Mahjong Club: Enjoyable Mahjong Experience for Various Players

MASTER S THESIS Mahjong Club: Enjoyable Mahjong Experience for Various Players Author: Zhanxi Shi Advisor: Dr. Seth Cooper A thesis submitted in fulfillment of the requirements for the degree of Master of Science in the Game Science and Design College of Arts, Media and Design Northeastern University Boston, Massachusetts March, 2017 1

Mahjong Club: Enjoyable Mahjong Experience for Various Players by Zhanxi Shi Mahjong has been released for years, and lots of localized rules have been designed to make it enjoyable in different ways. Based on a set of competitive mahjong rules, this project focuses on minor adjustments to create an enjoyable experience for different groups of players. Iterated tutorial mode and balanced ability system contribute majorly to this project. Submitted in partial fulfillment of the requirements for the degree of Master of Science in Game Science and Design in the Graduate School of the College of Arts, Media and Design of Northeastern University April, 2017 2

TABLE OF CONTENTS Abstract 2 Introduction 4 Background 4 Playtest 6 Iterative design 7 Conclusion 14 Reference 15 3

1. Introduction Mahjong is a traditional tile-based game with a history of hundreds of years. Similar to chess, Mahjong is built up with many strict rules. Different from chess, the players need not only strategies but also luck to win the game. Due to the disturbance of randomness, players need a lot of practice before they know how to improve their win rate. Even though mathematical calculations can improve players performance, but the complex arithmetical procedure is not enjoyable to most of the intermediate players. Hardcore players may spend hours practicing and searching for the optimal solution, but the innate randomness of mahjong could suddenly turn over the situations and confuse the player. Traditionally, mahjong is an analog game. Previous players set up different mahjong rules to meet different kinds of playstyles. Comparing to the competitive versions of mahjong, these local rules usually sacrifice randomness for skills or the other way around. In our project, a digital version of competitive mahjong is implemented. Featured designs are elaborated to help different groups of players enjoy mahjong. Limited by the scope of the project, this is a single-player version, and the other 3 opponents would be controlled by A.I instead of human players. 2. Background It is easy to find electronic copies of mahjong games, but there are huge differences between them. Many players recognize mahjong solitaire games, which are not what we are designing. Based on the mechanics of the traditional mahjong, there are many variations across the world. Mahjong is played with different rules in different areas. In China, it is played differently than in America, in Japan different than in Hong Kong, in Great Britain different than in France, in one region different than in another(jelte, 2006, p. 11). In this project, a complete competitive rule system of mahjong is preferred to keep the game balanced initially. There are two versions of competitive mahjong rules: Official International Rules and Riichi Competition Rules. This project is based on Riichi Competition Rules, also known as Japanese mahjong rules, because the players are required to remember less to start playing Riichi Competition Rules. Limited the scope to only digital mahjong games with Japanese mahjong rules, there are still many forerunners. However, most of the Japanese mahjong games are developed for mahjong experts. Simple graphics and no introductions are common characteristics. Tenhou.net is a popular online Japanese mahjong platform. Thousands of players from beginners to masters are active all day. Due to its simplicity, many mahjong artificial intelligence(ai) is developed and tested on this platform against human players. Tenhou.net is an excellent choice for pure AI development, but it is only appealing to hardcore players that already know how to play mahjong. There are also mahjong games developed specially for beginners. スーパーリアル麻雀 is a 4

series of Japanese mahjong games with more than 20 years of improvements. These series design attractive stories and characters and simplified mahjong from 4 player game to 2 player game. It is an acceptable introductive game collection for beginners, but definitely not enjoyable for mahjong experts. Both of these examples focus on specific player community. There are also Japanese mahjong games developed for all player types. Video arcade games are usually installed in public areas, so all the people nearby are intended players. Different from mobile games and PC games, arcade game players do not have much time for exploring. According to the research on arcade gamers preferences, players prefer familiar games with simple, easy-to-understand rules and play(david, 1990, p. 11). However, it appears that video-arcade-gamers apply a very similar aesthetics to the computer and non-computer games(david, 1990, p. 10). Consequently, a successful arcade game should be easy-to-understand and challenging, but not impossible, to win. In 2011, Sega released a mahjong arcade called Sega MJ5. With years of improvements, Sega MJ5 has successfully meet players expectation of mahjong video arcade game. Successful mahjong games for all player communities not only appears on video arcade machines but also on consoles. Mahjong Dream Club is a console game published on PlayStation3 by D3 Publisher in 2012. With systematic tutorials for beginners and collectible power-up abilities, this game is attractive from the aspect of mechanics. Reviewing the game from the aspect of aesthetics, we notice that Mahjong Dream Club offers flexible camera movements and charming environments and visual special effects. These features improve visual performances vitally and make it more like a real mahjong game. Even though personal computers are built with hardware with similar or even better performance as consoles, we can rarely find mahjong games that include both elaborated mechanics and astonishing visual effects and experiences. In this case, PC platform is chosen as the intended platform for creating the enjoyable mahjong game for both beginners and masters. 3. Playtest Since this project is designed for players with different levels of skills, the playtests are held with 3 groups of people. Group 1 includes 3 people with no experience in mahjong. Group 2 includes 7 people with experience in Japanese Mahjong. Group 3 includes a pool of players that are familiar with other mahjong rules besides Japanese Mahjong. Playtesters in group 1 mainly focus on the effectiveness of the tutorial mode. All the playtests are held with 1 to 1 interviews. The playtester is given a learning task before each test starts. Feedback is collected after the playtester successfully completes the task or gives up when confused. Afterward, the project is improved to a newer version and another test would be held. 5

