« SLurl | Main | Hamlet who? »

Mar 06, 2006



One thing that struck me about the differences in our articles, Nick, was that yours focused on labor while mine focused on production, although we are certainly dancing around the same topic. Tracee Wolf of IBM sent me a really interesting e-mail after reading my article that asked: "How do you measure productivity?" Ted Castronova has done this famously using traditional econometrics in terms of the GNP of EverQuest http://papers.ssrn.com/sol3/papers.cfm?abstract_id=294828 . I think in some sense, he is also looking at this in terms of labor vs. value. I'm not sure how you would 'measure' productivity in a play environment, but that might be an interesting question for someone like Ted or Tracee to chime in on. I personally am more interested in these artifacts as a function of social currency rather than monetary value; how to measure that would be an interesting research question.

The outsourcing of labor is another interesting trajectory. We know that people outsource, for instance, ‘Adena farmers’ in Lineage, low-wage workers who farm for game currency to sell on the ‘black market.’ This creates interesting class and even race tensions, such as the Lineage 2 scenario described at State of Play 2004 by Constance Steinkhueler. Here, Adena farmers typically took the roles of female elf warriors (primarily for farming efficiency reasons); as a result, this race/class in the game began experiencing racial slurs and attacks by players who associated it with Adena farming. So here you have a situation where labor and play are literally converging one of the emergent byproducts of this are resulting racial and class tensions as a result.

Finally, your comment about the cancer screening scenario brings to mind James Surowiecki’s new book The Wisdom of Crowds (Anchor Books, 2005), which has some really interesting ramifications in the MMOG world; this book documents the benefits in some applications of throwing a large number of diverse ‘non-experts’ at a problem. I have actually been working on some game concepts that try to take this approach to problems like the environment and space tourism: if you have a large enough quantity of a wide enough range of people, you might actually get some pretty creative ideas and do so much more efficiently than traditional means. If these people perceive what they are doing as a ‘game,’ this might produce the added benefit of people being less risk-averse and more experimental. I guess for me the distinction may have to do with creativity. To what extent is being a pharmaceuticals manufacturer in Star Wars Galaxies creative? Or is it really more routinized labor, like being a factory worker?

More to say…but I’ll wait to hear your responses.


Celia - I'm glad you pointed out the distinction between your emphasis on production and mine on labor. What struck me in reading over players' narratives was their use of words, such as tedium and obligation, to describe their own game-play. I also like the tension in the pair of articles between production as creative force vs. labor as oppressive force.

I've also thought more about gold farming since State of Play II, particularly along the lines of Jerry Kang's reactions. What fascinates me is how race/nationality is now invoked to create the social category known as "gold farmers" (rather than the other way around). For example, players who pass the "English Test" aren't "gold farmers" even if they farm all day. But ironically, players who don't speak fluent English (i.e., French) are at risk of being branded as "Chinese gold farmers".

Even though gold farmers have completely blurred work and play, I feel it is the case of the casual players (performing equivalents of the pharma task) that are most interesting, because they are working without realizing it. And indeed, the cancer-screening idea was something that came up during a group brainstorm where The Wisdom of Crowds had been reading material.

I would frame the pharma example as neither creative nor routine, but rather as a problem-solving task of moderate difficulty. I think that's what led us to the pharma example - could we harnass this 20-hour-a-week collective brain power?


Ok, i feel invited to take part. Here's my point: "Even though gold farmers have completely blurred work and play, I feel it is the case of the casual players (...) because they are working without realizing it."

In fact there are people, let's call them sweat shop gold farmers, who work for money and generate virtual goods, to be sold for daytime cash. To those people gold-farming is an office-job, not very much unlike other office-jobs. So i don't see any blurred line here, this is about making a living.

Putting hard cash aside, a lot of people would work hard and to the risk of RSI in an online virtual world - just to gain a social advantage in their chosen peership. I take it as a matter of fact that people are playing games to fit their needs: social acceptance, individual success. Carrying out a 'work metaphore' in a game does appear to us as daytime work, but with a huge difference: virtual work (not as in sweat shops) is always related to individual gain or benefitting the peer group. Virtual work is not - and is unlikely to be any time soon - taking part in any industrial production process (not even in the huge ship-building activities of eve online), there is no 'alienation' (yes, i'm talking Karl Marx) stepping in between the worker and his product. It's a pre-industrial work culture, roughly to be compared to the DIY culture of our time. Our common picture of work is formed by the industrial production environment, so we need to understand that in-game work is not 'work as we know it' but the original, pure form of 'because i want to' or 'let's just do this'. So to my understanding the 'blurred line' is not drawn between work and play, but between work (as in: general motors assembly-line) and work (as in: building a little wood house for my kids and me to sit in). This is what games are for and *imho* some MMOGs are quite on the right track.


