The conflicts of virtual football players

The site I’m studying, Goal Line Blitz @ http://goallineblitz.com, is a massive multiplayer online role-playing game (MMORPG). There are different kinds of interactions that can take place, so different rules have to be made for each part. The general guide and FAQs for the site are @ http://goallineblitz.com/guides.html. In the forum section, there are rules about conduct in the forums @ http://www2.hawaii.edu/~karhai/ics691/glba2.tiff. There are also sub-forums, each with potentially slightly different rules. There are general guidelines that are common across many communities, e.g. limiting hate speech, name calling, etc. Since it is also a game where strategy plays an important role, there are also rules against publicly posting private team strategies or private messages in public forums (something that a player angry at his team or owner might do), or colluding with other teams to steal money, etc. Most of these rule-listing threads are “stickied,” so they remain at the top of the forum listing, and gives abusers no excuse about not knowing the rules.

There are different levels of severity when rules are broken, and the reaction by the administrators is usually applied accordingly.

Example 1: http://www2.hawaii.edu/~karhai/ics691/glba3.tiff (warning, lots of expletives). My first example is a player who is complaining about his team and his owner. These types of posts come up all the time, and the reactions can be very polarizing, with people taking extreme stances on each side. In this case, the admins quickly locked this thread, although from viewing the user’s profile, it seems he is still playing the game everyday.

I think it was a good reaction by the admin to stop any trouble before it starts. It was obvious in this case that the general population was against the original poster, and his barrage of insults didn’t help his cause. Without getting into the details of the game, the player ended up in a bad situation – but only as a result of his own actions – and then posted it in the forums just to complain. It was clear no one sympathized with him, and the responses he elicited prove as much.

I think this would make a good example of using community oversight. The responses to the original post were unanimously negative, and there was no one who supported the complaint the original poster made. While this is not exactly parallel to Cosley’s MovieLens example, it shares the same push toward a Web 2.0 method of relieving administrative duties by delegating to the masses.  In the GLB site, there would be no literal contribution made; the contribution would be the policing of the forums to prevent hostile threads from popping up. I believe this would greatly add to the value of the discussions, and increase the sense of community to the whole site.

Example 2: One of the previous general managers on my team got really frustrated trying to recruit players and posted a thread in the player-recruiting forum loaded with profanity. He was subsequently banned from the forum, and the admin sent him a private message (screenshot @ http://www2.hawaii.edu/~karhai/ics691/glba1.tiff. I could not find the actual thread he posted in, it might have been deleted). He responded to Rick (the admin), apologizing for his posts, explaining his frustration, to which the Rick accepted his reasoning and after a week, removed his ban.

I think this was another good admin response. The admin could search the forum for the user’s previous activities, and he could find any history of rude forum behavior. Not finding any, banning him for a week is a fitting response to the player’s outburst. Handling the situation in private messages instead of the forum also helps prevent the player from being embarrassed, and easily defuses the situation.

It’s interesting that the different sub-forums have different levels of rule enforcement. In some of the forums focused on a particular league, there is a lot of trash talking, but since most of it is generally meant as rivalries, the “no insults” rule is very lax. Kollack’s USENET example (from 1996!) shows the importance of virtual group boundaries with respect to behavior. And the “kill file” that was limited to USENET is now a readily available feature (re: user ignore) for any forum. Does having all the different sub-forums help? Especially if an admin can pick and choose what rules to enforce in each one?

Example 3: My last example is less of an interpersonal conflict, but with the game itself. Since the community is based around the game, I wanted to give an example that showed it. http://www2.hawaii.edu/~karhai/ics691/glba4.tiff is a response by the game creator about a cheat/exploit in the game. It was a topic that had been discussed in public, but only as a rumor without any actual confirmation until the admin posted that announcement. Afterwards, we (the public) checked on the profiles of some of the players suspected of cheating. Most were still playing the game as normal, but there were a handful whose accounts no longer existed.

Again, I think the admins did a fair job. Usually, most of the admins on the site can manage any conflicts or disputes, but this was a game-wide exploit and reached the game creator’s attention. He was more lenient that I would have guessed, giving the cheaters a chance to come forward before banning anyone from the site. Being banned from the site is a serious reaction; many parts of the game cost real money to play in, but that money is not all used at once. Any player could easily still have money deposited in the game that he would no longer have access to if he were banned. There are definitely legal ramifications that could arise from these actions. After the incident was handled, the public continued to speculate about the lack of “noise” from the players that were banned; there were continued rumors about the site creator quietly resolving the situation to prevent public outcry from any banned players.

