Why did I teach them to spit acid at me?

So lately I've been mostly focusing on AI, or any lack there of. Unfortunately with Ruby we've gone the hard route and wanted bad guys that are actually bad, not just lumps that wiggle or pace that hurt to touch. Nope it's time for some good old fashioned these things are trying to kill me. It's certainly a more painful route to develop for, taking a vast amount of more planning and code than the simpler style of enemies but we're going for a game that you don't get to just memorize. I may have planned out this AI, essentially teaching it everything it knows, but it doesn't mean I know exactly what it's going to do.

Pattern recognition is a standard feature for a lot of old school games, where players are able to analyze the game and reliably predict the actions of opponents to a high level. While we can certainly see a great fondness for games using this classic design method to overcome processor limitations, we like to make things a bit more of a pain in the posterior for would be game guide writers and cheats. Now we don't have the time or resources to invest in truly complex intelligence for our monsters, we do have a big improvement in processing power than the classic games and so we'll put it to use.

Coming Robot Apocalypse? We did our part!
(image found on geekologie.com)

Giving our monsters decision making abilities allows us to, I believe, improve the realism of the world, and provide a much more challenging system. Players can't just read a guide and know how to be good, our system won't generate results too predictable (or at least that's the plan), forcing the player to instead invest in the game a level of thought and care to develop their skills and character that is the heart of RPGs. This is one of the reasons we are releasing episode 0, it gives players the opportunity to learn about the protagonist of our story, and develop those skills they'll need to survive the adventure ahead.

That being said, we are still dealing with a mobile game, and a limited budget. Our creatures don't formulate chess-like strategies, and given the story of episode one, may not be the most realistic representation that players will encounter in our game. So we've struck a balance, we aren't going to settle for pre-plotted movement and blind undefended monsters that offer little more than a terrain hazard, but players aren't going to be overwhelmed by a coordinated opponent of thousands orf accurately predicting automotons. Where is the sweet-spot? I wish I knew, but when you're playtesting, you know you're resolving your AI issues when you start dying a lot. I take an awful lot of them out with me, but you can't just look away and expect them not to turn you into tonight's supper.

Episode 0 is certainly a special case and you'll get to find out why, but as you get to explore the world of Ruby Rust I think you'll find a very satisfying responsiveness and vitality to the world, and if I've done my job right, every so often get yourself right stomped.