I asked Celia the question 'how do you measure productivity?' because I have been asked this question and am uncertain how to answer it in a way that capitalist-minded or empirical-minded folks would like to hear. And since it tends to be a capitalist-minded person asking the question, I’ve always felt obliged to try to answer the question from the standpoint of how their bottom line is met. But I cringe when I consider the answer to the question based on this point of view. In many cases, it seems like an impossible question to answer from that point of view. There is more to productivity than money.

The government uses the GDP as a measure of productivity, but there are those who believe positive growth of GDP is not sustainable and therefore not a good measure for productivity. When I was at the ">http://www.id.iit.edu/events/strategyconference/2005/index.html"> Design Strategy Conference in Chicago last year, Josephine Green of Philips Design gave a fantastic ">http://www.id.iit.edu/events/strategyconference/2005/fullschedule.html#green"> presentation in which she spoke of the GDP as an unsustainable goal. This was the first time I’d heard of an alternative national goal: happiness, measured as http://en.wikipedia.org/wiki/Gross_national_happiness ">GNH (Gross National Happiness). She argues that a person’s happiness increases until an annual income that supports their basic needs are met. At that point, happiness remains stable and more money doesn’t necessarily mean improved well-being; it is the point of diminishing returns in a person’s quality of life. She also pointed out that continual rises in the standard of living has consequences for our environment – which we depend on for our social well-being.

The New York Times reports on how a small Himalayan kingdom of Bhutan is trying to apply this measure of well-being in a real way.

With this kind of thinking, I challenge myself to answer the original question in a way that does not address monetary issues directly. Instead, I wonder how we might create a measure of productivity (well-being) based on happiness. How do we create a measure of the social condition that is understandably standardized?

Now I can start to see a clear and convincing discussion of game activity contributing to the well-being of organizations as well as individuals. However, I do feel we have some work to do to win over the minds and hearts of those who ask the question from the capitalist point of view.


First, I want to address Nick's comment about the 'pharmaceutical industry' in SWG. [Nick: I would frame the pharma example as neither creative nor routine, but rather as a problem-solving task of moderate difficulty.] This is VERY important point because it speaks to the very core of games. Typically, games are goal-oriented activities that often involve problem-solving. So SWG pharma is a case where a real-world metaphor becomes the framework for a game goal. What I’m finding tends to happen with emergent behavior is that players begin to create their own goals which are not part of the designers initial intention. What happened with Uru was that when the game closed, players created the meta-game of ‘staying together’, which then became a problem to solve collectively. They used both their ingenuity and the skills they had learned in Uru and other Myst games to accomplish this collective goal.

One of the key points in the Productive Play paper is that that players are making these efforts on their own and for their own satisfaction rather than at the behest of an ‘employer,’ which also jibes with some of Fritz’s comments above. Like hobbies, this is part of what makes it ‘not-work.’ We could thus call these sorts of endeavors ‘ludic labor,’ that is, a high degree of effort applied within the framework of a play intention. This is a sort of ‘game studies’ spin on GUI innovator Alan Kay’s concept of ‘hard fun.’

However, the example of Adena farmers is interesting because it basically turns a form of ludic labor into labor-for-pay. The flip side of this is ‘playtesting sweatshops’ where many game designers now begin their careers. These usually male, usually college grads (some of them my former students) are overjoyed to have the opportunity to play video games all day for money. Over time, however, the grueling hours and obstacles to progressing through the hierarchy often transform a dream job into the drudgery of factory work. So these are interesting breaches in the traditional demarcation between work and play.

Finally, regarding Tracee’s comments, I really like this GNH concept. I’m wondering if we could develop some similar concepts for play? In terms of ‘productivity’ I have been thinking about finding ways to study creation of player-made artifacts in places like There.com and Second Life. This would include looking at the percentage of play activity that takes place around artifact creation, as well as what types of artifacts are created, and what their in-world monetary value says about the values of the community. For instance, in Second Life, custom avatar designs come at a premium and are much more costly than houses or vehicles. In There.com, where female avatars have only been able to wear pants, the recently introduced black mini skirt has become one of the most expensive and sought-after clothing items. So part of what we might be able to look at is the role of production in community, and what these artifacts represent in terms of their meaning and value within the social system of the game.