The players that were initially suspected by the general public were Dibbell’s griefers, posting things on the forum that would ruin the game for others, but skirting the admins’ wrath by not explicitly breaking the rules. Bragging about their players and teams, they celebrated “success” whenever a team disbanded, or a player quit the game. They even hinted at the exploits they were using, claiming it was just “part of the game.” Eventually, the game creator stepped in and basically nixed all the exploits he could find, and banned those responsible. Arguably, the banned players could claim they had done nothing illegal, or against the game’s rules. But Internet, in this case, is very serious business, and it was obvious that banning a handful of players (and possibly dealing with them quietly via back channels) was the best solution to keep other members playing (and paying).

Example 4: http://www2.hawaii.edu/~karhai/ics691/glba5.tiff. This one is pretty self-explanatory. A user created multiple accounts, most likely to gain in-game points through referrals. We do not find out what happened to the user and his many accounts, but most likely all of them were banned from the site.

I’m not sure if this fits under “Interpersonally exploitative” in Gazan’s rogue behavior list. There are certainly users in the site suspected of having multiple accounts just to hide behavior, but in this case, it seems more likely they were trying to exploit the game’s referral points system, considering that this user didn’t even try to use a completely different name. This lets me bring up a point I had regarding the list of rogue behaviors: how does one correlate action with motive? To be fair, I think the examples on the chart are very clear and I would make the same conclusions. But as the referral cheat shows, for better or worse, there can be different motives for conducting rogue behavior that would exempt the user from being a rogue.

The secret identities of virtual football players

Your online identity is a big part of any interaction you will have on the web. It identifies you to other people, and serves as a basis for any common ground. For my working definition, I am studying the site http://goallineblitz.com (I’ll refer to it as GLB), a web based MMORPG built around playing American football simulations. I joined the site a little while back when we were studying social capital. I’ve since been more active on the site, and have a general idea about the population, codes of conduct, etc. The users on the site are very loosely tied together; they come from all over the world (although the majority seems to be from the US), and their knowledge of American football ranges from none to people who used to play competitively. The percentage of participation is very high; the game requires interaction in order to succeed, and so everyone must be involved. Team owners, general managers, and players must all work together to build a successful team. However, users not only interact with their teammates, but also with rival players. This can be quite different, as there is a fair amount of flaming on the public forums (team forums are private). Additionally, each user controls at least one, usually more, simulated football player. Users build their players in a similar fashion to many RPG games, they give them strength, speed, and other attributes. Based on these properties in GLB, a user’s identity is based on their background, forum interaction, and how their virtual athlete performs.

There are some similarities and differences with Ploderer’s study of bodybuilders. Unlike users on bodyspace.com, GLB users do not share any common ground on why they joined the site, there is little emotional support for each other, and it has no connected offline passion-centric activity. But there are also some similarities; almost all interactions are online, with few offline interactions between users, and users are judged based on developing a particular component (their bodies, or a virtual character), using it in competition, and watching it improve over time.

Compared to Hodkinson’s study of Goths, there are mostly differences. The Goths transition into LiveJournal gave each one of them a voice and presence. An individual blog site is much more than the simple profiles available to GLB users. Goths are united by something more than bodybuilder passion – their common ground is an entire subculture. The friends on the Goths LiveJournal pages are mostly other Goths, whereas GLB friends are usually their teammates.

I’ll give two scenarios of users in GLB.

User Stephenthegreat (owner of a team in GLB, and also has a player on his team) logs into GLB. He checks a public forum where people ask questions about their players, and sees a discussion where someone is asking why their linebacker is not playing well. He responds giving him a suggestion based on his own players. Knowing there was a game last night, Stephenthegreat checks and sees that his team won a close match against a division rival. He makes a congratulatory forum post in his team’s private forum, and then a “good game” post in the public forum. He sees that the other team has already made a thread about the game, commending him on a good game plan. His team has another game coming up shortly, so he goes back into the private forum and makes another post with the next game’s plan. Finally, Stephenthegreat logs off.

Stephenthegreat’s public forum post, helping another player with their build.

Stephenthegreat’s private forum post, advising his teammates on their settings for the next game.

User Joseph Fernwright (owner of two GLB players, on the same team) logs into GLB. He checks the statuses of each of his players, and notices that one of them, the quarterback, has gone up a level last night, giving him extra points. He reads the quarterbacks forum and determines that his strength attribute is low, so he uses his new points to increase it. Then he posts on the team’s private forum informing them that his quarterback is stronger. His other player, a running back, played in a simulated game and won the day before. He “watches” the simulated game, and sees that his player did very well. He goes into the public forum and finds the discussion where a member of the opposing team said his running back would do very poorly because of his poor build, and that his team would lose. He makes a response, “laughing” at the opposing teams, and makes a sarcastic brag. Satisfied his players are ready for their next games, Joseph Fernwright logs off.

Joseph Fernwright’s private forum post, starting a thread to collect information about special awards that can be shared between teammates.

