Something that occurred to me recently is that there are three versions of every roleplaying game in existence.
First there's the game as written. This takes the form of the main rules, supplements, adventures and so on. In many cases, there will be some sort of discussion of tone, mood and so on (especially in games after the 1980s). This is where the game designers vision of how the game will be run is laid out and is perhaps the most artistic part of the proceedings.
Then there is the game as envisaged by the GM, how they put their interpretation on the rules and setting. Again this can be quite artistic, certain parts of the game will call out to the GM for exploration and those are the bits they'll focus on. This (usually) involves creating a story or at least situations and having an idea of where the game will go. In truth this is probably the most tenuous part of the process because the game here only exists as notes and ideas as opposed to a rulebook or something that is played. Nonetheless, its an important part of the process and in my opinion counts as a game in its own right.
Sometimes this is going to be the same, or if there are tweaks they'll be quite minor. If you're interesting in playing Dr Who, its unlikely you're going to move the game too far from the TV show, because that would defeat the point of playing at all. On the other hand, a game like Vampire the Masquerade has managed to produce many different forms of campaign, from the high game of underhanded politics to dark superhero games with Desert Eagles and fangs.
Lastly, there is the game as it is played. And this is the game that really exists, if only because its the one that makes contact with the real world.
The three games tend to be quite different. The game might read as something steeped in the Gothic. The GM might envisage something where players' choices can damn them as they are dragged into the dark mysteries that confront them. In play, well 90% of the time it seems to resemble Monty Python's Holy Grail on crack: with a higher body count.
This situation can lead to problems, just from the way gaming works. Information is vital here and one thing we need to remember is that whilst a mode of gaming is implied in the name of a game it doesn't necessarily follow that that's what the GM is bringing to the table. Dungeons and Dragons implies going into dungeons and killing, looting and burning your way to wealth. But the game of it could also involve a group of watchmen in a fantasy city dealing with the local underworld (the game, of course might eventually lead to a trek through the sewers, killing, looting and burning your way to wealth, but the framing is different to plucky adventurers out for a quick buck).
For example, if I say I'm running Dr Who, but that tells the player nothing about the intentions I have for the game. It doesn't communicate the idea that there'll be adventures throughout London's history, from Weeping Angels in Roman Britain to the near future where an aging population is dependent on servants grown from the Flesh for their day to day care. This might create dissonance between GM and players - where the GM wants to run a tight series centred on the mystery, the players want to bop about the cosmos doing not very much apart from getting into trouble. With these two conflicting ideas in place the game will stutter and burnup in a matter of weeks.
So there's a responsibility on both sides. The GM has responsibility not only to have an interpretation of the rules but also to communicate what he or she wants to do with the game. The player has a responsibility to take that on board and to say what they want to do. I know it can be argued that sometimes knowing what you want in a game takes time, you have to get to know your character; learn what they can do and orientate yourself in the setting after all. But, surely you as a player know where your interests are in terms of game play unless its your first game (and even then, you know what you like in terms of other media)? Having played games where at least 50% of the game was combat, and I felt sidelined because my character wasn't particularly fighty*, I can say that's not for me and if I were a player in a new game one of the first questions I asked would be about how much combat the GM was planning to introduce. As a GM I'd expect my players to do the same, if only because there's nothing worse than a game where the players act like knots on a log.
Players also have a responsibility for making sure their character fits the game; no point in playing if you're not going to buy into the basic ideas the GM is using. And GMs have a responsibility to check the character proposed is right for the game, not just in the sense of looking at how much damage they do, or if the build** is broken, but in the sense of whether the character concept fits into the campaign. Returning to our city watch campaign, a character whose dump stat is Charisma might not be such a good idea for example and its the duty of the GM to call the player on it and ask them to rethink the scores they've assigned.
This is why I think Session Zero is so important now; I've become acutely aware of the pitfalls associated with the human element of roleplaying. Never mind system queries and quibbles, what derails games is people (disclosure: I'm not proud of it but I've got my share of game kills). Whether the GM heads straight into the campaign without explaining what its about and gets the tone horribly wrong, or players design a character that are just wrong for the style of game, sometimes wilfully, lack of communication is the chief culprit to games expiring too soon.
Getting a good groundwork in early and keeping communication flowing are the only ways I can see to reconcile the three games into something workable. It's something that should be as natural as deciding what film to see or where to go to eat. But... we're dealing with gamers here and sometimes even the most determined efforts fall flat. Players can be frustratingly 'nice' and pass the buck, or feel that its not their place to suggest things because the GM's 'meant' to do the work or might use their ideas against them. In contrast GMs might not want to share ideas because it will ruin the 'surprise' or make things difficult, even when the most sensible course of action is to get players on board (and these days I feel that it is always a good idea to do that).
The fact that communication breaks down so much (and looking at the various killer GM and when players don't get it threads on fora, this is something that plagues the hobby) suggests that the hobby is still pretty adolescent, working around the idea that the GM is the person who puts in all the work rather than it being a team effort where everybody contributes equally. As time gets tighter and other media become more sophisticated (I'd argue that computer gaming is already in a position to make the traditional dungeon bashing campaign obsolete but that's another column), gamers will have to adapt and make sure that everyone at the table is happy with what's proposed before the first dice tumble.
* In the unlikely situation of me playing with that GM again I'll be playing a clever combat monster and throwing in non-combat skills as play progresses.
** If such things are important to you, I'm not super keen on them but at the same time, I can see the point especially in some games. Again, talking to the GM about whether a build is important is a good idea.