Tales from the Rocket House
In games that are set up for a simulationist and immersive style (in English: the integrity of “what would happen” in the game world is held as more important than player challenge or “a good story,” and every effort is maintained to allow the players to get and stay inside their characters’ heads), it is incredibly important that the player characters be created carefully. Each character must have a reason to be proactive and involved (in this type of game, the “reluctant hero” who only wants to return to his family farm generally can just return to his family farm), preferably with goals and connections of their own that can be explored during the course of the game. Additionally, the group itself must have a significant level of cohesion built in. The characters must have a good reason to work with each other, and at least enough trust in each other to not immediately walk away.
I’ve learned this from experience. In a game where problem solving and tactics are the most important things, the player characters will work together because it is tactically wise to do so. In a game where the plot is the major concerns, storylines will sweep the characters along, again, allowing unlikely characters to work together. When the game focuses as much as possible on in-world and in-character authenticity, those connecting forces are absent, and it groups that have no reason to work together, travel together, or stay together fall apart, much as they do in real life.
This does not, however, mean that the player characters have to like or fully trust each other, nor that intraparty conflict will be absent. I’ve had player characters come to blows and almost kill each other, but it was many sessions in, and had been developing, in-character, for some time. It was also surprising, jaw dropping, and amazing: the kind of thing you can’t plan, and which wouldn’t be as much fun if you had planned it.
Of course, I’ve had other groups that worked together like well oiled machines. The problem with well-oiled machines is that it’s difficult to introduce a new component (in this case, a new player character)…
Introducing a New Character to an Established Group
In games with a more game- or drama- oriented bent, it’s easier (and expected, as a part of the [generally unwritten] social contract) that existing PC’s sort of see the “PC glow” and accept new characters into the group automatically, just to help keep the game flowing.
But part of the nature of, and attraction of, simulation and immersion-oriented play style is that the players really do get to stay deeply in character. This makes integrating new characters harder. The first thing to do is make sure the player of the new character understands that there is no “pc glow,” and so their PC will need to act accordingly. Entering a cautious group that has had (or very nearly had) members die, and which is involved in a juggling act of several potentially dangerous political and/or other factions, will require the new character displaying discretion, good judgment, and a reasonable degree of caution.
It’s important that a new player not be blindsided by this. To that end, it’s important for the GM (especially) to take stock of how the game’s going and how the characters think, what they value. This may limit the kinds of personalities that new player characters can have (though not as much as you’d think), but it definitely requires a certain degree of “engineering first impressions,” so the existing PC’s are more inclined to work with the new one.
Constructing a pre-existing link between the new PC and one or more of the existing ones can be helpful, so long as it is A) accurate and B) favorable. An old enemy isn’t likely to be welcomed into the group without plot manipulation (and plot manipulation isn’t allowed in this type of game). Certainly the new PC needs a driving and credible reason to work with the existing PC’s; a new PC should not be allowed into this kind of game without a reason to work with the party. But just as much, the existing PC’s need to see that the new PC is worth working with.
Conclusion
In the end, I think this is probably good advice for most gaming styles, but it is critically important in the style for which I wrote the Tarafore system. When the group dynamic has not been right, the resulting games were unpleasant and short. When the group dynamic has worked, however, they’ve been incredible.
It’s really the GM’s job to guide character creation so as to create a working group of player characters. The best way to do this, however, is through openness. Discuss it with the players, make sure they understand the importance of group dynamics, and help guide them through it. After all, we’re all friends here.

