Let's assume that we can break play styles into two categories: "selfish play," where you want to have fun, and "less-selfish play," where you want to have fun but you want everyone else at the table to have fun, too, and you're willing to sacrifice some of your fun to make the game fun for everyone.
What can be done to enhance the latter? Social contract? Tools? What can we as fellow players and gms do to bring out the latter?
Our very own Graham Walmsley has a book on this very subject: Play Unsafe. The core idea is to always build on others' contributions, never trying to block or frustrate the stuff that other players throw into the game. This creates a wonderful forward momentum where a lot happens fairly quickly, with high player investment and a large degree of the unexpected.
This core idea is developed in several ways (of which this is only a glimpse):
I've found that my gaming reflexes are hard to change since my attention is always taken up by plot, characterisation, problem-solving, or whatever else is going on in the game. So although I can briefly describe the core message of the book, the actual book is invaluable for steeping yourself in the right mindset to make these changes. Also, Graham just says it better than I can. I've read it several times, and each time it improves my play an increment.
The other nice benefit of this is that it is a personal suite of techniques that can be applied (almost?) regardless of social contract, so you don't need anyone else to start. Once you start, though, others will catch on.
To add to some already great answers:
Failure has to be "fun." If failure is not-fun, then players will do their best to avoid failure, and one's own failure is more not-fun than another player's. This avoidance leads to this "selfish" (self-preserving) behavior.
Let me try to unpack "fun" in this context. All outcomes need to be interesting. The game needs to be fun to play even when your character is failing at something.
If you succeed, something cool happens... if you fail, something cool still has to happen. This is really tough with traditional games derived from D&D... games with no shades of grey, which punish failure mercilessly. (Nothing like throwing down an awesome "one use per day" power and watching it fizzle... turning what could have been awesome into just disappointment and regret, because you had only one chance to be awesome in that way today.)
It's one of the reasons that PC death is generally off the table in all of my games, except when the player involved thinks it's dramatically appropriate. PC death generally isn't interesting and is rather not-fun, unless the player finds death a satisfying outcome in that particular point in the story. (Such as sacrificing themselves so the rest of the party can escape, and feeling this is an appropriate end to their character's story.)
When failure is interesting... when it leads to interesting story complications or interactions, then players are less afraid of failure. And players who aren't afraid of failure are more willing to take risks to help other players do awesome things. (This assumes that your players find story complications, etc to be part of their "fun". If not, then you probably have little hope with this group of players.)
So the most important thing I think you can do here is pick a game system that supports this kind of play instead of discouraging it. Players used to this approach can do it in D&D, but I think it'd be hard to teach people to think this way in that context, unless you're willing to tweak the system.
My biggest tweak when running "unforgiving" games is to introduce a "luck point" system, where the players earn tokens (often by accepting failure) they can spend to "fix" bad die rolls. (This is effectively what Fate does.) I've experimented with rerolls, +4 bonus, automatic success, and I find automatic success to work best... when the player really wants to succeed, making sure they can when it's important is key. I let players spend tokens on other player's actions... you could go one step further and make that the only option, so fixing a botched roll can only come from another player. This would emphasize the value placed on reinforcing others' fun.
But there aren't any really good mechanical fixes, and you have to keep in mind that you're trying to muck about with people's personalities... people play "selfishly" because that's the way they are, and they have to want to change if they're to play differently. You can't force that change on them. You can talk about it, and sometimes you'll find one for whom the idea of making others look awesome as a way of raising enjoyment for everyone is an epiphany. But gamers are competitive by nature, and some of them will never change.
The alternative is to surround yourself with the kind of people you want to play with to start with. This, too, is hard.
I think that selfish play comes from players mistaking the game for a competition. It's not entirely those players who are at fault either. Many GMs take on an antagonistic approach, treating the game as a matter of players vs GM. Even if the players see themselves as teammates, the competitive tone is still there if the GM is opposing them. I think the GM needs to set a tone of collaboration.
On an entirely different note, I think creativity should be encouraged as a way to enhance everyone else's fun. If a player is able to push something into the game, that has the potential to provide the other players with an additional entity to interact with.
That was vague, so here's an example. I just started playing in a Dresden Files RPG campaign. The system is based on FATE and gives players way more narrative control than I'm used to seeing. One of the ways players can narrate is by making checks to add details or aspects to a scene. So lets say your party is fighting a vampire in a skyscraper that's being built. You could use your craftsmanship skill to notice that floors are still being built and there's an unstable patch over there. On a successful check, you'd add that aspect to the scene and another player could later take advantage of it.
We haven't played enough for me to see how that works out yet, but I like it a lot in theory. It encourages people to build things in the game that other people can use. Even if you were playing a complete non combat character, you could spend the whole fight imposing new aspects onto the scene that other players could use to their advantage.
Vs. Selfish Play
Presumably, when you're playing a game together, everyone's there to have the same kind of fun. The places where we usually find gamers at odds with different types of fun usually points to the fact either they thought they were playing something different or they are ok with sabotaging the agreed upon game (see also "griefers" in online games... they're not there for the same kind of fun everyone else is having).
A common cause in rpgs is "My Guy Syndrome" where a player decides to do things that are unfun for everyone else, constantly claiming, "But that's what my guy would do!".
[From a Forge thread on dysfunctional play habits] on the topic 2:
Come up with as colorful a concept as possible, preferably somewhat irrational, so that you can carry out the following safety-measures from “in character” and blame the character for “making” you role-play in this way.
Encouraging "Less Selfish" play, or basically group fun
Once you get past those hurdles, the easiest solution is to try a game that has a "Fanmail" reward system - Primetime Adventures is the first game that utilized it, though it shows up in things like Tenra Bansho Zero.