Playtesters in group 2 mainly focus on the general game experiences. All the playtests are held online. Everytime the project is updated, a working build is uploaded to the drive and an announcement is sent to the playtesters. Usually, no specific tasks are given with the announcement. Since these playtesters have played at least one other digital Japanese Mahjong games, functional exceptions, and uneasy designs can be spotted quickly. With the report from this group, the general playability of the game is improved. Playtesters in group 3 mainly focus on comparative designs between this project and other mahjong games. No playtest is held directly within this group. Questionnaires are sent to this group to collect feedbacks from general mahjong players. Answers from this group ensure that the designs of our project do not violate the consensus of major mahjong players. 4. Iterative design Design and development of this project can be separated into several sections. A set of tutorials is designed for beginners. Ability system is designed for intermediate players. Extra helper UI is designed for smoother gameplay experiences. Setting and statistics are designed for custom adjustments and reflections. 4.1 Tutorial Mahjong is built up with many minor rules. To become a master, a player needs to remember all of them. However, to start playing mahjong, a player only needs to remember a small part of them. The goal of this section is figuring out the minimum collection of rules needed for beginners and the approach of teaching these rules. Design round 1: The first version of the tutorial has 3 levels: beginner, intermediate, advanced(see Figure 1). Beginner level includes 6 parts, referring to general mahjong rules. Intermediate level includes 12 parts, referring to unique Japanese mahjong rules. Advanced level includes 5 parts, referring to tricky optional rules. Each part is directed to a short video with subtitles(see Figure 2). Finishing the parts in beginner and intermediate level should be enough to play the game. It takes about 20 minutes to watch all these videos, which is quite a long time. In order to keep the player motivated, a bonus system is added. Finishing one part gives the player 5 points. Collecting certain points would unlock extra game contents like a character. 6

Figure 1: the first version of tutorial mode Figure 2: tutorial video with subtitle Playtest round 1: The collection of rules selected in beginner and intermediate level is enough for players to start playing the game, but the playtesters want more practice before start playing. The videos are taking so much time that even patient and curious player may not finish them. The bonus system is totally useless and uninteresting to players. Design round 2: All the videos are compressed together as one, and only the most important moments are kept(see Figure 3). The total length is much shorter(20 min to 5 min). The bonus system is removed. The beginner level now only includes the tutorial video. The intermediate includes a shuffle practice(see Figure 4). Shuffle practice is a basic training for scientific mahjong players. Some teaching software has implemented this mode(see Figure 5), but mahjong games rarely include this. The advanced includes a complete introduction to all yakus. The 10 commonest yaku are highlighted so the player can deal with most of the situations without remembering all 30 yakus at first. Figure 3: the second version of tutorial mode Figure 4: shuffle practice 7

Figure 5: Shuffle Practice on Maru-Jan Training Software Playtest round 2: The shuffle practice is very helpful in learning, which matches the conclusion from Cheung that combining games and learning activities can increase the learner s motivation(cheung, 2014, p. 153). The shorter video is more pleasant, but the terms in the video are confusing the players a lot. The chosen 10 yaku are not equally easy to be understood and memorized. Design round 3: Now the tutorial has only 2 categories: basic and advanced. The basic category includes the minimal information for a beginner to start playing Japanese mahjong(see Figure 6). The video is simplified into texts on three pages. Shuffle practice and yaku introduction pages are kept. Instead of recommending the 10 commonest yaku, the 10 easiest yaku are chosen for less learning cost. The advanced category includes only texts, and it is prepared for curious players that want to learn more after playing a certain number of games(see Figure 7). According to the result of the usability study by Mestre, web page screenshot tutorials have a better effect than screencasting tutorials(mestre, 2012, p. 3). As a result, the general format is as simplified as a collection of screenshots rather than hours of interactive video tutorials. 8

