"though the mob clearly lacked the ability to see the potential results of their actions -- or care about them."

Please explain this statement more thoroughly.

Mike, that's simple: it's much easier to suggest that someone else should put their livelihood on the line to stand up for a principle than it is to put your own livelihood on the line for that same principle.

Digg stands a greater-than-zero chance of losing court battles over the AACS key, as they explained on their site, but a horde of people with no skin in the game decided to make a stand, most of them not even considering that it could literally mean the end of Digg. That these people were not acting completely rationally is as obvious as the fact that most of them won't see it that way.

I certainly hope that Digg doesn't lose, but it was a stupid display of bullying by a mob that even makes that a possibility.

Very well written. I have only one minor issue with it: the recommended-filter at reddit will theoretically over time learn what stories you like and present them to you even if the rest of the users didn't like them. Though admittedly it's still far from perfect.

Great piece, Phillip! I like the fact that Netscape is the most transparant about the "up" and "down" votes that stories get. Each is displayed on every story, so in theory stories aren't "penalized" by a down vote, though I'm sure enough down votes trigger some kind of review by one of their anchors.

One of the most intriguing things about social news sites to me is that in the end they are made up of the people who frequent them. Thus Digg is and probably always will be tech-centric, Netscape has a broad American cross-section sort of feel, and Reddit is more of an eclectic hodge podge.

Ksero, the recommended filter at Reddit seems particularly awful, and has a really bad unintended side effect: people end up using their down-votes to train the filter, which causes them to go out of their way to downvote things they would otherwise ignore.

Basically, the down-votes do two things: one for the user, and another for the site readership. That's unfortunate, because people end up thinking about only one or the other.

Want to actually harness the crowd? Then use Subvert and Profit... we sell front page placement on Digg. See my URL.

Another benefit of anonymity, as I see it, is that it makes ventures like that advertised in comment #6 difficult or impossible to operate. If they can't know whether or not you've voted on an article, they can't really pay you, which means you won't do it, which means they can't sell it.

And only allowing votes from semi-random pages of links may make it difficult for people to vote on specific stories anyway, which would be another nail in the coffin of these types of operations.

I would like to say: bravo! This is an excellent follow-up to the recent Digg controversy, and delivered with intelligence and even-handed appeal. My own article on the issue is certain more one-sided though I am pleased it provoked some thoughtful reactions. I largely agree with your suggestions, but must (unfortunately) point out that they may be naive in nature.

As much as I would love to see anonymity (to avoid things like: Kevin Rose's Diggs all making the front page) I just don't see it happening. At the end of the day, these are all for-profit sites watching out for their bottom line. To my mind this model of social news can and will only change if and when these sites are run by and for the people, as opposed to being controlled by a few individuals in power.

Sound familiar? Yes, it is the classic political dilemma of democracy! My best hope: people will begin to see these problems in a new light and start to form organizations around such realizations, which in turn means having joint control over sites. Too simplistic? Probably. A nice idea? I'd like to think so.

Yes, I'm sure that these suggestions won't be taken. The methods that would *actually* harness the wisdom of crowds would be unpopular, so the sites will settle for *pretending* to harness the wisdom of crowds, while actually setting up cliques.

I'm not naive; I recognize this; my point was that harnessing the crowd *is* possible, and these sites aren't doing it. And no, they won't.

Well written. Considering the systems at the level of granularity of the individual may prove to be problematic however because it is very likely that users interact and form social groups formally or informally external of the system. These groups may be able to influence certain content, promoting it or demoting it within the system, without any external appearance of group behavior. Discerning this behavior may be difficult because many if not all, of these systems fail to provide means of internal communication between their members, leaving them forced to use email, IM, or some other form of communication exclusively, all mediums that are untraceable by the internal architects and the external community at large. (Not that I'm suggesting we the services implement them, but rather that it is a factor that plays into our inability to understand the social forces at work).

Nate, the real key would be to prevent people from picking and choosing what links they could vote on. Either present them randomly, demanding an action one at a time (StumbleUpon style, I believe), or present them in a random listing, but don't let people decide to vote on a link their friend told them about.

Of course, that would make the site much less pleasant to use, so it'll never happen. In which case we ought to stop pretending that we're trying to harness the wisdom of crowds and admit we're just hoping the popular clique likes the same thigns we do.

I must interject here that I believe StumbeUpon has succeeded in at least minimalizing the "mob" effect... What pages it sends you are effected by your interests, rating behavior, and voting habits. To see what other people have said about it, you must load the reviews page, almost totally eliminating the mob mentality while still being very, VERY good at providing pages the user is actually interested in.

I must admit I haven't spent very much time at StumbleUpon at all. I mostly spend my time on Digg and Reddit, with a bit of Netscape. I used to spend time on ShoutWire, too, but that was some time ago.

I've used StumbleUpon for Videos, but that's about it.

Ah, what I wouldn't give to see some good old fashioned mob rule in this country! The great thing about a mob - regardless of its relative level of wisdom - is that it refuses the dictates imposed on the individuals it's made up of, that are handed down by a small group of power-hungry corporate rulemakers.

