM, 2010. [Lub12] Lubitz, Kolja und Markus Krause. „Exploring User Input Metaphors for Jump and Run Games on Mobile Devices.“ Springer Berlin Heidelberg, 2012. 473-475. [Tre11] Trepte, Sabine; Reinecke, Leonard. „The Pleasures of Success: Game-Related Efficacy.“ CYBERPSYCHOLOGY, BEHAVIOR, AND SOCIAL NETWORKING. Edition. 14. 9. 2011. 99