Is Digg Wise?

This morning I went over to Digg.com, the godfather of "user powered content aggregation." There I discovered that Digg users had this to say about the content they had powered:

• 889 votes for a picture of a McDonald’s ad on a campus that was based on the premise that since food is more difficult to steal than music, you might as well get your food cheap.
• 530 votes for an article about Terrell Owens, the receiver for the Dallas Cowboys, trying to kill himself (literally this time, not figuratively like all the previous times).
• 313 votes for a rather thoughtless but still provocative rant about America losing it’s standing as the number one economy in the world (apparently our problem is not enough taxes on the middle class – who would have thought?)
• 39 votes for an article about funding for Wallop, a new social networking service (Microsoft’s entry into the “Let’s kill MySpace” market).

I didn't find any articles about Bill Clinton's recent tête-à-tête with Chris Wallace, but I did find a lot of people interested in a girl trying to blow up her SUV with her sweater.

Digg is presently one of the top 20 trafficked website in the world, and its users search far and wide for interesting new content. The Digg user community has spoken loud and clear: silly ads are more important than America’s economic position or the funding of a new business. Why think about the merits of a political argument when you can watch a Jeep go "boom!"? Different strokes for different folks, I guess.

Digg is based on a “Wisdom of Crowds” concept. And since I am a big fan of both the logic and the possibilities of that idea, one would think that I am a big fan of Digg. But as my analysis above probably telegraphed.. not so much. When the crowd that I am depending on to tell me what I should read thinks that cheap McDonalds copy is more important than the ability to create jobs in the future I tune out. I am left with the conclusion that the Digg crowd is not that wise, and I don’t just mean that I don’t agree with their preferences.

“The Wisdom of Crowds” was James Surowiecki’s breakthrough book which brought the heretofore mutually-exclusive concepts of "Wisdom" and "Crowds" together in the common lexicon. According to Surowiecki a “Wise” crowd has the following characteristics:

• It is diverse in composition
• The individuals are well informed and have access to the same data sources
• The opinion of one individual doesn’t unduly influence the opinion of others in the crowd
• There is an objective method for aggregating individual input into a group conclusion

When a crowd meets those criteria it is amazingly adept at predicting the future, including what is important and what isn't. When a crowd doesn’t meet those criteria, however, it runs the risk of thinking that McDonalds’ copy is a better use of your time than why somebody would try to kill themselves or why America may be in economic trouble.

So if user powered content is based on the “Wisdom of Crowds” then we have to examine whether the “Crowd” meets the “Wise” standard. In most cases the answer is clearly “no.” Here are just a few of the reasons why:

Most crowds are self-selecting and therefore not diverse. There are a lot more geeks using Digg than grandmothers. And the smaller the crowd that is powering the content aggregation, the lower the probability of diversity of opinion or thought. If you intentionally limit the scope of the group aggregating information then it may be that you are limiting the utility of the tool because you are limiting the potential diversity of the crowd.

Most individuals are not well informed, or believe that being learned is the same as being objective. Modern social research confirms what many know instinctively: most people join and then believe. In other words, people tend to find groups they are comfortable with and then take whatever that group says as gospel (no pun intended). In an increasingly complex world it is not possible to be well informed without constantly considering divergent and seemingly contradictory points of view. But by its very definition a group is designed to ensure that all participants believe in a way that is coherent with the group's continued survival. And in many cases where people “believe then join” the individual's expertise is taken as God's truth (again, no pun intended). Most experts are in fact zealots for a certain point of view, and as recent research uncovers, partisanship is a filter though which all information flows, reinforcing the ancient belief that we are each creators of our own reality. Again, as you constrain the size of the crowd, you exacerbate the basic problems associated with this issue.

In small content arenas, there are clearly market-makers for information. What Jason Davis, John Sumser, Anthony Meany and Jim Durbin (to name just a few) say matters more than what the average recruiting blogger posts. This is not surprising since each of the individuals named has put a lot of time and effort into creating a market for their opinions (as have I, though not as successfully). In content markets celebrity matters… a lot. When Tom Cruise says that “Brooke Shields is being hoodwinked by modern psychiatry” it changes the way people the world over think about psychiatry, and consequently, the types of content they think are worthy of aggregation. This is at direct contradiction to the basic premise of wise crowds: a lack of an echo chamber fueled by mass market celebrity. And, again, the smaller the “mass of the market” the easier it is for one person to sway the crowd in their favor.

And finally, Digg is not an objective method for aggregating opinion, as recent attempts to game the system have clearly shown. It is not a “one person, one vote” system. It is a system that rewards partisanship and individual ambition over collective wisdom.

The logical question to all of this is whether the Digg model can work for smaller domains. That would be entirely dependent on what you mean by “works.” If you define success as pushing the envelope and trying to expand the possible community, then the answer would seem to be "yes."

On the other hand, if you define success as expanding the strength of the connections within the community, the relative quality of the information available to that community and the overall strength of the information markets associated with that community… well, the jury is going to be out for a while on that one. If I had to guess, I would say that you will see some people profit wildly from such innovations and some people decide to turn to other channels to build relationships and the content that flows from them.

In the meantime, I'm going over to YouTube to check out that little guy who dances like Michael Jackson. Now that's pretty cool.

Comments

[Recruitomatic]

I came across this last week and thought it an interesting read, somewhat related:

http://www.rashmisinha.com/archives/06_01/digg-crowd-psychology.html

Sometimes I think we overlook that other people’s blogrolls, for example, are a better read of what people are reading than who digs this and who tags what. Clearly, when it comes to the consumption and ranking of content, not all interactivity is as overt as we find among the crowd, is it? And sometimes, forensics is more interesting to an enquiring mind than fashion is a guide to anything. Yes, sometimes good content does require a lot of digging for.

Great post, Jeff.

[Colin Kingsbury]

Digg's technology content is really good, though as the "gene pool" gets diluted with growth into less-savvy users the links become more banal. But it's still good enough that it's pretty much stopped me from browsing sites like Slashdot and Ars Technica on a regular basis.

Outside of technology, their content stinks. Therein lies a lesson.

[Jeremy Langhans]

Leave it to Jeff Hunter to have readers with solid vocab ... I actually had to look up "Banal"

http://wordnet.princeton.edu/perl/webwn?s=banal

Jeremy Langhans
Sr Sourcer - Human Resources
Panasonic Avionics Corporation
Direct: 949-462-1458
26211 Enterprise Way
Lake Forest, CA 92630
Jeremy.Langhans@panasonic.aero