Social Software & Gaming: User Content
by Shannon Appelcline
No sphere of study every stands on its own. This has been one of the central focuses of this column, as I've talked about mythology, movies, comic books, fiction writing, and many other fields, all with the purpose of seeing how their experiences can expand upon our own in the field of game design.
This week I want to delve into an entirely new field of study, which I know will offer us new lessons learned: Social Software. Social Software is, in some ways, a foundational ideal of Skotos. Christopher Allen, the founder of Skotos, was involved in Social Software (then called GroupWare) through his previous company, Consensus Development. Today he still writes on current trends in social software through his blog, Life with Alacrity.
The ideals of Social Software are pretty simple: using technology to manage, improve, and facilitate human interactions. One definition by Clay Shirky simply says that "Social Software is software that supports group interaction."
The classic Social Software is probably Lotus Notes and its brethren, wherein multiple people became able to contribute to a central database of information through distributed workstations. Nowadays a new type of Social Software is making a strong appearance--social networking software. These are the annoying programs like Orkut and LinkIn that constantly generate messages about how you should join one group of another to connect up with your network of friends. Though different in purpose, social networking software tends to have the same core ideal as Social Software of all sorts--managing interactions through technology.
One of the core problems with Social Software is also a core problem in the online world: managing user content. As games grow, particularly as we turn to more user-oriented creativity models, as I discuss in Trials, Trimumphs & Trivialities #67, Creativity & The Online Gamer, we'll be increasingly faced with the problem of how to make sure that good content is made available & bad content not. Ultimately this means adopting specific models for how to publish reader-submitted content.
Thus, this week I've looked at how Social Software addresses this problem by investigating three different asynchronous Social Software models, explaining each of their solutions, then considering how that would work in the online gaming world.
Social Software The First: Forums
Everyone has probably interacted with one of the simplest types of Social Software: the forum. Here at Skotos we use vBulletin software, which is a pretty robust & well featured forum program. Forum software itself is an evolution of traditional bulletin board software, USENET newsgroups, and mailing lists, none of which have the same importance as they once did.
The idea behind forums is the simplest type of Social Software: a venue for readers to converse on like topics.
The model forums use for managing user content is moderator disapproval. In other words, anyone can submit anything, and moderators only step in if something troublesome is said, and then only to remove those troublesome items.
This sort of model is great for allowing for the dynamic, free exchange of ideas in an asynchronous, yet real time format, but the biggest problems (spammers, troublemakers, harassment) are innately caused by that same format; because anyone can post anything, anyone does.
Skotos avoids most of these problems because of its size, but over at RPGnet the forums only survive because of the dedicated work of a team of volunteers, ready to sweep down and beat the hell out of a malcontent (or at least delete their posts) at a moment's notice. (This all leads to a whole other topic of discussion, Dunbar's Number, which says that groups over 150 start to break down, but that could be the subject of a whole other Social Software-gaming crossover article.)
The other problem with this model is often a low signal-to-noise ratio; though the conversation is all fun and interesting it sometimes doesn't convey as much information as other models can.
Social Software The Second: TvTome
TvTome is a good example of another model of user content management. Here you have guides for a number of different TV shows, each one containing information on all the different episodes.
The model it uses for managing content is a combination of editor input and moderator approval. On the one hand, a dedicated team of volunteers sketches out the skeleton of each information. Then, individual readers can submit additional information within specific topics.
I often watch TV shows on DVD and when I go back to TvTome after the fact it's often a very good resource, providing information about individual episodes that I hadn't been aware of. However it's less useful for brand new shows as they air because there can be a real lag betwen the episode showing and posts to the guide getting updated. I often finish watching an episode of Lost a few hours after it airs, for example, rush upstairs, and find just a couple of sentences on the show.
The signal-to-noise ratio is much better than for a forum, presumably because readers are explicitly trying to provide information, not just conversation, but still bad things slip into the guide. I often find the worse to be "know-it-alls" who try and show how knowledgeable they are by pointing out "goofs" in each episodes--most of which aren't actually mistakes. This type of false information being introduced into these databases points toward weak or poor editors, but also shows an innate problem with content editing being done by singular people who aren't as experienced (or hardened) as a profesional editor would be.
Social Software the Third: BGG
One of my favorite sites on the 'net is BoardGameGeek, a database of board & card games, complete with articles, reviews, session reports, links, pictures, and lots of other information for every single game.
BGG started off with a similar model as TvTome, based largely on moderator approval, though some types of input are actually free approval, meaning that no one has to OK them. As a result their database has grown at an enormous rate. Personally I've submitted 42 pictures, 3 session reports, 20 "articles", 2 games, and numerous other items to the database. (As a side note, BGG offers another interesting mechanic: when you submit certain types of information you're rewarded with "geek gold", which can then be turned in for avatars and other rewards. This type of reward structure is probably yet another topic for another day.)
However, this database growth has caused the same problems as seen in forums and at TvTome, a slow degredation of the site as good information becomes lost in bad. Thus new ideas have recently been introduced: reader rating filtering. Now readers can actively rate a few sorts of information on the site, and as a result the good ones rise up and the bad ones disappear.
Unlike the problems with moderators seen at TvTome (and BGG too, as it happens), the entire community is more likely to have time to give careful consideration to any individual item, and their group response is more likely to be in tune with the community as a whole.
With the introduction of these initiatives, info on BGG is slowly becoming more useful again, though it continues to face another problem: user lassitude. We actually have rating tools in our own forums too, but they aren't used that much. There's the same problem at BGG: some types of information are getting more rating, but others appear and are rated just a couple of times, not enough to give a meaningful response.
(Of course one of the definitive web sites, slashdot, uses a similar method; I've heard that one called reputation filtering, and their community is sooooo large that items do tend to get rated a lot, at least until a thread starts to peter out.)
Social Software & Games: Redux
Player creativity is definitely one of the ways that online games need to go to expand. Games like Second Life have clearly realized this, and built success upon the idea. It's truly the only way that sufficient meaningful content can be created for a voracious audience.
However, the more user content that we allow into games, the more troubles we have separating the good from the bad, the signal from the noise. As we introduce user content systems into games, filtering it will become the next challenge. Fortunately there are already some answers in our nearby field of Social Software. Moderator Disapproval, moderator approval, free approval, and reader rating approval aren't necessarily the final word, but they're definitely an introduction to how another industry has already dealt with this issue, and how we should start thinking about it.