Main Area and Open Discussion > Living Room

How Digg Gets Everything Backwards.. And How to Fix It

<< < (8/9) > >>

JavaJones:
Note: all of these previous replies of mine were made without having actually read a lot of the stuff above because I was looking at an older version of the post/thread. :D

Anyway, interesting Mouser, the more I read about this stuff, the less I think agree with you. :D In fact I think the simplest fix for Digg is just to remove the reward for "Power Diggers". In this case, since there's no financial incentive (generally speaking), the reward is notoriety. So take away their fame, their profiles, etc. What matters (or should matter) there is not *who* Dugg something but *that it was Dugg* by enough people to get it to the front page. I grant that this doesn't address the problem of people being attracted to good headlines alone, or sensationalism, but I think a much more fundamental issue is the "elite" problem discussed a great deal in these articles. I think the Shmula.com "Digg as a game" article has some very good stuff to say on this which also happens to agree with me:
My Proposal

If Digg were to be relevant again, it must tackle the problems of GroupThink, Conformity, Paradigm Shift and Information Cascades.
To tackle Groupthink, make it truly democratic again — do not profile Top Diggers or elevate anybody higher than anyone else. This includes no special weights on previous digging history, etc. — level playing ground for everyone, no monarchies or philosopher-kings.
To tackle the problem of conformity, do not show profile or # of votes for up-and-coming dugg articles. Just show the article link, with no profiles or votes attached to it. As a compromise, only show the profiles and votes on the articles that make the digg front page, but make them un-diggable from the front page.

Doing the above 2 items will most likely fix the Paradigm Shift and Information Cascade problem.
--- End quote ---

That is coming from someone versed in game theory, and as he shows Digg can be very well visualized using game theory principles. So it stands to reason it could be fixed using similar analysis.

The idea of a representative voting system for the experts means that users can vote on (or rate) experts in much the same way they currently rate stories has a number of benefits.  Voting on experts based on their long term editing choices seems more rational and likely to lead to considered decisions as opposed to instantaneous mass voting based on a glimpse at headline titles.
--- End quote ---

This is interesting - who is to say people won't just positively rate bad reviewers? This just makes the choice of reviewers as big a problem as the choice and voting of stories, IMO. It shifts but does not solve the problem. The problem, at least beyond the "elite" issue discussed above, is that people are basically stupid and easily influenced. They like vapid entertainment and are headline driven. In fact isn't this almost how Digg works already, just more formalized? How else do people become "elite" except by implicit "votes" of support? How can you know the same flawed social dynamics will not govern the influence of your "elite editing body"?

You are trying to solve a fundamental problem of humanity with a rearrangement of the pieces, while still maintaining some semblance of democracy. I say if the people are the problem (and they are :D) then forget about it, just make it a benevolent dictatorship: problem solved! :mrgreen:

Meanwhile StumbleUpon doesn't really seem to be trying to do the same thing at all. It operates on the same fundamental principles, but - aside the "buzz.stumbleupon" page linked above, which is acknowledged to be "not the point" - it is focused more on suggesting things based on user history and predicted preference while browsing. It essentially takes the direct user element out of the equation in terms of choosing stories to look at. Obviously you can choose to look at something else if it suggests something you don't like, but without a "front page of headlines" it's really trying for something much different than any traditional media. Much more like viral marketing in that the majority of it happens person-to-person and behind the scenes rather than collected on some aggregator ("front page") like most news outlets. So while there may be lessons to be learned from SU (don't let users vote from front page, for example), the goals are different so a direct comparison isn't really useful IMO.

I must say I'm very interested in the influence on popularity of each of these sites' various approaches to these things. Is Digg popular because of these problems? I see it as a distinct possibility.

- Oshyan

app103:
Take away the profiles.

List all links in the order in which they were submitted.

Votes are a sidenote and do not elevate the position or order in which links are displayed except on a top 25 page buried elsewhere on the site, after voting is closed (yes, you can only vote for a limited length of time).

All links start out on a today's links members only page and are pushed to the front page for all to see by experts who evaluate links submitted and watch the activity in the comments section of each post, either approving and pushing or rejecting (much in the way it is done with the DC blog)...rejected links staying on the today's links members only page. Spam links are of course deleted.

All commenting is done anonymously, with only the internal system able to identify who the commenter is. Nobody will know who submitted the link. The identity of the submitter only matters if the content is inappropriate. (almost 4chan /b/ style)

wait a minute...except for the anonymous commenting and lack of profiles, this sounds like Fark.com  :-[

And on Fark, submitters are encouraged to make the most sensational headline descriptions for links as possible. Everything is done with a sense of humor. And if you want to see more links and access the members only area, support the site and become a paid member.

mouser:
A very interesting new article that focuses on what i have been saying for a long time is the key problem, that digg (like google) incentivizes bad behavior:
http://www.shmula.com/216/digg-a-lesson-in-freakonomics

the key to fixing these things has to be removing the incentive for gaming the system, because quotes like this from digg founders (and the similar stuff is said by google) is just wrong:

We have sophisticated anti-gaming processes. We are spending a lot in R & D to prevent gaming. Motivations don’t matter.
--- End quote ---

just like google's adsense has now become a playground for click fraud, and has created an entire shadow web of fake websites designed to get search engine traffic, until you fix the incentives, you are turning our internet into an engine for manipulation and gaming by those who make money by figuring out how to trick algorithms into bringing them traffic.

JavaJones:
Agreed. Removing the incentives isn't always easy, of course. Especially as in the case of Google where money is involved. It should be fairly easy for Digg though. But I would guess they realize a good part of their popularity seems to stem from the very same things that make gaming such a problem: social incentives, status, etc.

Btw interesting that the Shmula post essentially recommends commodifying Digg for a similar reason to that which I think is at work here with the DonationCredit system. I still intend to post more on that one day...

- Oshyan

Mizraim:
After reading the whole post, I still don't know what digg is... did I miss it or am I just naive?

Navigation

[0] Message Index

[#] Next page

[*] Previous page