Dynamic vs Random
Keen has another post up lamenting the stagnant nature of modern MMO game design, while suggesting devs should instead be using ideas from games that came out 15+ years ago and nobody plays today. Uh… huh. This time the topic is mob AI and how things would be so much better if mobs behaved randomly dynamically!
Another idea for improving mob AI was more along the lines of unpredictable elements influencing monster behavior. “A long list of random hidden stats would affect how mobs interact. Using the orc example again, one lone orc that spots three players may attack if his strength and bravery stats are high while intelligence is low. A different orc may gather friends.” I love the idea of having visible cues for these traits such as bigger orcs probably having more bravery, and scrawny orcs having more magical abilities or intelligence — intelligence would likely mean getting friends before charging in alone.
The big problem with dynamic behavior in games is that it’s often indistinguishable from random behavior from the player’s perspective. One of the examples from Keen’s post is about having orcs with “hidden stats” like Bravery vs Intelligence that govern whether they fight against multiple players or call for backup. Why bother? Unless players have a Scan spell or something, there is no difference between carefully-structured AI behavior and rolling a d20 to determine whether an orc runs away. Nevermind how the triggers being visible (via Scan or visual cues) undermine all sense of dynamism. Big orc? Probably not running away. If the orc does run away, that’s just bad RNG.
There is no way past this paradox. If you know how they are going to react based on programming logic, the behavior is not unpredictable. If the behavior is unpredictable, even if it’s governed by hidden logic, it is indistinguishable from pure randomness. Besides, the two absolute worst mob behaviors in any game are A) when mobs run away at low health to chain into other mobs, and B) when there is no sense to their actions. Both of which are exactly what is being advocated for here.
I consider the topic of AI in games generally to be one of those subtle designer/player traps. It is trivially easy to create an opponent that a human player could never win against. Creating an opponent that taxes a player to their limit (and not beyond) is much more difficult, and the extent to which a player can be taxed varies by the player. From a defeated player’s perspective, there is no difference between an enemy they aren’t skilled enough to beat and an unbeatable enemy.
You have to ask yourself what you, as a hypothetical designer, are actually trying to accomplish. That answer should be “to have my intended audience have fun.” Unpredictable and tough mobs can be fun for someone somewhere, sure, but as Wildstar is demonstrating, perhaps that doesn’t actually include all that many people. Having to memorize 10+ minute raid dances is bad enough without tacking convoluted mob behavior outside of raids on top. Sometimes you just want to kill shit via a fun combat system.
Themepark MMO players enjoy simple, repetitive tasks – news at 11.