The only way players get rewarded is by the other people in the group, giving "Fanmail" tokens (points, etc.) for doing things the group finds entertaining. These are given on the spot, not at the end of the session, so players can modify their actions during play as they get a better bead on what everyone is finding fun.
From the GM perspective, there are a number of tools I use to make the game more cooperative. We play a very social-heavy, skill based game, so this can really work. I am at the office, so this will be quick and dirty and less than comprehensive.
Break up the players into small groups with unified goals and unified in-game support systems. Our game is called GuildSchool, and all players have multiple guild/group allegiances...setting up cooperative small and mid-level goals, you enhance a lot of the game. I love setting up 'odd-couple' groups like this in town...
Whenever possible, I encourage the players to use each other as resources to gain access to guilds/churches/organizations. This includes a certain subconsious 'us vs the world' reinforcement.
Allow for additive skill use in your systems. This is a wonderful mechanical way to enhance and change gameplay. When One player is using a social skill like bribery and the guy behind asks if he can add part of his indimidate skill, or if a sage with mechanical understanding wants to advise on a difficult bit of trap removel, our system is set up to add up to 25% of the additive skill to the success % of the person using the skill. The system teaches them to cooperate. Since experience is only given in skills that are used, the players also get exp rewards for doing this.
In addition, I give a roleplay exp bonus at the end of every session. Part of this is based on working together. I only started this a few years ago, and it has suprised me the amount of competition amongst all my players in who gets the biggest Roleplay exp bonus.
I also dump alignment, as it is usually a within-party deconstructor. But you can de-emphasize if you prefer. Personal opinion.
This isn't normally an issue in the groups I play with, but occasionally it comes up when a player from a different gaming background joins us. The tack I sometimes take is to explain collaborative play as a test of one's own creativity.
Artists, programmers, designers, and people in other creative disciplines frequently talk about how constraints breed creativity. They will point at some of their best work and tell you it came not from being unfettered, but from having to reconcile two or more seemingly irreconcilable creative goals.
Maintaining your own goals as a player and providing goals for your character drive that feeling of fulfillment we all seek from roleplaying. If you treat the goals of other players not as roadblocks, but as useful creative constraints, the emphasis shifts. Instead of being forced to watch out for everyone else's enjoyment, you are testing your own creativity by working with the other players to create and reinforce constraints.
An example: The Paladin wants to find the evil dragon and slay it. The Thief is unconcerned or possibly even annoyed by the Paladin's goals, because they interfere with his own goals of sneaking in and getting the treasure unobserved. The creative constraint for the person playing the Thief is to make the Paladin's approach work to his advantage, to work with the Paladin's brash, straightforward attack support his stealthy entry.
Obviously there are times when this embrace of creative constraints runs counter to what a player who is used to thinking in tactical terms might think is the optimum choice. I work to minimize that by rewarding unconventional actions. Opponents frequently expect the "best" choice, and are often caught off guard by a more creative but theoretically less perfect choice.
With any luck, after a while it dawns on players that it's intrinsically valuable for everyone at the table to enhance each other's fun, and the trick of thinking of it as a creative constraint is no longer necessary.
This post is based around my experience as a GM and from my experience only
As I see it, there are three ways to change it: Rewards in terms of advancements (XP, Gear, etc), rewards in terms of bennies and the like and rewards in terms of compliments. While I'm strictly against rewards in the form of advancements, I really like the other 2 types.
An XP reward is a great tool for the short term. In the short term it will make the players happy that they chose a particular way of action, while also making them crave for more. I see with it 2 main problems that detract from it and turn it into a damaging gaming nightmare.
The first one is that it is not that different from bribing the players. Instead of promising the players money we're promising them to level up faster and more often. Personally, I have a slight problem with bribing my players as my sense of morality prevents me from doing so.
The second problem is that by teaching them that "good actions" equal XP rewards, we teach them that things that don't grant XP are bad things. This is far more problematic than it sounds because it a) prevents us from ever stopping rewarding the players, so no real progress has been made, and b) discourage players from coming with things of their own as we don't reward them for their style of play. Furthermore, we won't see those beautiful actions like the ones done by players who add a little sparkle of detail here or there, for example.
The little benny is a beautiful thing, both pretty and charming. When used right, it can reward in ways that are far greater than most other ways can, but when used wrong it can destroy stories. Although this danger, I still think that it is a great tool, when used right.
The most important thing to consider when adding bennies, as far as I can see, is that the players will have most of the power to grant it. I truly believe that there's no one who knows better when the players have fun than the players themselves, and as such they shall grant it.
Saying that, I use a little different approach, which is still quiet close, but not exactly the same: It's a system that I borrowed from a great indie RPG called Awesome which states something like that, when implemented into my games: "If at least one player think this was cool or fun or anything of the like, the action succeeds. If the entire table thinks it was cool or fun or entertaining then the action succeeds and another good thing happens."
If there is something that will make me, personally, less selfish in my gaming (and far better overall) it will be without any doubt a compliment. Be it from a player or from the GM, saying that something that I did helped to make the game for them will make me craving for doing that again. When they think that I'm cool, I'm cool and when they think that I help to make the game, it's exactly what I do.
In other words, try saying to your players when they did good stuff that it was cool or fun or entertaining or even all of the above. It will surely make a difference.
To that I can only add a seconding to Carl:
Failure has to be "fun." If failure is not-fun, then players will do their best to avoid failure, and one's own failure is more not-fun than another player's.
Making failures fun is one of the greatest tools in order to achieve this aim.
Normally I'm against this, but if you state explicitly that this is the purpose, having a system by where XP is awarded by other players on subjective grounds would be the way to go about this. Someone who plays selfishly won't get much XP, whereas someone who plays selflessly will get a lot more. You'd also have to ensure that XP as a reward enhances the fun of the player getting it.
