Where's the catch?

Thomas and I recently bantered simulation in MMOGs (why not more of it).  Before this David Edery wondered in a fine essay whether online gaming could be the lubricant for applying "the wisdom of crowds" to real world problems using simulation.  Earlier this year I wondered aloud whether Jason Lanier's beautful rant might also be directed against crowds in virtual worlds (are they overrated). 

Fortunately, Henry Jenkins comes along and forwards a great post to prod our muddle:  should we distinguish a "collective intelligence" from the "wisdom of crowds"?

In Henry's words:

The Wisdom of Crowds model focuses on isolated inputs: the Collective Intelligence model focuses on the process of knowledge production. The gradual refinement of the Wikipedia would be an example of collective intelligence at work.

He cites Edery's spin on the impact of the wisdom of crowds to game play:

Crowd intelligence can fail (and fail spectacularly) when there's too much information passed between members of the crowd. Members start to alter their opinions based on the opinions of others, which skews the results. The online communities that build up around any popular game would seem to promote exactly this kind of skew.

From Jane McGonigal Henry suggests that mechanisms such as competition and population diversity might help avoid "group think" and encourage "feature(s) which enable us to process information in more complex ways than could be managed by any individual member."  I recommend reading Henry's complete essay (Fn1) before jumping to comment here.  He mentions related points raised by Raph.   

Yet I wonder this transfer to virtual worlds:  is the problem of groups in MMOGs, say, not themselves but the lack of virtual world processes against which they can cut their teeth and "generate valued knowledge production"?  Crudely, do guilds become mobs with organization because there aren't enough levers for them to pretend City Hall and dress wisely?   Of course I could be wrong about the mob bit, wink, wink.  In any case, be sure to also read David Edery's essay (Fn2) especially the bit about "SimCity-esq game" and "(t)here's gotta be a catch."

-----------------------------------------------------------------------------------------

Fn1.  Confessions of an Aca/Fan: The Official Weblog of Henry Jenkin.  Collective Intelligence vs. The Wisdom of Crowds.  November 27, 2006.

Fn2.  Game Tycoon.  Using Games to Tap Collective Intelligence.   September 4, 2006.

Comments

[Thomas Malaby]

1.

Interesting ideas. Much of the ethnography of Linden Lab that I'm writing now will bear directly on some of these issues, and I can't begin to bring in some of that here, but, as the epidemiological game example from Ludium I showed, there remain quite a few challenges with aligning the incentives of the game with the aims of policy.

I also simply have to add that one of the surest indicators of the failure of the American academy over the ~60 years truly to educate students is the current rise and popularity of books like Surowiecki's. It's not that it's "wrong" in any particularly important way, by any means. It is that the idea which animates it is Adam Smith's, in The Wealth of Nations, where he saw not only the power of the "invisible hand" (the aggregate effect of lots of independent judgments), but also its core limitations. Most of those who trumpet WoN have never read it (as its shrill usage for a narrow political agenda shows), but its core insights on this issue of aggregate judgment delve deeper than any other thinker that I can recall. If a liberal arts education meant anything today, it would spell bad news for one segment of the publishing industry.

[Junn]

2.

How about this question? (If it is not proper one, blame my damned ignorance!)

Could games of any-kind be put to use in a similar way to IEM? Is it appropriate?

I'm not talking about 'educational' or some 'serious game'-like usages but real 'informative' ones like those from IEM. Before jumping into talking about right model, I'm wondering the scope and applicability of the agenda.

* As far as I know, Hayek's original defense of market is more fundamental than Surowiecki's applications. It may be fair to compare Hayek's idea with Levy's one.

[Michael Chui]

3.

The catch is that you have to build and present it correctly. =P

is the problem of groups in MMOGs, say, not themselves but the lack of virtual world processes against which they can cut their teeth and "generate valued knowledge production"?

This is something that I'm hoping that Raph's "Grammar of Gameplay" can answer. When I hear several of the veteran designers going on about lack of innovation or "This has been done before," what it shows me is that the actual game isn't very interesting. It's simplistic and linear. I've personally discovered that gameplay is increasingly dull; it's no wonder you have guild drama. It's much more interesting. It's no wonder we have raids and such; it's more complex.

[magicback (Frank)]

4.

We should make clear distinction between collective Wisdom and collective Intelligence. Wisdom implies heuristic process while Intelligence imples empirical, scientific processes.

Massive multiplayer gameplay can be used to derive collective wisdom and intelligence.

For example, a game about space exporation have two major gameplay. The first is mapping space and second is analyzing the stellar bodies within the charted areas for things of value. The former is a collective intelligence process (incrementally add charted regions to the master chart). The latter is a collective wisdom process as what is deemed valuable is subjective.

Now most games just create and use a fictional space. The collective efforts and results of the players have no real life application. However, if we use real data in the game, real world objectives like spotting planetary bodies or even signs of intelligent life can be advanced.

Wisdom and Intelligence is related, but not the same.

P.S. Are we now talking about a Web 2.0 business model? MMO gameplay + Amazon's Mechanical Turk processes + company willing to pay for results = profits?

[Andy Havens]

5.

What I would be interested in observing (one of my soapboxes, again) is exactly this; uses of MMOs to "do things" that aren't available in other game-spaces, enabled exactly by the vast-numbers of people now taking part.

We are starting to have hugely participative entertainments outside the realm of VWs/MMOs, where people are (at the lowest scale of participating) at least searching on the web for content (as opposed to watching a fixed show at a fixed time on a fixed channel)... that requires a minimal level of interaction, though it may not be "game-like" enough for us grognards.

