While artificial intelligence and machine learning are nothing new, during the last year, AI technology has been one of the most-discussed topics — not just in the world of technology and gaming, but in almost all news around the world.
Advocates say that it can help humans in almost every field, while critics advise caution and scepticism. What impact will AI have on esports and gaming, and how can it be used to help the broader industry?
Competitive advantages
One of the most interesting things AI can do for esports is help teams and organisations streamline their training.
A day in the life of a professional player includes several hours of training and practising, often including scrims (practise matches) with similarly-ranked teams and players. The results of these matches are then analysed by analysts and coaches and are used to find weak spots and plan for improvement.
There are a wide range of AI-powered tools that aim to analyse the performance of players and teams, and each one of them is tailored to the needs of a specific game, such as VALORANT or League of Legends.
For example, AI tools in FPS games use match replays to tell the player how fast they reacted, whether their positioning was right, and how well they aimed. This information is often graphically shown to paint a better picture of an individual’s in-game performance.
Chris DeAppolonio, Chief Innovation Officer and Interim CEO of Evil Geniuses, told Esports Insider that the company uses AI tools for this exact purpose, through their partnership with Hewlett Packard Enterprise:
“Through our partnership with HPE, AI provides our coaches and analysts with a third eye to identify patterns and trends. This helps us improve performance, strategize to win, and uncover talent from across the globe.”
Additionally, the usage of ‘bots’ — software that helps players gain an unfair advantage (such as guiding the crosshair to opponents in FPS games) — has long been a topic in esports, AI can also be used to fight these practices with great results. Scientists from the University of Texas have recently made advancements in using artificial intelligence to detect cheats and detailed them in a paper published in the IEEE Transactions on Dependable and Secure Computing journal.
The idea was to use AI to scan packets of data that a user sends to the server, and figure out which ones have unusual segments to them. For a cheat to work, it must send commands to the game’s server which are not present if a player plays fair, so AI algorithms based on the server were able to detect abnormal pieces of information coming from the client PC, thus pointing the finger in the cheater’s direction.
Although still in its inception and only working in Counter-Strike, the tool has proven to be very effective in figuring out who cheats, and who doesn’t.
Non-competitive uses
Outside of esports competitions, AI is also used for a wide range of purposes. A notable example is automating content creation for esports teams and organisations by using AI tools that automatically create highlight reels and clips to be used on social media and other platforms. This enables those employed by the organisation to spend less manpower on creating clips from streams.
The AI tools can ‘see’ what is happening on the screen during matches, and based on a large number of parameters figure out whether action on the screen is noteworthy. This includes impressive plays in competitive games such as multi-kills in the likes of VALORANT or CS:GO, or exciting duels in the likes of Fortnite or PUBG: Battlegrounds. Content managers then pick which one of the clipped videos they want to use and create social media posts with them.
On the other hand, companies such as Shikenso Analytics use AI for a completely different purpose. The analytics company’s software also watches the streams and games, but focuses on branding and partnership placement. The tool can then calculate the amount of exposure a certain brand has on stream, which then enables both teams and their partners to gauge the value of partnerships.
For example, if a sponsor’s logo is placed on a team’s jersey, the tool can tell the exact amount of time the logo has been featured on the stream, and estimate how much value this generated.
As esports education programmes grow across the world, the inclusion of AI technology can also help support esports courses.
Craig Skilling, the Esports Coordinator at American Public University System (APUS), told Esports Insider that while the university is yet to use AI in its programmes, there are many talks held about what can be done to improve the overall quality of education for the students.
Skilling noted that the university — which runs online Bachelor of Science esports degrees — works to educate students about the usefulness of AI, and often encourages students to try and use AI as a resource and productivity tool for their daily lives. The university’s esports club, however, does use AI in its daily activities.
“APUS’ esports club embraces AI and includes the technology in its daily operations. The students utilise AI for various purposes, such as creating graphics and media content for the club. This enables students to enhance their AI knowledge and increase the visibility of the organisation.”
However, AI’s fast growth needs to be taken with caution due to the fact that not all issues and problems in esports and other industries can be solved by AI. There are arguably many companies simply trying to include AI solutions in their offerings because AI is in fashion now.
Still, just like in other industries, AI is here to stay in esports — and used right, it can help players, organisations, students and brands make the best out of their investments in the industry.
Supported by American Public University System