Post by Eorl @ 12:05pm 16/04/14 | 0 Comments
League of Legends developer Riot Games has taken to their blog to reveal just some of the more recent changes that they've been making to their rather popular MOBA, namely in the AI bot development section.
Design Analyst RoamingNumeral headed the discussion, revealing that even high level players spend a lot of time practising against bots, and both co-op and PvP players have in the past asked for more “human”-like bots. Players requested less predictable bots that focus on last hits and combos rather than just “tougher” ones.
We surveyed both Coop-vs.-AI and PvP players to find out what they thought about bots. One very clear piece of feedback was a desire for more “human” bots: ones that are less predictable and focus on the same game elements that players do, such as last hitting and nailing their combos. Surprisingly, "tougher" bots was a less frequent response by comparison. Based on this data we set our sights on more-human bots as the overarching goal for this update, especially since such changes will make harder bots more interesting down the line.One of the key improvements to come from this discussion is bot threat evaluation, which the team at Riot believes has changed for the better. Akin to first-person shooters, the previous systems looked at how much damage has recently been received and uses that to determine future threat. The new systems instead try to deal with scenarios by looking not only at "health but also how many spells the bot and any nearby enemies have available."
Of course, other requests from players weren't as easy to fix. One of the more common suggestions was allowing bots to jungle or wall jump, which Riot has said is actually impossible to do, due to the fact that bots don't actually know what walls are or that they even exist. Jungling shares a similar story, with bots unable to determine what paths are "safe" and what are optimal, unless the team completely re-designed the pathing and terrain systems.
Much more can be found in the full blog post, including a look at how these fixes were achieved, and also how Riot optimised bot code to compensate for their new processes.