1. N.Z. Bear on November 30, 2003 4:02 PM writes...

Clay -

Strictly speaking, the example you quoted represents the "last straw" in a series of behaviors I considered abusive with regards to the Ecosystem. To be precise for your readers, the quote above actually refers to a lesser-known part of the Ecosystem which provides rankings based on SiteMeter visit counts (as opposed to based on links). (That ranking can be found here: http://www.truthlaidbear.com/TrafficRanking.php)

However, the individual in question has also demonstrated a history of attempting to manipulate the traditional, link-counting part of the Ecosystem as well, so your point remains quite valid.

I continually am struggling to provide a service which both offers a simple, easy-to-understand "map of the territory" of weblogs, but at the same time is resistant to abuse. With this latest incident, I'm afraid, no elegant programming solution was available, and so I've had to take a more crude approach and directly limit one individual's participation in the ranking system.

Solution unsatisfactory, indeed: but the best I can do at this point...

Best regards...

-NZB

2. Lucas Fletcher on November 30, 2003 5:36 PM writes...

I am curious to what extent authors link to alpha blogs in order to have a chance of "getting in on the action", having the alpha blogger take note of the link and link back at some point. Call this the suck-up factor, and if it exists would be yet another feed-forward factor.

3. PZ Myers on November 30, 2003 6:06 PM writes...

Like this, you mean? http://gevkaffeegal.typepad.com/the_alliance/2003/09/the_blog_war_an.html

Even us delta/gamma/epsilon/wherever-the-heck-way-down-I-am webloggers get e-mail requests for reciprocal links, and clearly there are a lot of pathetic people out there who want an "instalanche". What's really strange is that people can consciously set up systems that rank sites on the basis of the number of links or number of visits or other such trivially quantifiable parameters, and then express dismay that participants strategize to increase those parameters. When you define one group of people as "higher beings" and another as "insignificant microbes" on the basis of links, of all things, you don't get to call it a nice neutral term like "map of the territory", and you shouldn't be surprised when people struggle to get out of that microbial ghetto by increasing the number of links they get.

4. The Commissar on November 30, 2003 7:09 PM writes...

There is a perfect example of "alpha bloggers," right here. No, not with the "scandal" itself, but with its coverage.

I am "The Commissar," referred to in NZB's post. I mentinoned this to NZB and he then took appropriate action. All this occurred within a farily small group of "non-alpha blogers." (Although NZB himself may be an alpha.)

But now, with Jeff Jarvis and Clay Shirky "on the case," it seems to be a biger deal.

A perfect example of the role of "alpha bloggers."

BTW, I am very happy in my role as a "gamma" or "delta" blogger, and do not begrudege Jarvis nor Shirky their prominence.

5. Ross Mayfield on November 30, 2003 7:25 PM writes...

Lucas, what you are talking about is Preferential Attachment. We connect with connected nodes to gain the benefit of their connections.

And Feedforward and Feedback have huge implications for the emergent properties of the system.

6. N.Z. Bear on November 30, 2003 8:29 PM writes...

PZ, you said:

"What’s really strange is that people can consciously set up systems that rank sites on the basis of the number of links or number of visits or other such trivially quantifiable parameters, and then express dismay that participants strategize to increase those parameters."

Well, yes, it is dismaying, when the "strategies" are clearly cheating, as (in my opinion) the SiteMeter games at question here were. (If you want the full description, head to my site and scroll, if you have the time/stomach for it). I mean, taken to its logical conclusion, doesn't your statement suggest that people should expect, and not be dismayed at, people cheating at *any* form of measured, competitive endeavor? Should we have no standards of fair play --- and no dismay when those standards are violated?

"...you shouldn’t be surprised when people struggle to get out of that microbial ghetto by increasing the number of links they get."

Where did you ever see me express "surprise" at this behavior?

Surprise, no. Disappointment, yes. As I said before, I try to run a fair system. Does it surprise me that some few individuals, out of the thousands who participate, try to push the limits of acceptable behavior and cheat? No. Frankly, it surprises me more that I haven't seen more of this kind of nonsense.

Anyway, I've been the first to say that the Ecosystem shouldn't be taken too seriously. (Literally. Quote from my very-first post about it, over a year and a half ago: "if you're gauging your own blog's success by any measure other than how much enjoyment it gives you, well, that way lies madness") But even if a game is meant to be silly, and for fun, that doesn't mean that those playing shouldn't be expected to exercise a sense of fairness in how they play.

7. snr on November 30, 2003 9:02 PM writes...

Whenever there's a perceived benefit, there's incentive to cheat. The penalty (cost * likelihood) needs to be large enough to outweigh the benefit to effectively discourage cheaters. It's human nature, it's who we are. If it has value & we think we can get away with it, we cheat. Expecting fairness in the absense of enforcement is unrealistic.

s/n:r

8. PZ Myers on November 30, 2003 9:23 PM writes...

I'm not saying that cheating is good, but that in your system there is no reward for fair play, or for other less quantitative measures. When the only metric you use is the link or the visit, and you rank them by that metric, you are going to incite competition. The people playing also perceive rank as having a value. I note that someone has commented on your latest post, saying "In fact, when I've tried to introduce new people to the world of blogs, I often point them to the main page of the Ecosystem, so they have a list of the top sites." It's being regarded as more than a silly game by a lot of people.

