Back in 2003 when I started on my first moba (Dota Outland), hero design really was a case of "anything goes". At the time my heroes had six abilities: the six coolest things I could come up with, or find graphics for, that fit the character model. That led to some funny heroes, but there wasn't much synergy or elegance.
Since then, my hero design theory has come a long way. I attribute that to spending seven years working on RoW1 with complete freedom to try out whatever ideas I wanted, and being incredibly fortunate to have players who gave responsive, detailed feedback about what was working, and (more often) what wasn't.
At this point, I'm comfortable with the general "structure" I want for heroes in RoW2. Structure is important: the hero is a player's way of interacting with the rest of the mechanics in the game, so a "good hero" should always be built mindfully of the game it's featuring in.
I also know a lot of the ideas I'd like to include, both mechanical and thematic. Some of those are based on RoW1 material, others are new.
The challenging part will be slotting the ideas into the structure to assemble complete and compelling heroes. I need to do an exceptionally good job of that, for 30+ heroes at a minimum. I don't think there's any excuse for less; not in 2018 when everyone has had enough time to think about how kits should be constructed, and the competition is of such a high standard.
Thus: Hero Planning! How do you plan a cast of 30+ heroes, without spending the time to crunch through them all? And how do you ensure heroes will be compelling and fascinating, rather than just good?
Many of my Dota Outland and RoW1 heroes had abilities that could do stuff, but you didn't play the heroes differently to any others. Show up, use your skills, build items, end the game. My better designs have tension: a potent strength, coupled with a serious weakness. The player has to contort themselves to leverage the strength and mitigate the weakness, which produces new kinds of play that are exciting and fresh.
With this in mind we can plan out 30+ tensions, and then work on any subset of those without knowing specifics for the rest of the cast. Neat!
So how do we produce a list of 30+ tensions that a player might find engaging? One option is to look at game mechanics, but today I want to talk about modelling tensions from human needs.
Human needs are familiar to everybody: they're the natural tensions of being alive. If you have unmet needs (and almost everybody has at least some), you won't feel great about them. There's pressure to change, or to create change.
The proposition I am working with is that "giving heroes a distinctive strength and weakness" can be evolved into something more compelling by having the weakness negate a human need, while the strength pinpoints and satisfies a human need.
To give an example, one of RoW's heroes is Rasputin. He is unconditionally visible to the enemy team at all times. That negates the human need for space: he's not going to get any alone-time, is easily ganked, and is constantly responsible for any information his positioning leaks.
At the same time he satisfies a need for presence: the enemy team will be constantly aware of him, and he has a moderately large threat radius. You can bet some players would have a lot of fun with that! With the right supporting abilities, this tension produces a compelling hero.
So: that's what I'm exploring right now. You can check out this needs inventory for a full list of human needs. I won't necessarily work with all of them, but I think this approach is promising because it's rooted in humanity and psychology in a way that working from game mechanics isn't. I hope the resulting heroes might be experiences that will resonate with players.
Feel free to leave thoughts, comments, or tell me I'm crazy!