This is a non-sequitur that came up in a recent conversation, and it’s something that many experienced MMO developers and players are well aware of.
In online spaces, emergent play is as important as social play.
Emergent gameplay behaviors (”unforeseen interactions outside of the original intent, which frequently provide an unexpected result”) can exist between players and the system, between the players and the AIs, between AIs and AIs, and so on.
They can exist between anything that interacts with anything else.
A game system that fosters emergent behaviors is more likely to give users the ability to entertain themselves in your 3d world/2d interactive environment/web based spreadsheet game for many more minutes/hours/weeks/months than you’ll be able to create content to keep them engaged in a way that’s mutually beneficial.
Emergent play lets people experiment “harmlessly” with pushing the boundaries in a way that the same behaviors in social play would be unacceptable or detrimental. (e.g. NPCs don’t walk away from a product or brand with a negative impression when they’re “experimented upon” by curious players.)
Degenerate gameplay is generally undesirable. I use that in a literal (not moral) sense: “A strategy/path of action/combination of resources or interactions that is both unforseen and so beneficial that it becomes the sole way to play. Not partaking in that specific, narrow path of activity either outright precludes “success” in an environment, or drastically reduces the amount of fun a person can derive from an experience.”
Degenerate gameplay is a small subset of emergent gameplay.
Given this relationship, attempts to systemically pre-empt degenerate gameplay frequently have the unfortunate side effect of outright preventing beneficial emergent behaviors. In a multi-player online environment, this can be a significant contributor to a failure to thrive.
It’s important to address the worst of the worst ahead of time — That’s one place where knowing where to strike a balance comes in — but with targeted solutions, despite the fact that targeted solutions frequently require more effort over time, both in maintaining an awareness and and being able to address the correct problem. However, that cost is variable, and never guaranteed to occur.
The other place balance comes into play is having a good sense of which potential problems are safe enough to address if and only if they become real problems. Fixing some problems before they actually exist often comes with an immediate cost that’s best left unpaid until (and if) you need to.
In short - Creating an environment that maximizes its potential to succeed as a whole is far more important than creating one in which all potential for degenerate play is pre-emptively stamped out.