Joseph Fernwright’s player, the green bars on the right tell us that his build is very good.

Joseph Fernwright’s public forum post, sarcastically insulting the players on the opposing team.

So now the question: How are online identities shaped and expressed through online interactions in this community? Well one thing I was curious about was Honeycutt’s study on Twitter, and her focus on using the @ sign for messaging. The method applies to a message meant for a specific person. However, it is also publically readable, so it’s meant to be “heard” by other people as if it were a public conversation. Twitter also has a private messaging feature, but it’s almost a secondary option; why bother clicking through several times when you can just type “@name” in the main screen? On the other hand, would users use it even if the private message option were also on the main screen? In other words, did the technology steer people to use the @ sign, or would users have preferred the “public conversation” anyway? In GLB, a user’s identity is partly shaped by their interactions in the forums. But since there are private team forums separate from public forums, people can and do act differently in each. Someone who is polite and cordial in the team forums can be the same person who is an instigator in the public forums. Their interactions give them multiple identities, depending on whom they are presenting themselves to.

The mass interaction in GLB is also different than those in Whittaker’s Usenet study. There is much more participation by the players, mainly driven because GLB is not a place people go purely for information, and where lurking can be beneficial. Cross posting also occurs a lot because users often have players in multiple teams, which play in different leagues. Many times, discussions in one league forum will be very similar to discussions in another league forum, since the game is still the same; GLB is still the common ground for any discussion. (On a side note, I had a small problem with how they measured thread depth in that paper. There is a figure that shows an example of a discussion thread, where most replies are usually a reply of a reply. But they completely ignore the interface available to users. Where is the reply button? Many times on forums, it’s just easier to click the last reply button, even though you’re responding to a post several messages earlier. Newsreaders were not very complex, back in the day, and I would not be surprised if the technology at the time steered people toward hitting the “last” reply button.)

GLB is certainly one of Wellman’s “more sparsely-knit communities” that higher use of CMC has afforded. There are very few strongly linked people in the site, and the users come from all over the world. Their only common ground is the game itself, and only use the site as such. But it is only in a networked society would the game have succeeded.

In the scenarios, Stephenthegreat’s actions make his identity “an experienced, smart, and classy person.” He actively participates in discussing strategies publically, and will respond to complete strangers when asked about it. His successes, both in his team and in developing his player, are clear to everyone, raising his status in their eyes. His interactions in the forums, both private and public, are always polite and humble, never making fun of people or pointing out his achievements.

Joseph Fernwright’s actions identify him as “a user with good players, but never contributes and can be a jerk in the forum.” He has never posted to help other players with their development, even though his players are successful. He makes it quite clear on the forums, both public and private, that his players are very good. Whenever his team wins, he makes sure to gloat about it to opposing teams.

online recommendations, real advice, my oopsie, quantity v. quality

I chose to compare online social recommendation systems vs. real world advice seeking for this week’s assignment. In doing so, I found the first complication of Dr. Gazan’s suggestion to complete the readings in order. I generally like to markup my readings, and also type my ideas out as I’m finishing each paper. I went through Duguid, Haythornthwaite, and Geisler before I realized I had a problem. I assumed “online social recommendation systems” were related to reviewers’ recommendations because that was more closely tied to the comparison with real world advice seeking. As such, all of my ideas and text were based on that. Then I reached the Lerman paper, where I realized my mistake, and the term actually referred to different filtering mechanisms, like ones that Amazon and Netflix use to offer suggestions such as “people who bought your item also bought these other items.” I didn’t want to delete everything I had already written, so I thought I could just compare both online systems (recommendation and reviewer) against real world advice seeking.
I have some biases going into this, based on personal experience. For example, I’m currently trying to purchase a new digital camera. There are tons of review sites, and online retailers will have a section for user feedback. But which sites do you trust? Which reviewers do you believe? Offline, I can ask friends directly about their experiences with cameras, but how many people do I know with a particular brand? Will I even know someone with the model that I want? Ok onto the readings…

As user reviews, online social recommendation systems fall under Haythornthwaite’s lightweight peer production (LWPP). Most sites that allow users to write about a particular product have minimal rules, low entry and exit costs, support anonymity, and require no history of contribution, each a property of LWPP systems. One question that arises – both in the paper and in my mind – is what the motivations for individual contribution are? Offline contributions would immediately reward the friend/reviewer with satisfaction in helping a friend, and an increase of how much knowledge they have in the eyes of their peers. Trust in the relationship has been built up over time, and I can determine how much weight to attribute to his opinion based on the amount to which I trust his expertise and motivation. Unfortunately, this does not translate to the online world; from my perspective, there is no difference between one reviewer and another.

