When players disagree about how to play a game, the game is disrupted. This is usually easily resolved: the rules define how to play the game, thus reducing the chance of misaligned expectations. However, there is at least one place in role-playing games where "how to play" may be unspoken and, in fact, undiscovered, allowing the clash to become systemic. This is the clash between mission and event.
Before I get going, let's say up front that we are talking here about refereed games — games where one player is acting as a gatekeeper of the rules and as a keeper of secrets. He is directing or mediating or facilitating play by having an idea about what will happen in the game and by organizing player input to create a session that's consistent with everyones expectations (genre, verisimilitude, participation, and so on).
Starting at the beginning, the first collision is the game opener: the referee supplies either a mission or an event. One may be disguised as the other, which can be the source of confusion in expectations. That is, the referee is either going to give the players something to do or something to react to. He will probably have a clear idea of which he is providing but, in many cases, is unlikely to explicitly communicate to the players which he has provided. And here's where some friction lies: if a player expects a mission but never gets it, he will be frustrated and spend his energy trying to find the mission — trying to deduce the referee's intent when in fact he may have none. If a player expects an event but receives a mission he may balk and refuse the mission as uninteresting opting instead to wait for the event — or engineer it. When players at the table have different interpretations of the referee's opener, we get havoc unless someone realises the failure and speaks it.
Many games expect the mission to be the opener. Some games make this absolutely explicit — Dogs in the Vineyard, Mouse Guard, and Paranoia to name a few. Some seem to expect it but may not communicate (may not realise!) this. Most versions of Dungeons and Dragons fit in here, though I note that in fourth edition there is a distinct move towards making it explicit.
Some games expect the event to be the opener but I can't think of an example that makes it explicit. The fact that it is not stated in so many games, however, leads to many tables interpreting the game one way or the other and creating a local culture with a specific expectation of mission or event, and this table may find they have difficulty integrating new players or participating in other tables' games even though the same system is deployed.
When all players at the table have the same expectation for the opener, a style of play emerges that is distinct. Mission play acknowledges the artificial focus on a specific goal outside the game and gets down to achieving that goal immediately inside play. The referee is control of the direction and can judge success or failure. This style of play is goal-oriented and effectiveness of the players can be measured. Obviously these things are appealing — they create play that tends not to wander and where victory can be announced.
In mission play, the opener implicates a mission. The goal might yet be unstated, but a goal will emerge. The opener might be an explicit mission: the players are handed a mission sheet by their superiors, for example. More commonly the mission is implicit: players are told a rumour about a great treasure in an old dangerous ruin but are not explicitly told to go there. This lack of explicitness is the dangerous space in this kind of play — it's where a mismatch of expectations will cause players to fail to engage the rumour. Worse, if the players have not identified the fact of mission versus event play, the referee may hide the mission — "you all start in a bar" — expecting players to actively search for the mission before starting it — "the bartender, now that you've bribed him, has a story to tell about an old ruin". This is probably one of the most common sources of friction: the refusal to acknowledge and engage the choice of mission versus event.
Mission play is very well rewarded by specific game mechanisms. Declaring a mission as part of the rules (and rewarding its resolution, again, as a rule) drives play forward very effectively. There can be no confusion about what is expected. It does, however, constrain play to this specific type, which is more focus than role-playing games typically have and therefore might be seen as out of genre.
Event play is a much fuzzier animal. The players may have less of their game-brain devoted to the artificiality of the opener and instead direct their characters to react to events and create the story themselves. In this case less is expected of the referee and more is expected of the players, which will obviously create frictions if the expectations are mis-matched: the referee can easily be seen to "not be doing his job" yet it might not be recognized that the other players have one to do which picks up that slack.
In event play the opener is an event. Something happens that is intended to make the players react but that does not imply a correct solution. Thus the players are forced to engage their characters to decide how and what happens next. This places a burden on the referee to prepare for practically any response and a burden on the players to react without a forced direction of action. These freedoms can be paralyzing without system support — ways in which the system gives your character power over the story.
Event play is rewarded by rewarding players for engaging their characters. Making choices according to stated characteristics of their characters would warrant reward, for example — a character that has a specific character trait that implies "Heads headlong into danger" is given a game-mechanical benefit when his response to nearby boors is to flip their table and punch them in the eye. In this way conflict is created out of event by players manipulating system through their characters.
Successful hybrid play happens when players have the power to establish missions through system mechanisms. This increases the burden on everyone — the referee needs to respond to a wide variety of player choices but now must also find ways to make specific stories explicitly stated as desirable by the players interesting. he players themselves now also have the burden not only of reacting to events to drive play, but also to declare goals for themselves and plan to achieve them. In a way this happens with well established event players anyway, but actually stating it mechanically can be onerous. Reign does this with it's Mission attribute (one of the character's Passions) and it can work well in either form of play — either the player accepts a referee's mission or extract his own mission from the opening event. Or defines one before the event, giving the referee something to work with in designing the event and its surroundings.
The bottom line of all this is the importance of deciding what you are playing and communicating it: the choice between event and mission play entails substantially different pressures for players and referees including different kinds of preparation and even different ways to engage the system (which in many cases can remain exactly the same and facilitate opposing forms of play as needed — few games can be said to intend one or the other approach exclusively). Different players having unspoken differences in this particular expectation can easily fail to be communicated and yet be very destructive, either through deliberate inertia or insistent exploration of details that are not relevant to play (as other players read the opener).
JBB: I find _Mouse Guard_ interesting in this respect as it's very explicit about the mission structure, and uses events to "twist" the plot, triggered by failed rolls to overcome obstacles. The game is divided into a GM Turn and a Player Turn; during the latter, players are prompted to generate their own events. This then provides material enough to construct a new mission.
BJM: That's cool — I had kind of thought of it as a binary proposition (you're playing one way or another) but the idea that you can chain these expectations together is pretty illuminating. Makes communicating what you expect that much more important if you're going to switch like that. I think that's the way Reign works too — you set up an initial scene (event or mission) but players create distinct missions to be addressed. How you do that (again, explicit mission or reactive event) is not mechanical though.