Here’s a thing that I tried in The Devil, John Moulton and Medical Bay Three. Creating a character works like card drafting in 7 Wonders or a booster draft Magic: The Gathering tournament. Each player gets a hand of cards, each of which has a piece of background information on it and some mechanical effect for the game. You look at your hand, choose one card and pass the rest to your left. Then you look at the hand that was handed to you from the right and choose another. Thus, each PC acquires details and information as you go in a semi-random manner. (When you choose between the last two cards, you discard the one you didn’t pick. That way you never are forced to take a card you didn’t choose.)
Why do this? Because it has a few advantages, some of which are obvious and some of which are more subtle.
It solves the blank page problem by providing obvious fictional prompts. You can sit down at the table with no character idea and start receiving cards. As you proceed, you learn about the PC bit by bit and construct an image of him or her in your head. One playtester noted that he rejected later cards based on fictional ideas he had forming based on earlier cards.
It creates characters that are guaranteed to have specific aspects, but are still surprising and interesting. If you put specific details on cards, you can be pretty sure that someone will pick that card sooner or later. If you want a group of noble knight who struggle with temptation, then you might make a quarter of the deck temptations that the knights have to face. And you know that those character elements will appear in the game, but you don’t know for sure what the combinations will be. Characters are made unique by taking two or more cards and explaining how those pieces fit together. The player’s imagination finds ways to fill in the gap between the cards, making each character unique even if each piece of the character isn’t.
Card sets can be manipulated and customized. A GM could reject a card and pull it from the deck if they don’t like it. Good for them. Or an expansion/supplement might add new cards in character creation. Or you could replace the character creation deck with a separate deck entirely to completely rework the game’s initial starting point.
You can control how many people do one particular thing. This could be mechanical matters, such as being the most skilled in swordplay or something. Or it might be a fictional matter: if you only want one heretic cast out from the church in the party, then you only put one such card in the deck. In a superhero game, you could guarantee that each hero has a unique set of powers rather than having two players step on each other’s toes. Or you could have a character aspect be really common: maybe there are seven dwarf cards, and only one exiled princess card.
Cards can hold a bunch of information. Cards in The Devil, John Moulton have a piece of background, a leading question for the player to answer, a suit and a mechanical effect. You could probably increase the information on there if you wanted without problem. Or you might divide the card up between opposing bits of information that you have to choose between by how you orient the card. This much text would be clumsy to work with for dice and lookup tables (Though Fiasco does so pretty well.)
Establishing tone and genre expectations. If you want a very specific kind of background or ability, you can define it on the cards. For The Devil John Moulton, I had a pretty specific idea for what abilities the demons grant the PCs. And a few of the cards provide example of demonic pacts that illustrate the tone and style. Now, even when the players don’t choose the card, they still read it and internalize a bit of what it gets across. If acts as an example of what I’m going for, but in a quiet, seamless manner. After drafting cards, players have to author their own demonic ability, and so far the players have created new abilities that fit the creepy weird vibe the examples showed them. Yet I didn’t have to describe that style or tone to them directly at all.
Secrets are interesting here. If I get a card with a bit of secret information and pass it on, then I’ve still read it. And I know that a PC is likely to have that secret, but not guaranteed (the card might have wound up discarded). So I’ll be looking for the PC who is actually looking to betray us, or being blackmailed or what have you. If you’re making a game about players secrets, then this could be a really interesting and fruitful technique.
You have many options to control how drafting works. Different drafting mechanics will have impacts on the decisions your players have to make. I’ve been working with individual hands for players and hidden decisions. But you could also have all the cards visible to everyone and choices made in public, so that players can use that information to influence their own choices. This could be important if, say, you have a superhero team and want a spread of abilities. You could have specific suits or kinds of cards and demand that each player choose at least one of each. Or not let a player choose more than X number of one kind of card. Or let each player choose one card from the entire deck before drafting begins. Daniel Solis suggests a card game drafting mechanic where each suit of cards is shuffled into a separate deck, which could easily work for games where each suit is race, class, background, superpower, etc. http://danielsolisblog.blogspot.com/2014/02/solving-design-problems-by-reducing-not.html
Randomness, then player choice, rather than vice versa. The random aspects for the player to make a meaningful choice, instead of the random aspects decreasing the player’s interesting choices.
Scalability. By having a larger or smaller starting hand, you can control the amount of detail a character gets. A simpler, more cartoony game might start with a small hand and broad archetypes on each card. A more finely detailed game might have a much larger hand, and each card gives you a specific skill, feat or special power. My current project ties the cards into Cthulhu Dark based death spirals and die sizes, so the hand size determines campaign length as well.
Players act simultaneously so this phase of character creation goes pretty quickly.
Specific to the game I’m working on, it reinforces themes about difficult choices that you may not ever be happy with. You have to choose one of the cards in your hand, but you don’t get control over what cards are there. So you might have to pick something you wouldn’t go for otherwise. Your character, ultimately, is the result both of choices you controlled and outside forces that you didn’t. Just like real human beings, you don’t decide the hand that life dealt you, but you do get to make some choices about how you deal with that hand. (This may or may not fit the theme of other games, of course.)
Any mechanic or game dynamic will have some limitations and drawbacks, of course. In this case, the system can cause some problems for a player who already has a character concept in mind. Similarly, the thematics of the decision making might not fit a game where it isn’t about making difficult choices in reaction to a world you didn’t create. For certain kinds of power fantasy, the difficult choices might not work well. Also, I haven’t figured out yet a use for the deck once play begins, which is a bit of a flaw. The cards are full of information that only relates during character creation. Possibly if they were labeled with playing card ranks and suits, or Tarot style art then you could reuse them. And selling a roleplaying game with custom cards is probably more difficult than selling a game that just uses regular cards or your standard gamer dice.
I’m sure that there are a host of other problems that other people can see in this idea that I’m failing to notice right now.