Geisler’s suggestions for improving tagging conventions can also apply to online recommenders. He suggests that tags might be limited by system constraints, a similar problem in some reviewer sites. Back to my example, I had been looking for a camera that works well underwater, but CNET’s 5 star system doesn’t give me a lot of information, basically requiring me to read through every user post. A few more specialized sites have different ratings for particular features of the camera (e.g. picture quality, price, etc.) but they are still very generic and do not begin to cover many cameras that have custom features.

CNET’s not useful reviewer recommendation system
Bestbuy’s slightly better system

(At this point, I read the Lerman paper so I’m going to switch from reviewers to recommenders. But before I do…)

I would argue that reviewer recommendation communities (and I use the term “communities” loosely) lack much of the social capital that offline advice givers have. Much of it has to do with the fact that most users are unknown to us, and we’re not sure whose opinions we can take seriously (I personally suspect any positive feedback to be written by employees of the manufacturer). Information given by any reviewer will also not be as rich as that given by someone you are asking in person. Any questions you have can be immediately answered. On the other hand, personal advice seekers are limited by their offline social network, which is not always large, nor can it guarantee that they will find someone who has the information they need. There is no obvious solution to either side, but both sides can leverage each other’s positive features by meeting in the middle. A more integrated reviewer system in the near future might be linked to our personal social networks. User Kar-Hai in Facebook might be karhai in Amazon.com and KH in bestbuy.com. Knowing that information might be able to reveal who the reviewers are in any online retail site (if I am linked to them in some way), and from that, increase our trust level for their opinions. Even at more than 1 degree of separation, hearing that a friend of a friend entered some input might be more useful than not knowing anything at all. This essentially helps online systems benefit from our offline trust networks, and offers a much larger potential group of offline individuals from whom we can access advice from.

Lerman reviews one major difference between digg and reddit: collaborative vs. social filtering as a mechanism for news story selection. In both cases, the majority of stories seem to be the result of a popularity contest (whether explicitly with digg or implicitly with reddit), another example of Web 2.0’s democratization at work. Although the social recommendation model seems to be taking over, as pointed out in a footnote that reddit recently added a friends feature. Ignoring whether or not socially recommended news might be interesting, there is also the concern of accuracy. Leibenluft uses Yahoo! Answers as an example of a question and answer site that might do more harm than good, given that many answers can be off topic, misunderstood, or completely incorrect. Similar to what I’ve mentioned about reviewer systems, how do we know whose answers to trust? And while I know which of my offline friends I can trust for certain advice, am I likely to know someone who can answer my question every time I have one?

Offline, if we seek advice, we generally go to our friends. The topics of discussion are always in areas where we share a common interest. While digg and reddit might offer news stories that we would not have heard of, it is just as likely going to show news that we don’t care about. Oppositely, we will always discuss common interest stories with our friends, but risk losing an opportunity to hear about something new. From a social network analysis point of view, it is another version of the argument between weak ties (online aggregate news site) and strong ties (offline friends). On top of that issue is the problem of accuracy, or as Leibenluft sees it, quantity vs. quality; it is also at the core of how I see the benefits of online (quantity) and offline (quality). Again, I imagine an integrated social-trust network to use the strengths of both sides. For example, if digg used a variation of social filtering that included news from your personal social(-trust?) network, it would provide news you found interesting. You could extend that to some amount of degrees to help provide different news coming from outside your core network, but still within limits of your areas of interest. Similarly, question and answer sites could also use an integrated social-trust network (as with reviewer recommendations) to help ensure that you could assign some trust value to the answers being given. From the offline perspective, you would gain the benefits of the online world, namely quantity. Asking a friend any question should automatically be directed to his friends, and his friends’ friends, and so on.

Lastly, I wanted to make a brief comment on Gazan’s analysis of social annotations. Answerbag.com is clearly an example of how successful collaborative filtering can be, and while admin moderation in conjunction with the ratings system helps maintain the community, it doesn’t keep off topic discussions from entering (which eventually led to conversation vs. education) nor maintain 100% accuracy in answers. However, some of the 8 major decision points are good examples of taking offline affordances (from book annotations?) into the online world. Given that Answerbag.com has reached critical mass, the decision points are helping all of that quantity have some quality. Bringing quantity to the offline world is a much harder task. If I asked my friend a question, could he automatically post it in Answerbag, Facebook, Twitter, etc., at the same time? Would an answer show up on his mobile phone a minute later? I think the 2 integrations involved – online and offline, and multiple social networks – would help our offline world gain the quantity that the Internet has always provided. Knowing that we are still maintaining a connection through friends of friends, no matter how long that extension, still provides some assurance that the advice we seek comes from an individual that I am somehow linked to.

Can technology make social capital “cheaper” ?

