In the past few days, I've been thinking a lot about Dungeons and Dragons, about how each edition's strengths and weaknesses. Maybe I'm just prone to daydreaming and should be doing something more productive with my time, but that detail hasn't stopped me for 32 years and it won't stop me now.
This past weekend, in my bi-weekly game, I threw in an encounter that I wasn't certain the party could overcome. Or in other words, I knew when I designed it that the monsters weren't balanced against the PCs. In fact the encounter was very unbalanced in the monsters' favor. That is, if the PCs simply chose to skirmish them and fight them conventionally. Indeed, on the first round, the PCs did exactly that. And realized the creatures' Damage Reduction made them almost invulnerable to all but the most focused and lucky attacks. On paper, the fight had all the makings of a Total Party Kill.
What I learned was that this imbalance, this unfair advantage in the monsters' favor added a huge element of dramatic tension to the entire combat. Everybody had to coordinate, think outside the box, and invent - on the fly - tactics and strategies to neutralize the horrors. Because these desperate measures offered no guarantee of success, the excitement was heightened. Every roll of the dice held dire consequences in the balance. And when the party ultimately prevailed, there was a huge sense of victory. Of overcoming long odds. It was and remains pretty clear that the encounter could easily have gone the other way, and that the next game session might have been one of rolling up new characters.
Which brings me to the concept of an "Elegant Imbalance" in game design. By making the tools of an encounter too predictable, that is, making the odds too measured, the encounter loses dramatic tension. If the numbers and mechanics are too clear, if it is too obvious that the group will be fine if it has all the bases covered and if every weakeness has a compensating strength, then all the dice rolling and arithmatic becomes an exercise in weighing odds. A group of PCs engaged in a balanced encounter are almost assured of victory, and even a string of encounters, where party resources are expended at a steady rate, becomes a calculated string of victories. Defeat becomes a feature of bad luck, of the dice being fickle.
The point I'm trying to make is that, the rules of the game should support encounter design as an inexact science. Not random, or arbitrary, but hold within it the real possibility that despite a well-planned and executed strategy, defeat and death remain a strong possibility. Sometimes, that uncertainy should affect the DM as well, not-knowing if they may have, in the effort to create an exciting story, doomed their heroes to glorious defeat and death.
I've been a DM for over three decades now. I've been fortunate enough to run more adventues and campaigns than I can easily count. Out of all of them, those times when the entire party of adventurers perished, either by going in over thier head, or because they ran afoul of an unbalanced encounter are among the most memorable. My friends and I still tell stories about these defeats years and even decades later.