The play-testing is a great example, Celia. This also ties in with the AOL/Microsoft volunteers, and to a certain extent the newbie guides in There/Second Life.

Fritz said: Virtual work (not as in sweat shops) is always related to individual gain or benefitting the peer group ... So to my understanding the 'blurred line' is not drawn between work and play, but between work (as in: general motors assembly-line) and work (as in: building a little wood house for my kids and me to sit in).

But what if doing pharma in SWG was doing actual cancer-screening? You think you're just creating virtual objects in a game, but the combined responses from hundreds of gamers are being used for diagnostic purposes for a medical company (something that is currently outsourced to India). And you're paying a monthly fee to do work for this medical company. And what if building a house for your kids to sit in was actually quality control work in a motor assembly plant?

I don't think the pre-industrial or industrial style work in MMOs is interesting. I think it's the post-industrial work (knowledge workers, information management, distributed systems, collective brain-power) that is on the brink of being leveraged here. That's the interesting (and scary) blurring to me.


This discussion brought up several aspects of work and production in the virtual space. There is (1) an active crossing the line by gold farmers and beta testers, (2) work for it's own sake (Celia's ludic labor), including the production of happiness (Tracee) and (3) intellectual productivity, diguised as game routines.

Agreed, Nick, this "Welt am Draht" scenario is scary. What if Blizzard comes up with a glorious 'World of Starcraft', requiring not only giving up your privacy (read WoWs EULA) but performing routine tasks and problem solving repeatedly, so this 'surplus labor' can be sold to data processing companies? And yes, pre-industrial needs may drive millions of gamers into the virtual world, but a post-industrial abuse of virtual leisure activities should be dicussed and stooped in it's earliest stage.

But, and here the abovementioned blurring extends to my vision, do you see real-world examples for this, beyond countless 'what if' scenarios? What could a group of gamers do better than a 'Seti'-like cluster? And how could intellectual work, of a scientific grade, really be cloaked as in-game activities? I don't think that's possible. Do you?


I think maybe some of this is possible... I think the prospect of this blurring is interesting and in some ways inevitable -- but it's not easy to break up interesting work into this kind of system, there are some hurdles:

* roughly speaking the reliability of a person completing a task successfully is inversely proportional to the complexity of the task times the person's level of training.

e.g. a brain surgeon's task is very complex, but she also has a lot of training -- hence her reliability in completing a surgery successfully might be fairly good. If I, on the other hand, attempt to try brain surgery, my reliability will be approaching zero because I have no training in that field.

* if your results aren't reliable you must invoke additional filtering/verification steps to increase reliability.

For example, the cancer-screens use offshore labor -- and the claim is that redundancy offers increased accuracy -- however, what if everyone was trained incorrectly in one detail? Perhaps if they had training equivalent to the doctor's they would notice the error in training (this relates to the first point) -- concensus does not necessarily equal accuracy, so ultimately a trained doctor will still need to review the work.

* in parallel computing, much of the focus is on relatively *simple* local tasks that produce complex global results. So if the tasks aren't simple in this way the untapped labor won't be very scalable or productive. [Parallelization] In some ways, the problem to get work out of MMOs is like the problem of writing parallel algorithms.

* finally, even if I have a simple local task that provides a complex global result I still must have a complex system in place that organizes all the "units of work" into a coherant dataset for analysis.

SETI at home is a great example of a complex infrastructure for batching jobs, handling failover, and stitching the results back into a coherant dataset for further analysis. But it's also incredibly specific and very limited to the problem it was designed for.

In some sense, we have moved the complexity of "working" from one place into another. Instead of the worker picking up a paycheck and going to a workplace and doing work for a manager, we have automated the managers, the worker stays home, and they are "paid" in terms of entertainment or a sense of accomplishment.

Hmm, this also gets me thinking about Celia's question "how would we measure productivity?" Maybe "entertainment" and "accomplishment" have to be considered in the concept of *value* (i.e. like the joy of learning). It would be interesting if gamers thought they were actually getting "paid" a positive net value (entertainment - subscription = net value > 0)... perhaps this net value idea might shed some light on a productivity measure.