I joined 2 radically different online communities, in both context and in regards to how they implement social capital and trust mechanisms. The first community I joined was http://goallineblitz.com (username Griffmann). It is the site of an MMORPG game based on American football where you can build players, buy teams, organize strategies, plan budgets, etc (I’ve had friends pushing me to play the game for some time so it was a good excuse). The community of players seems to be very ride ranged, from guys who played high school and college football, to Europeans who have a fascination with “American football,” to computer gurus who have never played sports before. Because of the different roles you can play (at its core, it’s still an RPG), people can and do have unique behavior, although not in the same sense as the Gleave paper. In public forums, team owners always have an authoritative voice while individual players’ opinions do not seem to carry as much weight. I could see how owners, general managers, players, etc, each have their own patterns of interaction that could be mapped and studied. It also seemed to follow along Eryilmaz’s results on establishing trust, where those members who were higher up could make a statement and their word would be taken as fact, where I, as an agent with very few players and no management or ownership role, would have any statement doubted or rejected.

On a tangent, I had several issues with that study because it didn’t seem to deal with real situations well. I understand their ideas concerning trust, and to be fair, a system that can help guide users on how much to weigh certain information is not an easy task. But I think that context is a much bigger concern than the theory supports – that is, if someone dials 911 and reports a fire, it doesn’t matter if you’re a homeless man on a payphone or the mayor, the dispatcher will send a fire truck. Eryilmaz’s study provides some great information about how to apply a trust model to develop a risk evaluator, but situational context pushes any judgment of trust into a huge grey area.

Back on topic, administrators control forum moderation in the site. There is no functionality for ratings, or user input of any kind. That eliminates a lot of issues regarding trust, as admins can jump into a discussion and act with God-like powers: confirm or deny rumors, ban users, lock threads, etc. I tried one type of interaction similar to the previous week’s assignment, an attempt to insult a user’s suggestion and also troll for responses. I’ve seen team owners and general managers do this, and receive a lot of support, as people assume they are generally correct. From a bottom-of-the-barrel user perspective, I doubt I would get the same treatment.

http://www2.hawaii.edu/~karhai/ics691/glb1.tiff

http://www2.hawaii.edu/~karhai/ics691/glb2.tiff

Surprisingly, I received a few responses of support to my blatant insult, which I imagine has a lot to do with the lack of democratization in the site. Since individual users have no way to rate me, or reduce my functionality in any way, my simple insult did not reach an admin’s radar, but even if it did, it probably wouldn’t trigger a reaction. I only realized recently that while the site supports an online community for a MMORPG game, it is also a business. Why ban users when we are the main source of income?

The second community I joined was http://www.surfermag.com (username: bacardi), a community of surfer enthusiasts from different locations. I looked at the users in light of Williams’ attempt to measure social capital and Ellison’s paper on applying that to Facebook. Separating social capital into bridging, bonding, and maintained might be an oversimplification, but it served the study. Other studies have shown that SNS’s such as Facebook serve to maintain weak ties, which this study also supports (and interestingly points out being able to do this on the “cheap”). * Unlike the football RPG site, surfermag offers the ability to rate both threads and users. I looked around several threads and tried to find one where I could try out interactions based on user whose roles I had found to be focused on answering or helping. I tried asking similar questions in a thread and private message; I sent a user a message after giving him a high rating. Unfortunately, I haven’t been able to get a response in either case, so the results of this portion will have to wait.

http://www2.hawaii.edu/~karhai/ics691/sm1.tiff

http://www2.hawaii.edu/~karhai/ics691/sm2.tiff

The first screenshot is a message I sent a user regarding pictures he had posted. I asked a similar question in a threaded discussion. The second screenshot shows the user’s profile, including their rating. I’m still unsure how much weight user ratings carry, as I can’t see who rated me, nor if there are any consequences other than as a status. There does appear to be a similar form of roles played by particular users, just like the football site, but without the same game-play competition and business-oriented control, the users support each other’s roles much better.

* This is the point where I’d like to do my final project. Williams and Ellison both discuss “online” social capital, but they do not necessarily bring the technical components of social capital into their study. Rather, they look to convert offline social capital into their online equivalents. While I do not believe in technological determinism, I do accept that technology plays a strong influencing role in how social capital is created, maintained, and lost online. Furthermore, I don’t think we can simply translate offline bridging and bonding forms of social capital directly into the online world. In the form of a question (or maybe 2), I’d like to know what kind of role technology plays in online social capital exchange, and how can we (if at all) measure it?

How do we get users to contribute?