Figure 6: basic category. Figure 7: advanced category Playtest round 3: Instructions and information given are clear enough for the player to complete it. Playtesters can start playing a quick game without serious problems. As a result, this version is kept as a final version. 4.2 Ability System Player gets bored quickly if the same content of the game appears repeatedly. Players get bored more quickly if they lose control of the relationship between their actions and outcomes. Ability system can set up a primal feeling of control so that players feel less confused. Proofing the effect of ability system, we need to review the history of mahjong. Mahjong has a history over 200 years. However, mahjong has only been accepted as a member of International Mind Sports Association(IMSA) for 1 year. Consequently, mahjong is recognized as a gambling game. In the 1980s, gambling games were not allowed in public places in China Mainland. Similar to Las Vegas, Macau and Hong Kong ended up as casino clusters. In the same time, Hong Kong movie industries produced many tv series and films about mahjong. In most of these videos, the lead actor uses supernatural power to beat the system and win the day(desmond, 2005, p. 70). These gambling movies became famous quickly in China and around the world. According to the research by Chi Chuen Chan and Keis Ohtsuka, mahjong players in Chinese community are often susceptible to many irrational or superstitious beliefs(chan, Kes, 2010, p. 20). Multiple researchers confirm that these movies enhance Chinese mahjong players illusion of control. Nonetheless, no consensus has been achieved on the existence of strong illusion of control from mahjong players in other countries. In this case, we suppose mahjong players in different regions have a similar physiological preference of controlling the game. Under the condition that all the players want to learn the supernatural skills and beat the others, the special ability system is designed to fulfill their dream of control(desmond, 2005, p. 71). Comparing to the traditional mahjong rules, each player is given a character-related special ability that affects the randomness of the game in different aspects. No dominant or dominated abilities under the same level of strategy and reasoning. These abilities tend to focus on one 9

simple playstyle, usually based on one yaku. With the help of the ability system, the feedback of players action is simpler to be observed. Using different characters would teach a player the advantages and disadvantages of various playstyles and help player finally form a unique mixed version of a personal playstyle. A type of playstyle can be defined combining personal decision and environmental randomness(sato, 2017, p. 8). In this project, the abilities are all designed for more directed personal decisions in order to teach the player different playstyles. In this project, 15 characters with different abilities are implemented. Abilities are designed with 3 perspectives: focus, control, and skill. Focus abilities strengthen the player in one path to victory. These abilities are better at attack than defense and relatively easy to play with. Control abilities help the player balance the situation by controlling the flow of luck for each player. These abilities are better at defense than attack and the difficulty is about average. Skill abilities give the player extra information that can be utilized effectively by skillful mahjong players. These abilities are balanced on attack and defense, but the difficulty is relatively high. There are 6 focus abilities, 4 control abilities, and 5 skill abilities. In order to distinguish different abilities, each ability is assigned to one character with different photos and background stories. Players do not have to remember the bindings between the abilities and the characters in game since they are shown directly in game on the scoreboard. Characters are designed for a more complete persona construction, which has better performance in gameplay metrics and generates better game experience in long term(canossa, 2009, p. 4). From the feedback of playtest group 2, the ability system successfully simplified the reasoning phase and helps the player to make interesting decisions. However, balancing different abilities is a serious concern from playtesters. To keep the characters balanced, a matchup chart is created(see Table 2). Matching any 2 players and play the solo game for 10 times, the performance is marked on the chart. A mark of 0 means always losing, 5 means fair game, and 10 means always winning. The finished chart shows that the best character has an average score of +6. The worst character has an average score of -7.5. Comparing to Japanese Super Turbo Street Fighter chart with the extremes from -22.5 to +27 with 16 characters(see Table 1), the characters ability system in this project is approximately balanced. According to the research on AI scaling, a game with challenging AI is more entertaining than a game that is too easy or too hard(pieter, 2004, p. 2). Given the matchup chart, I confirm that each character has a chance of being a challenging against certain characters. Even though these characters may have minor difference in their winning rate, they all contributes to a challenging and indulging mahjong game because the player would not know who he/she would be playing against. 10