Also, regarding the "There" example: I'm not entirely sure the gamers are being exploited. Buying clothes in There is much cheaper than the real world -- how better to experiment with fashion trends and then take what you've learned into the real world? Here the consumer has been repaid with valuable experience and knowledge. On the vendor's end, they now have much better data on what consumers actually want to buy -- they have been paid in rich datasets that are leading them to manufacture only what the consumer really wants to buy (product optimization). "There" has been paid by both parties for playing "matchmaker" -- providing the raw world, the context, the rules, the content framework that enabled a new mode of communication between consumer and vendor.


One of the key advantages to using online MMOG labor for actual jobs is that it solves a common rl problem - the most challenging tasks seem fun the first couple of times, but then they get old real fast. An MMOG can get around this issue by distributing the task among millions of people, and only require the people to perform the task once or twice a day.

Say you are a chip manufacturing company, and you are working on chip layout compaction (making your designs take up the least amount of space possible to save on materials cost). You could pay Blizzard to place a quest in WoW that allows people to organize a bunch of odd-shaped polygons into the most compact possible layout. This quest would reward you with, say, a bank slot or an N-slot container based on how hard a puzzle you solve.

The puzzles are, in actuality, rl chip-layout problems where you are laying gates for a chip and trying to take up the smallest amount of space. The quest could get progressively harder (adding rl chip-layout restrictions) until you have people actually performing design work for the firm.

Now, this task by itself would not be fun for very long. But when people have the option of doing the quest, and they only have to do the quest for 5 minutes a day, they are much more inclined to do so - it would even be fun the first couple of times. If Blizzard gets a million people to spend 5-10 minutes a day doing this, they would accumulate actual productive work very quickly.

This implementation of placing rl work into an MMOG has another key advantage - by exposing the task to a large labor pool, it finds the freakish few people who enjoy this type of work even when they do it for several hours at a time. It is almost a type of reverse-auction where the majority of people will perform the task once or twice and lose interest, but some will grind for hours a day and become very skilled in performing the work. By definition, these hard-core people find the task fun, and will perform it for the least amount of 'money'.

Heck, you could even monitor someone's skill at your in-game task, and if they get good enough, hook them up with the company sub-contracting the work. Maybe the company will have found a gifted chip-designer who never knew he had it in him.


Thanks to everyone for a lively discussion. Here are a few comments on what's been said in the last "round":

Re Larry:
e.g. a brain surgeon's task is very complex, but she also has a lot of training -- hence her reliability in completing a surgery successfully might be fairly good. If I, on the other hand, attempt to try brain surgery, my reliability will be approaching zero because I have no training in that field.

If you believe in the ‘Wisdom of Crowds’ (Surowieki, 2004), then the key is to make the tasks simple and create a lot of redundancy in the system. This way you create a kind of emergent, collective intelligence. According so Surowieki, this approach does not work with highly specialized tasks (e.g, brain surgery), but works well when you have a large number with diverse skill sets (also key; homogeneity will ruin the effect as well), you generally get better results. Phil’s ‘chip efficiency puzzle’ is an excellent example of this. So, to answer your question Larry, I probably would not ask a group of MMOGers to remove a tumor from my brain, but I might ask them to detect it in the first place, or even to design a chip to replace it.

Re Larry’s reference to There
This is more on the lines of what I am talking about, e.g., player creativity. Since There is the primary world I’ve been studying, I can tell you a few things that make its formula work. It’s not just the idea that you can experiment; it’s also the idea that you can be a fashion designer, even if you’re not one in real life. In addition to living out a fantasy, being a successful designer in There brings with it a level of social currency. In addition, as is the case with Second Life, players can sell their items in a free market economy (e.g., prices are determined by what the market will bear). However, in There (and this, to me is the really interesting part) players have to pay to put their products on sale. So entering into the ‘artisan class’ (not a formal class but a de facto one) actually has a fairly high barrier of entry. On the other hand, the rewards are pretty high, as you can usually make back your investment in 2-4 sales of an item, depending on what it is. Conversely, in Second Life, there is no barrier of entry whatsoever. The difference is, in Second Life, there is a lot more stuff, but in There the stuff (which is also subject to management approval) is of a much higher and more consistent quality.

My point here is that were creative production is concerned, there is a high reward both in terms of social currency and virtual currency, so this becomes a highly motivating factor. Add to that the fact that players get to essentially play out a fantasy career (e.g., fashion designer, architect, etc.) that is often a departure from their real-life careers, and players are more than willing to pay a premium for this opportunity.

One more general comment—If find it really interesting how the discussion tends to veer towards the two poles of dystopic vs. utopic view of virtual worlds. Just something to take note of. (I’m clearly on the utopic side of the argument!)

The comments to this entry are closed.