I had 2 viewpoints for each of the readings this week, how it related to our Answerbag assignment, and how it stood on its own. The 2 big papers were Ridings and Ling. Ridings’ research was a study on why people joined online communities. She covered many of the possible reasons, including a need to belong, finding information, achieving goals, etc. However, she refers back to the literatures that cite these reasons, literature that covered face-to-face (F2F) groups. Much discussion about the basis for extending those motivations to include online communities. In fact, many of those examples have problems when being used in an online context. For example, she points out the most frequently cited reason to join a virtual community is to access information. But could the large number of information websites available be the reason for this? While one could join a bowling website to find out information about bowling, it is less likely one joins a bowling league just to ask bowlers about bowling. Another motivation she lists is recreation, specifically adventure MUDs; But does recreation automatically include social interactions? There are people that join real world communities for jogging or surfing, but those activities can be completely solitary events. Ridings’ recreation seems to imply only those that are socially interactive, which makes me think the recreation portion is not even relevant, only the social interaction. Another issue I took with the research is that while it was quick to point out that previous literature was limited in the community size being studied, this study only had 27 communities! One could easily find 27 communities for a single topic. In fact, it almost seems as if you could select a community just to fit into a category that you are looking for. For example, just taking a cursory glance at the sample websites, it isn’t hard to guess that communities for conceiving a child or PhD students would be under the category for social support, where the communities for back pain or wine would be for information. Ok I don’t think I meant to type (re: flame) this much about Ridings.

Ling’s paper tied a lot closer to our Answerbag assignment, and I used ideas from it when I posted to Answerbag. She lists several hypotheses about how to increase participation from users in an online movie rating/discussion community (which I also have a problem with), and then the experiment and results for each. I tried to use the suggestions that would counter social loafing in my Answerbag posts. Specifically, making users realize their effort (re: comments or answers) was important and unique (by responding to everything), and making them feel like we shared similar interests (by agreeing with what they said). Unfortunately, this did not work too well. I managed to get one question to have 8+ answers, but the rest were big failures (another one got 7). I decided to try the opposite approach, similar to the realization that Ling made regarding similar groups: that while such groups might help users feel more comfortable, too many people with similar ideas will reduce the amount of uniqueness. Heterogenic groups might serve to create better dialogue discussing topics where users will have opposing viewpoints. I asked several new questions that were more controversial (difficult to do without repeating questions that had already been asked), and then taking a devil’s advocate view when responding to comments. This approach not only gave me more people that responded, but also a lot more points rewarded to both my questions and answers! I even got a “one of the best answers” as a response! Some of Ling’s other hypotheses were harder to accomplish in our context, such as having goals or incentives, and in the end, I was only able to achieve 1 of the 4 goals (receive 8+ responses to several questions). I found it very difficult to sustain a conversation via comments to an answer, and in most cases, only 1 person would reply, and usually not more than 1-2 times. Regarding the scoring, I managed to get a few questions with over 20 points, and in time they might hit 40, but perhaps I wasn’t controversial enough.

The Tedjamulia paper describes a model for studies on contributions to online communities. I was a little confused whether the proposed factors could be practically used to attract new users or to convert current lurkers to become better participants. Many of the propositions do tie in with Ling’s research, and I do think they are valid. In fact, several of them are connected to Web 2.0 applications (i.e. ease of use and interesting content), which might help explain its success. However, I think one important factor they ignore is the value of any contribution. Users who make posts just to incite an argument, or to attack another person do not really contribute anything to the community. In fact, many such users might be the most “famous” person in a given community because of their high post counts, and willingness to attack anyone on any topic, but provide little to no knowledge or resources back to the community as a whole.

Speaking of Web 2.0, Java’s Twitter paper is a nice way to conclude what I feel like is more of a rant than a potential discussion about these papers! I was less concerned with all of the social network statistics, but more rather about why users microblog. I think it identifies with many of the motivations listed in the earlier readings: the feeling of being in a similar group (assuming those are the people reading your Tweets), the uniqueness of your contributions (obviously), and your contribution to the “group.” The last reason is based on an assumption that people are reading your Tweets at all, but as Java points out in the paper, most Twitter users were invited by friends to use the service, and so you can almost automatically assume that at least 1 person will care about what you are typing.

So does this mean Web 2.0 technology such as Twitter or Facebook fill the quota for how to maximize user contributions? Is this what makes them, and other Web 2.0 applications, so successful? It almost seems as if online communities didn’t really become increasingly popular and sustainable until the Web 2.0 revolution (based on my own experiences anyway). The Answerbag assignment was a good demonstration of user-contributed content, with a democratic method of determining value. And before I forget, my username on Answerbag is bacardi.

edit: Woohoo! One of my answers broke the 40 pt (59!) and I reached my 2nd goal! (a bit late doh) I was actually particularly happy with that answer 🙂

More about people v. technology (and which side Google seems to be on)

I’ve found that after I get a few pages into any article we read, I always go back to see when the article was written. Usually I’ll read a line and think, “That doesn’t happen anymore…” and then realize I’m reading something written almost 10 years ago!

