GDC is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

View, browse and sort the ever-growing list of sessions by day, pass type, topic, and format. With this Session Viewer, you can view GDC 2023 session details and speakers, and share your favorites via social media. You will be able to build your schedule and access it during the show via export or Mobile App, once live. Sessions do fill up and seating is first come, first serve, so arrive early to sessions that you would like to attend.

Pain and Gain: Obtained from Applying a Reinforcement Learning Game Bot to 'BarbarQ 2'

Xiaohang Xu  (Director of Technology, Electronic Soul)

Shun Bao  (Data Mining Engineer, Thunderfire UX)

Location: Room 2014, West Hall

Date: Friday, March 24

Time: 3:00 pm - 3:30 pm

Pass Type: All Access Pass, Core Pass

Topic: Design, Programming

Format: Lecture

Vault Recording: Video

Audience Level: All

In the online mobile video game, BarbarQ 2 by Electronic Soul, reinforcement learning-based game bots are applied to its 3vs3vs3 gameplay. These RL-based game bots are used as teammates or opponents of the human players. It turns out that the RL-based bots really improve the player experience.

However, there are a lot challenges in the development process, including the difference between game and AI, game-side bugs impacting negatively on AI model, difficulty of lifting the human-like level of game bots, and more. Fortunately, some innovative ideas have been implemented to solve these problems.

The high-quality game-bots are now available in the game. Some risks, ideas and benefits of applying RL-based game bots will be talked in the session.


Attendees will learn how to develop and apply reinforcement learning-based game bots in MOBA video game. They will also be inspired by their ideas of solving the problems met during the development process. Very importantly, attendees will understand risks they would take and benefits they would gain if they want to apply RL-based game bots into their games.

Intended Audience

This is for game professionals who have problems developing and deploying reinforcement learning-based game bots, or anyone wanting to apply RL-based game bots in their games but also want to know the risks and benefits.