The question of crowd wisdom vs. mob rule is very relevant in terms of the political climate online. The internet is supposed to be bringing democracy back to the Presidential vote, as I discussed in a recent post on my blog. But what I'd like to see is some of the power on display during social events, like the whole May Day thing on Digg, turned into actually getting underdog Presidential Primary candidates the attention, and funding they deserve. When will the mob power we occasionally see on display online translate into real-life activism in this country?

Anonymity will always undermine credibility. I don't mean to assert that no useful information can be learned from anonymous indiviudals, it's that a crowd full of geniuses is of little use if it cannot be distinguished from a crowd full of idiots. How about this: Organize the mob into small groups, then make certain the participants in those groups don't change quickly. Keep the same people with the same people. Smart groups will emerge, dumb groups will emerge, and the crowd, in its wisdom, will hopefully learn who to watch. An idiot can lower the IQ of a crowd, but they can't make a smart person dumb.

On comment #2:

For the first point, though it is easier to get someone else to stand up for a principle than oneself, normally few people would even do that. Few people who visit Digg would stand a chance trying to defend themselves in court, but by demanding a larger organization get involved something might actually get done. Additionally, just the fact that a large number of people showed their dislike of the current situation shows where popular opinion stands.

Though most people on Digg were just following the crowd rather than putting some thought into what they were doing, some believed that the exposure would be worth the risk. If this goes to court, and Digg wins, it would likely have a significant effect on anti-copying laws and their enforcement. If Digg does end, there are a number of other sites its audience could go to, e.g. Reddit, Slashdot, etc.

I agree with your article in general. Very good points, and well written. The big problem with this idea of a 'social network' being more wise (or, probably better put, less corrupted) than some group or organization is fundamentally flawed. I didn't realize what the actual 'experiment' was that was used as a basis... so your article really enlightened me there.

However, I think it is flawed in basic principal... for a 'social' network to be 'smart', they would have to.... 'think'. I would argue that this generally does not happen. While panels and organizations can certainly be corrupt, there is peer review, research, and usually thinking going on. Yes, this can still be corrupted or wrong, but I think is right more of the time than the 'popularity contest' that often takes place in 'social' networks.

I often see this problem present itself in reference to government systems. The idea that democracy in its pure form would somehow be good, ignores basic human nature. 51% should never rule. This is way, at its best, the USA is not a Democracy, but a Democratic Republic with a Constitution. The Constitution sets fourth some basic principals that keep human nature in check, and ensure that 51% can't overrule the basic principals.... such as 'All are created equal.'

ANY social network needs checks and balances to keep human nature in-line. Modern Internet 'social networks' are no exception. Organizations, groups, committee, etc are in the same boat. If the checks and balances are not strong enough, or fail, human nature will corrupt the results... no matter how big or small the group, or an individual.

-Steve

This was not only an excellent article, but long overdue to be written. How wise is the crowd? Even if the crowd is wise, it is always wise to yield to crowd wisdom? Are these tools even measuring the crowd correctly? And you're right, the way you describe how people use Digg is exactly how I do it when I find myself there, rather than following a DIgg button to the site.

This makes me want to go watch Gladiator, and write down all the quotes about what Rome is. That's often the ultimate question for me whenever I run something (like a forum) that's subject to the whim of the crowd as a large part of its success. Is Rome the mob? Is the mob fickle or of a larger intelligence than an individual? Food for thought, thank you thank you thank you.

Submitting content to Digg is easy; you just set up a Google Reader account with feeds from the most popular geek sites and check in on it every hour or so. When a new story hits one of those sites, you submit it to Digg.

Digg could eliminate the 20% altogether by automating this process. Just take an algorithm-driven aggregator, ala Google News, and make that the upcoming stories section of a social news site.

Problem solved. However...

I would like to see a community website that awakens human consciousness. That makes us better people. Digg seems to do the exact opposite.

When the masses turn the group vote into a popularity contest, you get people going onto sites like digg and molding the environment in their own image. People do bury things for less than noble reasons, one of which might be to get their own articles to the top.

Many commenters above are missing the point suggested by Francis Galton's observation. The point isn't that we need to separate the experts from the morons, nor that we need to able to identify everyone. In fact, both of those statements are taking, I believe, the exact opposite of what they should from the experiments.

The point was that while none of the non-experts in themselves knew what they were talking about, in aggregate they came up with the right answer, beating out the experts. I'll say it again, because some people missed it in the article: non-experts -- in aggregate -- came up with a better answer than experts. Your wrong answer and my wrong answer and the wrongs answers from a bunch of other people average out to something really close to the right answer.

Once you grasp the implications of that, you see why identity isn't an issue, any more than expertise is. The problem with expertise and authority is that people with it tend to rely on it, and it turns out that experts and authorities makes mistakes and have bad days, too. The idea of a large group of people is that a sufficiently large swarm of people ends up getting the right answer. Not individually, but collectively.

Digg, Reddit, Netscape, and the others, are all relying on experts, and think that because they let people *establish* their expertise on-site rather than carrying it over from the real world, this is somehow a brave new world. It isn't. It reward those with the most free time, but fails to take advantage of the very issue that it claims to rely on -- the wisdom of crowds.

Great food for thought!

I see a possible solution as a "trust" system for individual voters. Something that could downgrade the impact of those who vote only to vote, vote far too often, or too consistently for one contributers content. Part of the problem would be self-correcting (as in eliminating whackos) if a person just looking to mash buttons did not see immediate results.