How do you create a system of game mechanics that best creates the opportunity for interesting, emergent behavior in combination with each other?
Many games have interesting combo systems, and some are better than others. Magic: The Gathering and Terraforming Mars. Skill trees in Diablo III or World of Warcraft.
What are techniques for designing a set of mechanics that have interesting combos? What types of combos are there? How can a set of mechanics be evaluated for how interesting their ‘combo’s are? Is it a systemizable problem? How can it be made fun and comprehensible? What is the best I could do procedurally?
2 thoughts on “Hard Hitting Combos”
You know, my go-to for inspiration whenever designing systems for this goal space is STILL Harvey and Randy Smith’s 2004 GDC talk about emergence, and the stimulus-response system they outline there that the various Looking Glass alums have been using in every game from Thief to Deus Ex to Bioshock to Dishonored to Prey etc. etc. etc. It’s my favorite “standard” approach to the problem. I’d be interested in this report if you guys build on that and show us what’s next.
This is moving past combos as as such, but here are a couple of thoughts on emergence that I’ve been working on. I suspect — and would love to be able to discuss this with all of you! — that there’s a necessary and sufficient set of behavioral meta-attributes to get to emergence. That is, if some pieces in a game have behaviors that are
– generic in construction; general purpose rather than single-purpose, and useful in multiple contexts
– local in action: they affect other pieces “near” them geographically or organizationally
– modular in construction: they encapsulate their own behavior and honor the encapsulation of others
– and they are able to affect each other in ways that form structural or functional reinforcing or balancing loops
then they have a very good chance of creating emergent behaviors. Omit any one of these, and the emergence never appears or falls apart quickly.
The problem with emergence is that it’s almost impossible to see it just by looking at the pieces and their behaviors. It’s like looking at a pile of legos and being able to see all the things you could build with them (legos obey the rules above too). You have to create the parts, add the behaviors, and then see what they do when they function to see if anything interesting and stable arises from them.
I’d love to hear others’ thoughts on this!
Comments are closed.