I thought the Galston paper was the most comprehensive, but it was published in 1999! I like how he described similarities between the effects that both the Internet and television had on society; Weeks has the same point of view, discussing how the Internet is developing faster than social interactions can adapt. Galston also spent a good portion defending his argument that, “[Contemporary American society] … is structured by … individual choice, and the longing for community.” My first thought was, “Isn’t that what many Web 2.0 components are built for?” That of course, was when I looked up the date of the article. From that context, Web 2.0 technologies would certainly be considered socially driven, and Galston’s views on “volunteer communities,” low entry and exit barriers, etc., would make him a social determinist.

Bigge, 7 years later, takes a different point of view. He describes SNS’s such as MySpace or Facebook to be a “necessity, rather than an option.” He cites examples where users almost felt required to have an online presence; Rosen’s data on the virtual spaces of Hilary Clinton and Barack Obama agree with this analysis. Bigge describes those that resist joining SNS’s and blogs parallel to people fighting an addiction, further emphasizing the necessary nature and power of these systems. His paper sides with the school of technological determinism, and he makes strong arguments that it is technology in the driver’s seat, and users are forced to climb aboard or risk becoming a social outcast, both among friends and co-workers.

I like the social v. technical argument, freely admit to leaning on the social side myself, and will discuss it a bit more, but I suppose I will try to tie in the remaining readings. I liked the LaRose article, also partly because it was written so long ago (it’s becoming easy to think of 8 years as “so long ago,” especially in Internet time). In studying whether the Internet can be linked to depression, the paper cites many different examples both supporting and contradicting the theory. In the end, their study shows that Internet communication can actually alleviate depression in certain cases, and that depression might be caused by stressful interactions with the Internet itself, rather than other people online. I’d accept those results, because it also supports a socially deterministic point of view. If people were really going online because technology was driving them, it could certainly be linked to negative experiences, of which depression could be a result. It is simple to see that performing actions which you did not want to do, or were forced into by the devices themselves could clash negatively with the users. On the other hand, if social forces, such as a desire for interactions, community, or need to be heard, rather than the technology, pushed for people to get online and blog, join SNS’s, etc., then the actions would be more natural. LaRose’s results showing that the interaction with the Internet (i.e. HCI issues, connections problems) could be the major component in stress now make much more sense.

As for joining an online community, I signed up on http://androidcommunity.com (username bacardi), a portal for Google’s mobile operating system. Obviously, I just recently purchased the T-Mobile G1, which uses Android, so I’ve been meaning to get involved in a community and find out how best to use Google’s newest toy. I want to post 2 things that make sense with what I’ve been writing about.

The first requires a little bit of background. The G1 phone was released with some disappointing features, or lack thereof. For example, while there was a full querty keyboard, there was no virtual keyboard that some users have grown fond of. Another is the lack of video support on the phone. Both are found in many other similar phones, and would otherwise make the G1 an unattractive buy. Fortunately, the Android operating system is an open source project. Not only that, but Google has make it scalable, and promised to provide support for Android, especially as it gets used in more mobile phones. With that in mind, the newest update (codenamed “Cupcake”) for Android will actually include an option to use a virtual keyboard, and also have support for video. The important thing to note here is that it is not HTC, the phone creator, who is updating the phone features, but rather a joint effort between Google and the Android community. The fact that Android is open source fits the bill of a Web 2.0 culture which leverages user participation into creating a better product. Here is the link describing Cupcake’s keyboard features: http://androidcommunity.com/android-cupcake-screenshots-show-new-qwerty-settings-20090123/

The second is a little less exciting, and I lied about it having something directly to do with what I’ve been writing about. I joined in a discussion about the new video updates coming with Cupcake. To be more specific, it’s actually someone’s blog on the website. The site apaprently has separate components for discussions, pictures posts, blogs, social groups, and knowledge base. However, it seems to all tie together, and sometimes it’s difficult to tell where I am. I posted a simple question here: http://androidcommunity.com/forums/blog.php?bt=437 and also in a community discussion here: http://androidcommunity.com/forums/f8/all-cupcake-rumors-11020/index28.html. It doesn’t quite relate to the social v. technical question, but I’m curious as to whether I’ll get more information from a community forum or individual user blog. I guess I’ll find out in a few days.

What is social computing?

Social computing as an umbrella term for technologies, virtual spaces, and their associated communities seem to cover a lot of ground. Its relationship to online social networks, or “social network sites (SNSs)” as described by Boyd and Ellison is obvious. Examples of SNSs include MySpace, Facebook, LinkedIn, etc. All such SNSs are virtual communities built to support users in connecting with their friends or peers; all are egocentric, where each individual is at the center of their community. As such, SNSs are driven by the virtual space where the individual “lives.” From this space, each user can travel around his community to explore the virtual spaces of other users. The means by which they can travel will be dictated by the technology involved (which I will discuss in more detail).

