I have been following the great blog, Dungeon of Signs, for a few months now and I have to say that I love Gus' take on gonzo sword, science, and sorcery classic D&D. In fact Gus' play reports are what turned me on to Anomalous Subsurface Environment which arrived last week. A great purchase!
In addition to the piles of great content (tables, house rules, etc.) and play reports that Gus posts, he, from time to time, discusses his take on gaming and DMing. In this post he reduces running combats down to answering three simple questions. Should combat be...
1. Lethal vs. Obstacle?
2. Unpredictable vs. Managed?
3. Simple vs. Complicated?
Gus discusses the varied implications on play of how a DM answers these questions. I would have to say that, in my game, all of the answers are pretty much down the middle.
Lethality in my games is a real risk. PCs die and my players know that I am not going to pull punches. At the same time, death is usually the result of bad play (i.e. making bad choices) though a few horrible rolls might still be enough to do someone in.
I do like to throw a monkey wrench into the action once in a while... I do use critical hits and misses for combat and spells too, which certainly adds some chaos and danger. But generally my game is pretty predictable. Items which are consumed over time (i.e. oil, torches, etc.) are metered and consumed or exhausted with die rolls (e.g. a torch is consumed by rolling a 1 on 1d6).
I like to keep things pretty simple. The crunchier bits I like to take care of behind the screen. The crunchiest thing I have in my game is a spell casting roll (see my Magic tab). I am also hacking the new martial class/combat dice systems from the D&D Next playtests into my game. Character creation is not like it was in 3.5/PF, choosing not completely understood options from massive lists; it has become more of a Chinese menu thing... choose one option from each of the 4 following columns. Rolling up a new character takes a about as much time as it did when I played AD&D/2e.