In order to define dynamic content within a group, identities need to be defined. It all comes back to classes, to roles, to niches. In the simplest way I can describe:
[Online especially] Games need to be built in a manner that allows people to fill a niche. This is best achieved through avatars based on class. Develop the class that speaks to a playstyle and can perform needed roles within the context of the game.
Don't Tell Me What I Can and Cannot Do
I've heard it, I can respect it, but I also can't buy it: some people prefer classless games, giving the possibility to the gamer in defining themselves based on their interests. This prevents being pigeonholed into a certain task and/or unable to perform another of interest. I can appreciate it's a design point, but one better suited to a game whereby the gamer is the main, and only, participant.
The second multiple parties are introduced, so too are differing viewpoints. My desire to perform as an XYZ might be inhibited by anothers, and vice versa. Context, limitations, are therefor a requirement. A small barrier of entry involving choice- a necessary evil. We do not gravitate to more structure, we gravitate into chaos. Therefor, design with order in mind- disorder will naturally occur.
My Toy Can Beat Your Toy Up
We might establish a better understanding through metaphors. I submit an avatar is like a vehicle, and content is like the terrain. If I want challenging content, I might describe the terrain as rocky. If I want to tackle the content, it will be in my best interest to have a wide wheelbase. These individual points of contact act in harmony to achieve a single goal, but the wide wheelbase indicates specialization. The further from the center, the more specialized.
Applied to group context, we see the need for specialization. The beauty of class-based designs is that they discourage the entropy of people wandering into chaos from defined nature. Done properly, they allow this to occur for those who want the middle area, but also allow even further specialization, or mix-and-matching.
Lucky Number '3'
Quick read on trinity assertion. What this doesn't fix, however, is the inherent need for solo viability. If I have nothing but purebreds, I have complete ineptitude when it comes to doing anything. I need to be able to do at least some of the other two basic combat tasks.
Here are the basic combat roles in picture form. Consider that the area covered by each section is the conglomeration of abilities that might achieve the goals of tanks, healing, or damaging. Direct abilities, over-time abilities, area abilities etc.
Consider that the area covered by a 'class' is the amount of the given role they are able to perform through various abilities: more tools, varying tools, at the disposal proportional to the area covered.
The rock. The only thing he can do is take a beating.
The glass-cannon. The only thing he can do is dish.
The healbot. The only thing he can do is regen.
The problem with these systems is they promote the ultimate in group stability, but the worst in solo viability. So how does one design a system that allows for both? Place the limitations on a class that their 'center of mass' must remain within a given role's boundaries.
Expanding the Concept
In the examples above, I show the most basic combat roles and the solutions that might be attributed to them. But through no stretch of the imagination should one be limited to the most basic design! It is possible to build a need, a niche, for an 'interupting' role, a 'control' role, a 'buffing' role, a 'debuffing' role etc. etc. into the design of a game that calls for these to be filled. The simpleton 2D design listing the 3 trinity roles has been shown, but how about we add another role and another dimension? Here is what the design axes for a 4role system might show:
Quickly, things become complicated. And the aforementioned roles could quickly take design into 4D, 5D, 6D etc. But the system concept still applies, even if the human mind is grounded in thinking in 3D only. Back to these axes: one might place the basic 3 along with control, if the overarching game called for that as a fourth, defined role:
The goal is to keep the center-mass along an axis, while allowing the shape of the space to expand onto other axes. The center is always an option, but I can only do so much as an individual. I must pick my sphere of influence.
Sliding into the Finish
Building a system that promotes inherent specialization while allowing the player to choose between complete soloability or ultimate group contribution. Keep the center-mass of a class aligned with a role, but allow more branching-out options into other roles, the closer the gamer feels like approaching their avatar to vanilla-center. Maybe you limit how close an intended specialist can come to center. Maybe you limit how far an intended hybrid can stray. Maybe you rig which axes are adjacent to the present axis of design (for hybridization options). Maybe you don't, and allow the players to define all that themselves.
But what you do, is you develop the structure a playerbase needs to be successful in a solo vs. group context.