I wanted to jump-start a discussion of various point-buy character generation methods and get some feedback from the community as to which ones are most useful for fulfilling different design goals. Different game systems use different methods, and so it can get a little intimidating quantifying all of the different permutations of point-based character generation, so I thought I'd start up a list so everyone can share their opinions on the various methods. So here it goes:

1. Lump Sum Using this method, players are given a certain amount of points all at once and given free rein to distribute them amongst a character's attributes skills etc. as they please with no restrictions. This method has the advantage of allowing the greatest possible variety of characters, but risks imbalance by allowing players to pump all of their points into the most cost-effective areas. This method also has the disadvantage of preventing a 1:1 currency exchange rate unless all character abilities are equally useful/powerful (a tall order for most games with both attributes and skills). To my knowledge only GURPs and Champions use this method.

2. Lump Sum with Caps This method is a variation of the first, giving players a certain amount of points all at once and then setting restrictions on either the maximum value of traits that can be bought or on the points spent on certain traits or (rarely) both. This has the advantage of improved balance over the Lump Sum method by discouraging min-maxing but decreases flexibility and can result in characters feeling the same if there is a universal cap on traits and no way of manipulating the cap. This method also shares the problem of the first, in that it pretty much prevents a 1:1 currency exchange rate. Mutants and Masterminds uses this system, as does Shadowrun 4th edition.

3. Priority To my knowledge, the storytelling system games (OWOD/NWOD, Scion etc.) are the only ones which use this method which gives players a choice of priorities, usually expressed as "Good" "Fair" and "Poor" or "Primary" "Secondary" and "Tertiary". These priorities might be assigned to any given category of traits, for example Physical, Skills, Magic etc. depending on the demands of the system. This system has the benefit of encouraging character's with well-defined strengths and weaknesses, but impedes flexibility and can prove complicated for games with many categories of traits.

4. Directed Investment with bonus No system I've played has used this method, but it seems like an interesting idea. Using this method, players receive a certain amount of points for each category of trait and then receive a pool of bonus points that they may use to boost any trait. This would have the advantage of maintaining a reasonable level of balance and flexibility, but it's very much a compromise in that it doesn't allow nearly as much character differentiation as either of the Lump Sum methods. It also still has the problem of not preserving a 1:1 currency exchange rate unless the difficult task of making sure all areas of a character are of equal importance is accomplished.

I'd love to hear your thoughts on which method you feel allows for the best combination of flexibility in character generation and balance, and to hear any additional permutations of point-buy character generation that you have thought up or discovered yourselves.

1. Lump Sum - this works best for short campaigns where characters are created "as is" rather than developing over time.  Luke Skywalker is young but has potential, Obi Wan is old and experienced.  You can build and balance those characters (not that Luke and Obi Wan are equals) because the young guy has great stats, the old guy has great skills.

2. This is the worst in my opinion.  Caps means you don't think it is balanced but just want to limit the damage.

3. no opinion

4. Seems intriguing.  Like #2 the idea is to limit the damage of the imbalance, but seems simpler and more direct.

I like:
diminishing returns.  Either a scaling point costs ala D&D or a reduced benefit for increases ala Savage Worlds
limited points in attributes.  Attributes define the potential of the character, they gain skills and such but their attributes are fixed.

I don't know how exactly to phrase this but I don't think the various approaches have very much innate difference between them. It all depends on what you want. Should players be competing for the most effective point combos? Should the available point options make sure there characters stay within the parameters of the source material? Should the way points are spent flag what kinds of things the players want to be a focus of play? All of these things might or might not be design goals, and each of them would require or suggest a different way of approaching point spreads, competition and simulation recommend themselves to tighter controls, for instance. So what do you want the points to do, in the end, besides the obvious things?

John,
  I have played with some of these, I can relay you my experience:
Open Point buy works well, as long as three conditions are met:
1) The player has a solid idea of the character they are making
2) The player has enough knowledge of the system to buy all of the "essential" attributes
3) The adventure has a place for ANY character made.

Point buy with caps works a lot better in my mind. A better example of this is BESM, it still relies on:
1) A solid character premise
2) Knowing the things every character HAS to buy
3) Player and GM meeting half way about what kinds of characters the current adventure supports

Point buy in Pools is more like what White Wolf offers. I think this is a great way to make characters with some flexibility and in a very directed fashion.
1) With this sort of system, you can discover the character as you build it
2) this can also make sure that you are guided to get most of what you NEED.
3) This hitch is still there, but with teh direction provided by the various point pools, it is easier to correct a mistake...

I haven't seen a game that uses full-on priorities. I know that White Wolf Stats (Str, etc.) do this, but then that goes to the wayside after stats are determined. I know some designers were talking about using a priority system on all their Stats, skills, etc. But I don't know if those rules are available, or how they play.

