Ubiq wrote in his Zen of Design blog:

Someday, someone is going to make a solid purely social MMO, fulfilling the promise of TSO. This game, whatever it is, is going to capture its market the same way the Sims did, by trojan-horsing in on the early adopters, and then building it’s own hardcore market.

Without delving too deeply into the recent game vs. world discussion between designers, developers, and thinkers, I have to believe that in order to grab the players, it needs to look like a traditional game, but turn into a world as the users spend more time with it.

Consider World of Warcraft (WoW): make no bones about it, it’s a game. After six weeks, my wife is already starting to feel a bit fatigued and bored with it, as she’s essentially playing the same game with each new character, and with each level. Instanced dungeons and grouping are, for the most part, unwelcome distractions, and, at some point, even quest-driven games lose their ability to compel the player to continue working, and actively discourage the player from grinding.

Consider Ultima Online: this was a game disguised as a world. Sure, you could make changes by building and crafting, but monsters still spawned in the dungeons and at sea. It looked like you had limitless possibilities, and within the realm of bugs and some nifty designed features, you had a lot of activities you could perform. The mechanics remained, though: grind your way through skills. For my wife, UO was not terribly straightforward–the game didn’t provide a clear goal.

Consider, finally, LambdaMOO: this was a world disguised as a social space. Because it was “merely” text, the environment made it possible for users and developers to create anything they imagined: everything in a MOO is smoke and mirrors, words and thoughts. People came to chat with friends, and stayed because they could create their own rooms, decorate themselves, and present themselves in any way they wished. It’s lasted for years because it’s a social world, where the games range from the meta (programming your way out of a paper bag) to the literal (the MOO RPG).

To attract much of the current MMO audience, you need to present your system as a game, with goals. As part of the design, the players should be weaned onto the social and world-building aspects. World building helps create a sense of ownership. Social spaces create friendships. Both world building and social spaces can provide a way to retain (and attract) new users.

Some systems (such as Second Life) attempt to do just that: a world that you can build it, and that is specifically geared to be a social space, where games are created by other players. It’s a great idea, but it doesn’t seem to have had the market impact that a game like World of Warcraft has–Second Life didn’t market to the EverQuest/WoW crowd and didn’t manage to attract them.

Ubiq was right: to get a large network of dedicated social players, you need to bring them in through their friends, spouses, significant others, etc. The people willing to spend the money, spend the time, and take a risk on a game are the hardcore gamers. The people who are going to stick around for years are world-builders and socialites. Even in games like Dark Age of Camelot, players don’t hang around because they’ve bought house–players stick around because they’re in guilds and feel that leaving the game would mean leaving that social web. When WoW was released, many of those social webs were transferred, more or less intact, into the new game and out of DAOC. The game was old, but the social web was so important that people would continue playing just to spend time with friends.

A social game needs good social tools: customizable windows, distribution lists/channels, different ways of managing private conversations, etc. Consider what makes an IM client useful and port those features wholesale into the game. A social game also needs to recognize that there are out-of-game ways to communicate today (such as TeamSpeak and other VOIP applications) and indicate that the user is able to communicate that way as well. Finally, a social game needs accountability.

Accountability isn’t being able to call the GM or CSR to handle harrassment, and it isn’t allowing people to PvP one-another to handle griefing. Cory Doctorow’s novel Down and Out in the Magic Kingdom provides a perfect example of accountability in fantastic or non-material mileus: whuffie. Whuffie, essentially esteem, measures how strongly and well other people think about the character. Whuffie is given by other people when a character does something good or well that is appreciated, and taken away when the character is disliked or causes distress to others. Whuffie is a real-time measurement of what privileges and status a character has.

Whuffie is a wonderful way of making players accountable to one another: do something good (lead the group or raid well, transfer some gold from one character to another, craft a nice piece of equipment) and the characters can grant whuffie to one another. Do something bad (wipe the group, killsteal a mob, harrass another player) and the whuffie can be taken away. Designers and powergamers will immediately start thinking of ways to “game” whuffie gains. Go ahead.

Whuffie is used to “rent” things in the game. If your character wants a mansion, they can have it as long as they have high enough whuffie. If you wanted to get a horse, make sure you do good deeds for others and get whuffie to pay for it. Recovering from an unfortunate death can be a lot more comfortable when the character has whuffie to help pad the losses. Players should be forced to spend whuffie (and unspent whuffie should be able to decay, albeit slowly). Item costs in whuffie should scale to the total amount of whuffie in the economy, as well as to what the character can afford.

Players should also be encouraged to grant whuffie. A character with zero whuffie can still grant whuffie to another character, but large grants of whuffie should impact the character’s own whuffie by a similar amount. In short, whuffie grants should scale, and characters who have more whuffie can have their grant count more than characters with little whuffie. Finally, a character can’t grant whuffie to more than a few other characters within some indeterminate time frame, nor can they receive whuffie from another character more than a few times within another time frame. Whuffie grants should be tracked, to some degree, so that circles of whuffie gifting don’t spring up, promoting characters to higher levels of whuffie than they would have otherwise earned on their own.

There should still be a gold or cash economy to purchase items and exchange goods. Whuffie should not be able to be exchanged for items–I suggest simple “thumbs up” and “thumbs down” widgets to suggest the gifting and removal of whuffie from other characters.