A question about transitioning players to GM-less gaming already exists (and it's a good one). I, however, am looking for a more exhaustive description (thus CW) of approaches to GM-less gaming. As such, I'll describe the traditional effects associated with the GM role, and answerers can help out describe how gaps that arise from the removal of the GM role can be filled.
Note: the GM is not automatically the host of the session: I'm going to ignore location / snacks / agenda's, etc.
Traditional elements of the GM role:
- Cohesive story (generation)
- Appropriate threat / challenge levels, both strategically and tactically
- Neutral arbiter of inter-player dispute
- NPC acting
- Story / scene pacing
- social cohesion / group makeup selection / spotlight distribution (this one may fall outside scope, but is often associated, I think)
If I missed antyhing, please let me know in the comments and I'll edit the question!
Best Answer
Cohesive story (generation)
GMless games tend to have mechanics that are specifically aimed at generating a structured story. Often this is accomplished by giving the players choices and rules-assigned powers that directly or indirectly are about story elements instead of focusing on character actions. For instance, instead of having the player choose where to move their character and which enemy to target with an attack, the rules may instead ask the player to choose one of: 1) decide what's at stake in this fight, or 2) whether they succeed; and the other players together might be given the other option to decide. That's just one example—there are as many ways of handling such story choices as there are GMless systems.
By focusing the rules on the twists and turns of story and letting the players freely describe what goes on apart from the plot developments, a cohesive story can be made from everyone's input.
Appropriate threat / challenge levels, both strategically and tactically
GMless games generally pin difficulty levels to something other than the strength of the characters. It might be related to which of three "acts" the game is currently at; it might be based on a limited currency that the other players can use to set the difficulty.
GMless games tend to be very non-tactical. Many are strategic, but the strategy often lies more in managing the player's resources that govern story control or decision making rather than in-character strategy. Strategy and tactics for the actual characters is more often handled as a plot element than a player-challenge element.
Neutral arbiter of inter-player dispute
Most GMless games are structured such that it is always clear who has the final (or only) decision-making power at a particular point in the game, or over a particular game-mechanic or in-fiction decision. One game I know has a "no it happens this way" rules structure, where as soon as someone disagrees with what is happening the roleplaying is paused, everyone has a chance to offer a counter-proposal for what should happen, and then each player has up to five points to split among the proposals. The highest-ranked proposal is then played out by everyone together.
By making sure there is both a clear decision-making authority (and that everyone gets their chance to hold that authority as the game goes on), many games prevent the players from ever getting into a situation that can be argued over.
NPC acting
In GMless games, the player characters will as often be together as apart. NPCs are often controlled by players who either don't have characters active right now, or by players whose PC isn't the centre of the action.
Other games will give players the ability to decide the actions and choices of NPCs in their own character's favour if they are willing to spend the appropriate "plot resources" to "buy" that decision-making power.
Story / scene pacing
Some games leave this up to the group, knowing that the group will quickly get a sense for their preferences for scene and story pacing.
Some games make story and scene pacing integral to the rules. The end of a scene, or the beginning of a new act of the story, might be triggered by unrelated player decisions, by some kind of "decision power" resource falling to a specific level, or according to certain in-fiction milestones that are obvious or pre-determined during scene setup.
Some games have a rotating scene-framing authority, and a scene is over as soon as that player whose "scene turn" it is says that the scene is done.
social cohesion / group makeup selection / spotlight distribution (this one may fall outside scope, but is often associated, I think)
GMless games are rarely set up so that the PCs form a "party" or cohesive group. They often make it very natural to have PCs that are split up or even each other's enemy. So, group makeup in-game is rarely relevant.
Spotlight distribution is most often a part of the rules, either as part of a scene-framing turn mechanic, or as a "spotlight time" resource attached to characters, or some other organic way of regulating spotlight time.
Social cohesion of the player group is rarely addressed, except by making sure that the setup process of the game gives everyone a lot of buy-in to the story premise and characters. If everyone's excited about the game they're about to play, keeping everyone on-track is rarely a problem.