  In the end, the danger of point buy is two-fold:
1) Susceptible to min/maxing. More so the more different things the players have to buy.
2) Susceptible to underdeveloped characters (High Str and no Athletics skill. High Ranged Weapon Skill and no Perception, etc.).

  In the end, Point buy works best when there are less things to choose from to buy. Like the less skills needed the better.

  If you can build around these weaknesses, I think you will find that the strengths of Point-Buy are supremely worth it.

  Why do you ask (if you don't mind the enquiry)?

Well the reason why I started this discussion is that I feel that all too often game designers simply pick a character generation system (or particular subtype of character generation) pretty much at random without consideration of the strengths and weaknesses of each method. I also feel that codifying these strengths and weaknesses might prove useful to game developers from a theory perspective.

Masquerade,
I feel as if regardless of the type of game one is shooting for (whether narrativist, simulationist or gamist) one should try to ensure that the character generation system is balanced with regards to character effectiveness. Every system has some measure of a character's ability to affect things, the only variation is in what measure is used and what sort of effects are possible. And generally speaking it's not fun for a given player to have less ability to impact things with his character because the character generation system is open to exploitation. I'm curious though, since you mentioned generation systems with other goals, including serving as a way for players to state what sort of play they enjoy. What character generation system do you think would best support this?

Dug,
Thank you for your opinion. My thoughts on caps is that they can be done well if done in a limited fashion (as Shadowrun 4th edition does, by limiting only starting skills and attributes, as well as points spent on attributes) or if there's some way of manipulating the caps (as with Mutant's and Masterminds tradeoff system).

Any character creation system can be seen as a way to generate flags. Even D&D, if I put my skill points into Sneak that's a way to highlight that I want to be able to use stealth in play. As for effectiveness, or balance, this is not always a goal. MERP and to a lesser extent, the newer (Decipher) LotR, makes the Hobbits just plain out worse, and from a certain Sim perspective, this makes sense, and I've known people who enjoy the challenge to being less effective than other players. Another example would be Palladium's Heroes Unlimited, where random rolls can create massive differences in player's ability. This is okay, imo, because it reflects the way different heroes are in various comics, and because other factors are suppose to control player involvement/participation. I would also argue that various point buy systems are specifically designed to allow for variations in power between players of the same point cost (the White Wolf games, BESM) to better allow players to highlight whats important to them.

masqueradeball wrote:
Any character creation system can be seen as a way to generate flags. Even D&D, if I put my skill points into Sneak that's a way to highlight that I want to be able to use stealth in play.

Thats why I said "may." Unless the game system specifically says this is what the points do, people may do it to optimize their characters or just to get rid of points or because its a matter of course (this is D&D, everyone should take Stealth, or whatever).

I might not be understand you right, but here is what I think:

We used to use a limitless point-buy system, but we found that over time it was nearly impossible to balance challenges if anybody wanted to min-max one part of their character or another.  The acrobat would be the only one to cross any kind of an environmental challenge, the super-offensive character would be the only one with kills, and the tank would be the only one that could survive serious hits.

What we came up with was a very carefully tuned point-by system, with a progressively advancing level-cap, something like the way 3.5's skill system was built.  A starting character could have at most 5 levels in any skill, but once they got past x number skill points total, the cap moved up to 6, and they could continue advancing their favored skills.

What we've seen is that people choose different skills, but there is a set competitive level that people can expect to need.  If you plan on using a skill, you can have fewer levels in it, but if you want to be bleeding-edge competitive, you max it to the cap.  Once someone hits their cap on their favorite few skills, then they can branch out and try something else, adding some side-capabilities, and then when the level-cap advances, they can go back to their favorites to stay competitive.

It's worked out pretty well for us thus far.  It's important in this kind of a system to have several different ways to accomplish similar ends, but with dramatically different flavor.  We have Melee, Shot, Thrown, for combat offense, as well as some of the magick-ey skills, and we have dodge and parry for core defensive skills, with autonomy and grit for mental and physical defenses against magic and other similar effects.  Parry works better against Melee, but Dodge is better against shot or thrown.  Parry works against shot or thrown, but at a significant penalty, and same with dodging melee.

What happens is that people choose their own route to effectiveness, and stay pretty competitive.  They have strengths and weaknesses against their specific strategies, and enough of their own personality to keep things interesting and diverse.

So to make a short story long, the level cap works well, but only if there are good options among skills so that there's not the one 'essential' skill that everybody absolutely has to have.

I do like the M&M trade off system.

I wish D&D 4e and City of Heroes would have something similar.  I like the idea of archtypes, but I wish they were not built into the classes but a trade off.  Get +2 AC and do -2 damage sort of thing.  I like the idea of characters being balanced but choosing to be balanced just a bit differently than everyone else.