Although HPE only officially announced its entrance into the AI cloud market last week, the company has been working with artificial intelligence (AI) and large language models (LLMs) for some time. One of its partnerships – its collaboration with esports organization Evil Geniuses – is already seeing the benefits of that work.
Founded in 1999, Evil Geniuses (EG) is one of the oldest esports groups around today and has won over 180 championships in competitive gaming. EG’s primary objective centers around signing players and assembling teams from around the world to compete in professional tournaments across games like "Call of Duty," "Counter-Strike," "League of Legends," "Dota 2" and more. HPE and EG first announced their partnership at HPE Discover 2022 — not only providing the AI data-analytic resources but also integrating the edge-to-cloud GreenLake platform to power those services.
EG’s Chief Innovation Officer Chris DeAppolonio told Silverlinings during the recent HPE Discover event that the organization had always been data-driven in its decision-making, building tools in-house “very ad hoc.” But it quickly realized it needed a partner to take these tools to the next level. HPE’s end-to-end solution has gone beyond just cloud computing, he noted, with data scientists and software engineers helping build tools for EG’s coaching staff.
“Tech is the backbone of our esports competition,” he explained. “We can find those opportunities to put ourselves in a position to win or find undiscovered talent and offer more empirical opportunity for scouting and talent [development].”
DeAppolonio said that one of the greatest outcomes to come from the data-driven approach is the way it opens up opportunities for new players. “We want to be able to find people. Whatever their background or wherever they're from, if they've got the skill, they should be able to get up to the highest stage.”
Comms get an AI assist
The two companies built their first tool, an audio analysis platform, last year. This tool analyzes all dialogue in practices and games to help identify communication patterns, good in-game leaders and whether team members make the right call at the right time.The coaching staff isn’t allowed to speak during competition, so tournaments rely heavily on team communication.
“Maybe in practice, this one player talks a bunch. And then when they get on stage, the player doesn't talk and we lose, right? We wouldn't know that unless our coaches sat through hours and hours of data,” DeAppolonio explained. “We're able to use natural language processing, and AI to understand speaking delays, sentiment and tone so that we can better prepare ourselves for stage matches and get our players in the position to communicate more effectively,” he continued.
DeAppolonio added that the tool has found additional “interesting use cases” like finding unique and funny sound bites from players — the Leroy Jenkins in the group — that can then be used for advertisements and marketing.
Cream of the crop
The latest tool to come from their collaboration is a draft simulator using AI and ML, which they rolled out at the beginning of the year.
DeAppolonio explained in games like "League of Legends," there are player selection drafts. “So they all have different skills; they fill different roles. Some are healers, some are damage dealers, some are good at scouting, some are tanks [and] can take a lot of damage. There [are] billions of permutations.”
The draft simulator allows EG to analyze its own strategies for draft picks, predict which picks will be optimal in specific scenarios, and prepares the coaching staff “to a tee” for what other teams will pick.
While in the last season, EG’s team did not win, DeAppolonio noted that they were well prepared in predictively analyzing the other teams far more accurately. “I think the draft tool has helped us get to make sure we're starting off on the right foot because you don't necessarily win a match with the draft, but you can lose a match, similar to a chess opening,” he concluded.