Table 1: Japanese Super Turbo Street Fighter chart, Arcadia 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Sum 1 6 5 8 7 5.5 4.5 3 3 6 5 4 5 3 4.5-0.5 2 4 4.5 4 4 2 5.5 7 5 7 3 5 6 8 6 1 3 5 5.5 5 5.5 3 4 5.5 4.5 6 6 4 4.5 6 2-3.5 4 2 6 5 6 7 5 5 5 4.5 6 6.5 5 5 4 2 5 3 6 4.5 4 8 5 6 4.5 5.5 4 3 4.5 5 5.5-1.5 6 4.5 8 7 3 2 5.5 4 7 6 7 8 4 4.5 5.5 6 7 5.5 4.5 6 5 5 4.5 5.5 6 5.5 6 4.5 6 4.5 6 4.5 8 7 3 4.5 5 4 6 4.5 4 4 5.5 3 5 4 3.5-7 9 7 5 5.5 5 5.5 3 4 6 6 5 4 5.5 7 6.5 5 10 4 3 4 5.5 4.5 4 4.5 6 4 4 5 5 4.5 5.5-6.5 11 5 7 4 4 6 3 4 4.5 5 6 5.5 8 5 6 3 12 6 5 6 3.5 7 2 5.5 7 6 5 4.5 7 8 3 5.5 13 5 4 5.5 5 5.5 6 4 5 4.5 5 2 3 3 5-7.5 14 7 2 4 5 5 5.5 5.5 6 3 5.5 5 2 7 6-1.5 15 5.5 4 8 6 4.5 4.5 4 6.5 3.5 4.5 4 7 5 4 1 Table 2: matchup chart of the current project, all 15 characters included 11

4.3 Helper UI Both tutorials and ability system are directing the players actively. Considering these may not be enough to help all types of players to enjoy the game, a survey is delivered to the playtesters in group 2 and group 3(see Table 3). The goal of the survey is finding out any difficulties encountered when the experienced mahjong players were learning mahjong. Base on the survey results, helper UI is designed for the score calculation and situational observation problems(see Figure 8). Difficulty met while learning mahjong Count Solution Yaku: too hard to remember Score: calculation too hard Situation: not realizing the given information 5 Tutorial: Yaku 14 Helper UI 5 Helper UI Progress: not realizing how close is to victory Method: not know how to win basically Interaction: not know how to defend 3 7 7 Tutorial: Shuffle Practice Tutorial: Shuffle Practice Skill related, not covered Others: luck* Table 3: survey result for helper UI design 8 Fun related, not covered 12

Figure 8: Helper UI in-game screenshot 4.4 Statistics and Settings The statistics page is implemented for a detailed reflection. Setting page is implemented for personal adjustment. From the feedback of the tutorial playtests, certain players have the problem recognizing the tile patterns. Consequently, the switch for different tile pattern is offered. Two language packs are offered as well for localization issue. Understanding that these settings do not necessary change the functionality of the game, I offer the freedom for customization because different styles of learning work better for different people(james, 2005, p. 7). 5. Conclusion As we acknowledged, mahjong is a complicated game with so many minor rules. The complexity may prevent beginners from learning it, but the depth of the strategy makes mahjong enjoyable for master players. In this project, tutorials are designed to teach the player with the minimum effort. Ability system is designed to show the charming strategies more directly. Helper UI and other scenes are designed to improve general experiences throughout the game. From the feedback of the playtesters, combining these features successfully generate a mahjong game that can be enjoyed by both beginners and master players. 13

Reference Rep, J. (2006). The Great Mah Jong Book: History, Lore, and Play. Tuttle Publishing. Myers, D. (1990). A Q-study of game player aesthetics. Simulation & Gaming, 21(4), 375-396. Mestre, L. S. (2012). Student preference for tutorial design: A usability study. Reference Services Review, 40(2), 258-276. Lam, D. (2005). Slot or table? A Chinese perspective. UNLV Gaming Research & Review Journal, 9(2), 69. Ohtsuka, K., & Chan, C. C. (2010). Donning red underwear to play mahjong: Superstitious beliefs and problem gambling among Chinese mahjong players in Macau. Gambling Research: Journal of the National Association for Gambling Studies (Australia), 22(1), 18. Spronck, P., Sprinkhuizen-Kuyper, I., & Postma, E. (2004). Difficulty scaling of game AI. In Proceedings of the 5th International Conference on Intelligent Games and Simulation (GAME-ON 2004) (pp. 33-37). Cheung, S. (2014). Hybrid Learning : Theory and Practice : 7th International Conference, ICHL 2014, Shanghai, China, August 8-10, 2014. Proceedings. Sato, H., Shirakawa, T., Hagihara, A., & Maeda, K. (2017). An analysis of play style of advanced mahjong players toward the implementation of strong AI player. International Journal of Parallel, Emergent and Distributed Systems, 32(2), 195-205. Gee, J. (2005). Learning by Design: Good Video Games as Learning Machines. E-Learning and Digital Media, 2(1), 5-16. Canossa, A., & Drachen, A. (2009, September). Patterns of Play: Play-Personas in User-Centred Game Development. In DiGRA Conference. 14