Its relationship to Web 2.0 applications, of which SNSs are but one, are also apparent. Beer and Burrows detail several other examples, such as blogs and wikis. Blogs, like SNSs, are egocentric virtual spaces by its very nature, but depend on a community of followers to flourish. Wikis on the other hand, are built and supported by the community. Both blogs and wikis are forms of social computing.

Tenopir’s article, while not directly addressing “what is social computing,” actually brings up the point I’d like to make. She points out several concerns about Web 2.0; namely, that anonymity can be used to mislead and democratization of information reporting can degrade its accuracy. Both points are valid, but I would argue that the masses have and will adapt. To provide an umbrella question for Dr. Gazan’s umbrella definition, who is in the driver’s seat, technology or society? Are our actions determined (and limited) by Web 2.0 technology or is society driving the development of technology to support our needs? Tenopir shows the signs of a technological determinist; if she believes that an anonymous conveyor of information will deceive the masses, or that we will go to blogs and YouTube for our world news, then her view is that technology is herding the masses to behave in a directed manner. Instead, I believe that society will have the final say. My evidence is only in my own personal experiences and anecdotes. There are many times I’ve heard a “breaking” story first on a blog, or websites such as Slashdot, but never have I taken it as the absolute truth unless I’ve also seen it on cnn.com, nytimes.com, or another reputable news site. This may have served to move the major news sites from my primary source to a secondary/verifying source, but it also reinforces those news companies’ reliability. Another example is a site such as Digg, which takes the news portal idea and extends it to support the Web 2.0 democratization; Digg accepts user posted stories that come from anywhere in the web. It requires “votes” for a story to become popular, but most popular stories are usually ones where the attached discussion includes different sources, as people look for verification.

The two papers concerning blogs (Nardi, et. al, Herring, et. al) give another view of social computing. It’s interesting that both papers, published in 2004, never use the term Web 2.0; however, it’s clear that many characteristics of Web 2.0 applications are already recognized in blogs – defined as “possessing a socially-transformative, democratizing potential.” The in-depth content analysis provided clues that blogs served as a bridge between static web pages and CMC systems. Many of those bridging features (e.g. user generated content, asynchronous reader responses) are the beginnings of Web 2.0. Nardi’s paper supports my idea that blogs are socially driven, giving examples of users that develop blogs based on their audience, and others that started blogs because they were asked to. However, her research also reveals the speed of the internet; she made several suggestions on blog design, most of which are already commonplace in blog software: spell checking and other integrated document editing tools, file sharing, photo management, etc. As blogs became more popular, many social aspects of it were expanded. Instead of just an online diary (or radio shows, Nardi might argue), different media have been included to give a much richer experience to both writer and reader alike.

From a general definition of social computing, I find that it’s essential to bring in the relationship between the technology and the social (although I hope there are not Latourians in the class). Beer and Burrows strongly advocate research in “sociological description as applied to new cultural digitizations,” but I think the balance with the technological, as determined by the interactions between the two, is needed. Studying how users interact with different systems, whether by broadcasting, as in blogs, or networking, as in SNSs, or mass democratic knowledge contributions, as in wikis, there are different social and technical interactions involved in each. I’d be interested in – and also define – social computing not as a general description of the different technologies or communities that relate to Web 2.0 or any other online application, but rather how the “social” and the “computing” interact and affect what users do online.

user 2-way status bundle for second life

linden labs would like some documentation of a project we are submitting for a prize, and since a blog was acceptable, i’ve decided to use it.

i think the easiest way to document the user 2-way status bundle (what a mouthful) is to write up a use scenario.

i’m sitting at home one day, working on building new tools to help teachers inside second life. my second life office is a close replica of my real office, except i can fly between desks. all of the desks in our e-office also have pictures of their occupants hanging above it, and text underneath the picture showing their latest twitter status.

back to me… i’ve been stuck on a problem for an hour now, with no solution in sight. to my (virtual) left, i noticed something has changed. sam’s picture suddenly became much brighter, and his twitter status has changed from “changing diapers” to “just got to the lab.” just the person i’ve been looking for! i walk over to his desk and click on his face, and immediately get notified that an email has been sent to sam, asking him to join me in SL.

it doesnt take long before sam arrives right next to his desk. i give him an e-wave, plug in my microphone, and do what most people do when they hit a problem they cant solve: pass it on.

awareness and notification are 2 HCI concepts that must be properly addressed when designing any new system, especially one that is an entire virtual world. what the user 2-way status bundle does is provide both to SL users in-world, bridging them to their friends and peers in the real world.

ics 691 social computing

i wonder if capitalization is important for this class.