My problem with encounter balance is similar. A game is played by people. A game structured as part of well balanced mechanical breakfast seeks to challenge collisions between sets of statistics rather than the person playing the game. My character sheet is an object. The numbers written there have no need or care to be challenged.
True, but those numbers are generated and manipulated by a person. Plus in some cases it's more appropriate to challenge the entity created and defined by those numbers, depending on how much you want the play to be about the players-as-characters rather than the players-as-themselves.
Theres also an irony in that this can be read as "I don't like balanced encounters as the game isn't *enough* of a tactical wargame".
Connected with this is the terrible evil known as DM fiat. To those who would question why they should put any trust of success into the DM's hands I say this: As a player do you think your ideas and plans should have any influence on the outcome of your action?
DM fiat is, in my experience, a tool. It can be used well or badly. If I have crunched the numbers and set up my abilites such to execute a plan well and it is fiat-failed, then this is probably a bad use of it. If I am setting a situation up to deliberately fail (maybe its a appropriate to my characterisation) and it fiat-succeds, this is also a probably a bad use of it.
You have to trust your DM, and part of that is trusting them to rule consistently. Otherwise you risk playing Mother-May-I.
If the answer is yes then rulings/ DM fiat is your friend. Anyone who has ever played in a game and succeeded due to a well executed plan when the numbers said you should have crashed and burned knows this.
Honestly, I don't agree. If, in fiction, a plan has a one-in-a-million chance of succeeding, then it will always work. In the game, if the players make a plan with a one-in-a-million chance of succeeding, then surely they should either a) ride those odds, its their plan after all and to do otherwise is to negate thier choices and actions or b) come up with a plan with a better chance of success.
So it comes down to degrees of mechanics vs. rulings. The heavier the mechanics and statistics get the less relevant the people on either side of them are.
So a balanced rules-heavy system supposedly allows inexperienced DMs to run smooth games, at least thats what I keep hearing. A smooth running game does not always equal a rewarding or satisfying game.
No, it doesn't. But a game that does not run smoothly does not always equal a rewarding or satisfying game either.
As an analogy, lets take motorbikes.
My father owns two. One is a Norton Commando. Its a classic British bike and the ride and engine tone are incredible. You can fiddle with the engine to your heart's content and so on.
The other is a Kawasaki, and compared to the Commando its a souless hunk of metal. Engine is all black boxes and it sounds awful.
However, if you go out touring, the Norton is much more likely to break down, less fuel efficient and doesn't handle as well. You need to be sure you are carrying spares and the kick start, while fun, stands a good chance of skinning your shins. The Kawasaki on the other hand, is a much less stressful ride - it just works so you can spend your time enjoying the scenery and relaxing into the ride.
So, which is more satisfying/fufilling? Matter of taste. If you are interested in the visceral experience of the ride, the Norton, hands down, no contest. If you are more interested in the journey and destination, then probably the Kawasaki. If you were starting out in bikes, then the Kawasaki is probably a better choice for you too as it's vastly less work.
Essentially, thats why I prefer a balanced/smooth running game - I'm more interested in the journey than the mechanics. If I can trust the game to run with minimal input from me, then I can spend more time worrying about worldbuilding or whatever.