[Sidebar: then again... maybe searching for funny videos of talking cats, Photoshop tutorials, porn and guys dancing on treadmills *is* a kind of game... ]

Anyway... the point being that people are more and more willing to "do stuff" at the point of entertainment. We have been told (forever) that people don't want "interactive movies" where we pick the ending. I still agree. Good, long, narrative fiction is still very satisfying. But in short bits? The creators of "LonelyGirl15" adapted their bi-weekly, 2-minute long episodes on YouTube based on which segments got comments, and on the content of those comments.

That's not necessarily the "Wisdom" of the crowd... Except that ain't it nice (wise) to get (and give) more of what tickles ya?

Are there ways to embed "crowd-based" mechanics in games that would let us tweak the modifiers, inputs and rewards such that we could, over time, learn some interesting stuff... or at least be vastly entertained? I hope so.

I'd like to see, for example, an MMO where there are multiple "macro" routes to success. We've talked about this before (here and on Raph's blog, among other places), but right now, in classic MMOs, you get to choose whether to be rock, paper or scissors... but not whether or not you play rock/paper/scissors. A game where you could be successful by, for example, searching out clues and providing them to other players? Hmmm...

As far as "crowd wisdom" goes, you could present situations that give different options and rewards based on the number of players involved. And then see if, yes... a mob truly does, mathematically, perform "better" than individuals.

On this whole "Wisdom of Crowds" thing...

On the one hand, I'm not sure why we're surprised. A group of people is stronger than its strongest member. Why shouldn't it be smarter?

On the other hand, I'm not sure why we believe it. A group of people isn't faster than its fastest member. Why should it be smarter?

[Ola Fosheim Grøstad]

6.

Andy Havens:On the other hand, I'm not sure why we believe it. A group of people isn't faster than its fastest member. Why should it be smarter?

Problemsolving is a search-problem. Usually crowds cover more ground, but sometimes the solution is on the top of Mount Everest... (wisdoms of crowds, distributed cognition yadayada... hype)

[Mike Sellers]

7.

Crowd intelligence can fail (and fail spectacularly) when there's too much information passed between members of the crowd. Members start to alter their opinions based on the opinions of others, which skews the results.

This is succinctly articulated by the Demotivator that brings home the corporate group-think reality that "None of us is as dumb as all of us."

So yes, problem-solving can be seen as a search problem. But when (due to over-reliance on in-group information, as stated above) we all gravitate toward the same local minima, we all tend to end up stupider than any one of us might have otherwise.

Can small weakly connected groups (maximizing search space) in a virtual world be put to work on "distributed wisdom" problems? No idea, but it's an interesting thing to contemplate.

[Andy Havens]

8.

@Ola: Some problem-solving is related to search. Some is related to "grind." Some that is related to search is going to be good for crowds unless, as Mike points out, the crowd "gravitates toward a local minima." And, we assume, it's a minimally appropriate minima. Some that is related to search is bad for crowds, too, even if they don't cluster, because even if the crowds are spaced nicely, crowds aren't good at certain things; like differentiating between noise that comes from within the crowd and without. See Pooh and Piglet tracking hephalumps around the bush by their (Pooh and Piglet's) footprints.

Some problem-solving is related to grind. Cracking the human genome wasn't a spread-out search; it was doing highly specialized, repetitive crap over and over in a methodical way. You can call that a "search problem" if you like, but it relies on so much precision, that if you wanted to aggregate parts of the "search" to members of a crowd/group, those members would need to become so sophisticated that the "crowd" would be a crowd of specialists and blow the whole notion of group intelligence out the window. There were, if I remember correctly, only a handful of organizations in the "crowd" working on that particular problem, as it required such sophisticated tech and knowledge.

My point is simply this: if you need someone to say, "This is how we are going to define the problem such that we can determine whether or not the wisdom of the crowd is appropriate to this problem," you've essentially said that you need a "Crowd wisdom expert." And that becomes somewhat goofy, imo. Because if an individual is responsible for determining whether or not a group solution is better than an individual solution, we've got a kind of intellectual conflict of interest. And you can't ask the crowd whether it is smarter than a particular expert on any particular issue, can you?

Are we lifting things, finding things, racing or guessing the weight of beef at county fairs? Yes, there seems to be some times when the collective intelligence of groups of people can be harnessed to show us useful stuff. But, as of yet, I haven't seen a really good index of "Here is the crowd-trusted list of questions" and "Here is mob-stupid stuff."

I'm interested and intrigued, yes. And I do, generally, tend to trust the voice of the many over the voice of the few.

Unless that few is me.

Unless I need a doctor.

[dave]

9.

Tommy Lee Jones said, "A person is smart. People are stupid."

I agree Tommy. I agree.

[Ola Fosheim Grøstad]

10.

Andy: Some that is related to search is going to be good for crowds unless, as Mike points out, the crowd "gravitates toward a local minima."

Yes, that has been covered in research on brainstorming. You are generally better off having individuals work on the problem in isolation before they start to cooperate. That is just to ensure that you cover more ground. Nothing new under the sun. ;-)

Fashion is a crowd phenomenon. It may seem stable in a very limited timeframe, but it actually isn't. One multi-user game design challenge is to nurture the desire to be different. Explicit scoring systems tend to kill this... and designers resort to ugly expansion-tactics in order to artificially move the fashionable goals... Ugly, but state-of-the-art.