An Audience Participation Game with Difficulty Adjustment and Rap-style Commentary based on Audience Inputs Yiming Zhang * , Albertus Agung * , Pujana Paliyawan † , Ruck Thawonmas ‡ * Graduate School of Information Science and Engineering † Research Organization of Science and Technology ‡ College of Information Science and Engineering *†‡ Ritsumeikan University, Japan ‡ ruck@is.ritsumei.ac.jp Abstract—This paper presents a method for adjustment of game difficulty and commentary generation based on audiences’ input to be used in Audience Participation Games on Twitch. We integrate difficulty adjustment and rap-style commentary in our game system with an expectation to boost the game experience of players and audiences. The game allows the audience to influence the game difficulty in real-time by sending command messages meanwhile rap-style commentary will also be procedurally generated. Index Terms—Audience Participation Games, Game Commen- tary, Rap Music, Game Difficulty Adjustment I. I NTRODUCTION Recently, the dramatic growth of game live streaming has been highlighted [1], and Twitch is one of the most popular platforms [2]. Traditionally, game live streaming audiences can only watch the game, but to enhance the role of the audience, a game design concept called ”Audience Participation Games (APGs)” was proposed [3]. This concept has been known for its effectivity in promoting social connectivity [3]. We are investigated designs for enhancing the experience of both players and audiences in APGs and found that difficulty adjustment is one effective method [4] for promoting player experience, while the game commentary was reported to be effective in keeping audiences entertained and informed [5]. Additionally, musical-style commentary, such as rap-style [6] was reported positively affect emotions, thus promote mental health. Therefore, in this work, we present a design that com- bines difficulty adjustment with rap-style game commentary generation. In the proposed game system (Fig.1), a player broadcasts the gameplay on a live streaming platform, and the audience can influence the game state by sending command messages in real-time. These messages are processed and game difficulty will be adjusted based on the result, meanwhile rap- style commentary is also generated. Contributions in this study are as the following. 1) The first endless running APG based on Runner [7]. 2) An APG that allows the audience to adjust game diffi- culty through chat messages. This work was supported in part by KAKENHI KIBAN (C) 19K12291. Fig. 1. Screenshot of the proposed system. II. EXISTING WORK A. Dynamic Difficulty Adjustment (DDA) DDA is an idea of adjusting game difficulty dynamically based on the player’s performance so as not to make players bored, when the game is too easy, and not frustrated when the game is too difficult. The goal is to create a game flow that fits the game challenge level to the player’s skill. Traditional DDA solely focuses on the player aspect; audiences are not considered, which can make them feel disconnected from the game difficulty. Therefore, we presents a design for difficulty adjustment based on audience inputs. B. Game commentary generation Game commentary generation is employed to replace the need for human commentators with the capability to provide content to live streaming game audiences on a 24/7 basis. In big game competition events, it is common to see commen- tators entertaining audiences [5]. Recently a study by Jum- neanbun et al. [6] claimed that providing game commentary in a musical fashion can enhance experience better than in a non-musical fashion; they presented a rap-style commentary that not only improved game experience but also helped to relieve stress–this system is designed based on the therapeutic method called “hip-hop therapy.” However, in their work, they only focused on the experience of APG audiences, ignoring the player aspect (players in their work are AIs).