There are many ways to GM, but this one is mine.
It is pretty much like this, as opposed to the one defended in past on Twenty Sided (here, here, and here) and many other styles discussed in many other places. I won’t quite fuck the story as Merten did, for example.
My goal as a GM can be summed up into a short format: Force players to make choices as/for their characters, and impose the consequences of these choices on the characters.
This sounds heavy-handed, and, in a way, is. The key is in making the choices relevant to the characters (players tend to be interested in the fate of their characters).
For example, assuming typical D&D and good-inclined player characters: A medium-to-high level party rushes to save a village, only to find it swarming with hobgoblins who have takes the former residents hostage. A reasonably powerful (as in, defeatable, ut not a push-over) devil offers a deal to the party: Party gives the Powerful and Magical Artifact of Utter Evil to the devil and he will let the villagers go, and doing whatever the PCs want to the hobgoblins. If not, well, the hobgoblins are kinda hungry after the forced marching they endured to get here. Unless this is the very start of the campaign, previous choices of players, of perhaps failed skill rolls (ride to hurry there) are the reason for this event happening. What will the players make the characters do?
Maybe they decide to screw it and slay the foul beasts, probably resulting in heavy casualties on villagers and hobgoblins alike, possibly killing the devil, and maybe even one of them dying. Maybe they hand over the artifact and the devil lets the people go (with some potential surprises hidden in them, like some sired half-devils or such, in all likelihood). Maybe they come up with a cunning plan that saves the villagers and even slays the devil. Whatever they do, the players made the call. If they fail miserably, that was due to their own actions (and possible hubris or greed or trustfulness or soft hearts or whatever). If they manage to pull off a heroic rescue, it is truly a meaningful one, because the chance of failure was real. Maybe the devil gets the artifact and gets away, which means that the next adventure is pretty clear and the players will be eager to pursue it. Maybe the village is slaughtered, but the world as a whole is safe from the threat the devil and the artifact would have been, if they had managed to get together. The next bad guy will take a city hostage. The next one a kingdom. How long will the characters keep the artifact when the world around them is crumbling into darkness because they have it? When will they start using it? And what will the price for that be?
Many players would not enjoy that kind of game, because they will suffer wherever they go. That is the problem of those players. A game without trouble is a boring one. Piling complications atop each other is the way to go. At some point the characters will perish or solve their problems. This signifies an end to the story arc. There are characters, they face adversity, and either die in the process (resulting in tragedy) or triumph in the end, having revealed something about themselves in the process. This is what stories are about.
The way I use to get this sort of gameplay is: First make a starting situation and enough setting to get players started with character generation. Restrictions add to creativity and help focus the game. Make sure all the player characters are entwined in the situation and with each other. Make some bad guys or conflicting NPCs or whatever and make sure they are entwined in the situation and player characters, too. Ideally they want something from the PCs.
Once players engage with the starting situation, simply play the NPCs according to their motivations. Let player decisions and dice rolls dictate the outcome of the game. A well-built starting situation is robust and will not crash due to player choices or luck. Further, it will provide opportunities for further conflicts and such.
If/When the initial explosions end but game continues, the key is in throwing ard decisions at player characters and having them interact with each other as much as possible. To my knowledge, pure sandbox play, where an entire setting is built ahead of time, is not ideal for this, though a well-built sandbox certainly works. Goal is in making situations where the choices of players will form more similar situations.
When/If the flood of interesting consequences dies down, game ends. A new one can be started with same characters when the next interesting situation happens in their lives. An entirely new chars can be made. Or maybe some of the old characters continue, but some new ones appear, too. The ideal timing for this is when a new book full of shiny new prestige classes appears. Players can change characters at will without screwing anything up.
There are some aspects of the style which are far from trivial. A list of topics I hope to handle in the future (adding new one is encouraged), in no particular order:
- Building a good starting situation.
- Building good NPCs.
- Building good PCs.
- Using the dice in the open and not fudging.
- Using the NPCs in the open.
- Splitting parties, play based on PvP.
- Improvisation and preparation (they are the same thing), including using heavy rulesets.
- Diceless and freeform gaming, if I manage to master those.
- Secrets or lack thereof.