I would think that if you actually thought it was all just for fun and silliness, you wouldn't be at all concerned if a few people discovered weird little tricks to put themselves in your upper ranks. So what if they play around and get a "Higher Being" title? It's just a game. What harm have they done? Have they diminished anyone's success, or improved their own? Maybe it would give everyone a better sense of perspective if a few "Insignificant Microbes" displaced Glenn Reynolds from your top spot now and then.

9. Lucas Fletcher on November 30, 2003 9:33 PM writes...

If we agree that the number of incoming links is a "trivially quantifying parameter" in assessing the value of a blog then we should celebrate the gamers, at least in-as-much as they are adding noise to the system and thus decreasing any vestigial association between link popularity and worth.

Maybe when it becomes totally useless people will start to write what they really think, and not what they need to say to become popular. At that point there may be enough variety within the now stullifyingly homogenous blogworld to justify personalized search in leueu of the high-school mentality atavism that now reigns.

10. Clay Shirky on November 30, 2003 10:06 PM writes...

_Maybe when it becomes totally useless people will start to write what they really think, and not what they need to say to become popular._

This is not a likely scenario.

11. PZ Myers on November 30, 2003 10:22 PM writes...

Pessimist.

Don't you at least think it would be interesting if the gamers were successful, and hordes of weblogs became ingrown nests of self-referential links sprinkled with search-engine friendly buzzwords, all designed to maximize server load without saying much of anything? There is a kind of twisted evolutionary lesson lurking in the idea.

Oh, wait...I think I just described many of the 'top blogs'.

12. snr on November 30, 2003 10:25 PM writes...

At that point there may be enough variety within the now stullifyingly homogenous blogworld to justify personalized search in leueu of the high-school mentality atavism that now reigns.

A network is a mathematical function. If you want to change the distribution of connectivity, you need to know what knobs there are to be twiddled and how twiddling one affects the others. At any one time you can only have 2 out of 3 of Diversity, Equality & Freedom. Dampen one and the others will flourish.

Which 2 do you want?

s/n:r

(much thanks to Clay for pointing this out at http://shirky.com/writings/fcc_inequality.html)

13. Clay Shirky on November 30, 2003 11:08 PM writes...

There will always be an incentive to use traffic numbers in public speech -- "Well, I have 10,000 readers..." This in turn creates the twin incentives for auditing (hence Sitemeter and, at a higher level NZ Bear's work) and gaming (hence the people abusing Sitemeter to get higher on N.Z. Bear's list.)

However polluted the current system of counting traffic becomes, I don't think we'll ever get a system where the _desire_ for real traffic numbers goes away. People respond to incentives, and attention is an incredibly powerful one.

14. snr on November 30, 2003 11:29 PM writes...

"However polluted the current system of counting traffic becomes, I don’t think we’ll ever get a system where the desire for real traffic numbers goes away. People respond to incentives, and attention is an incredibly powerful one."

So they need an equally powerful disincentive to outweigh that. Removal from the status charts is pretty powerful, as long as they think it's likely to happen.

It's no skin off my nose, but some form of enforcement function is needed if you want to keep people from cheating. And it needs to be advertised. If you want to lighten the load on the owner, you could open an email line for people to report suspected cheaters & maybe use a couple volunteers to check them out & make judgements.

It's a lot of hassle to protect against childish behavior, but if you want it to be effective it's what you need to do.

s/n:r

15. Clay Shirky on November 30, 2003 11:41 PM writes...

s/n:r, not arguing against what NZ is doing -- I think that's the rigth answer. I was responding to the idea that if the gaming continues, people will somehow lose interest in traffic rankings.

16. snr on November 30, 2003 11:54 PM writes...

Yeah, I'm in complete agreement. As long as the ranking is anything more ordered than complete randomness, some will see value in the status to be gained from it. Hell, I wouldn't be surprised if people found value in high numbers even if they were handed out completely at random, just so they could say "mine's bigger than yours".

s/n:r

17. Lucas Fletcher on December 1, 2003 1:17 AM writes...

Clay, I agree with you that the incentive of gaining others' attention is a powerful one and will never go away. What I was suggesting (in a round-about manner) is that traffic and link rankings are not a good way to align the incentive of the blogger to drive traffic with the incentive of the reader to read quality content. I suggested personalized search as a better mechanism. It would have the effect of incentivizing bloggers to tailor their content and thus become more honest and individualistic. (What better a way to personalize content than being yourself.) In a world of aggregated search these traits may lower the average value of the blog across readers but with personalized search it would increase it. (BTW, the only thing in the way of this inevitable technology is that it is orders of magnitude more complex and expensive than what we have now. But More's law etc...)

18. nick on December 1, 2003 1:26 AM writes...

I just wanted to mention that what is called 'feed-forward' in the post would actually better be called 'positive feed-back' and what is called 'feed-back' more accurately called 'negative feed-back'

Feed-forward doesn't really have a meaning in control systems, but it does in other applications like artifical neural networks ;)

19. nick on December 1, 2003 1:30 AM writes...

ok my bad...

http://www.sapiens.itgo.com/control/

feed-forward control systems DO exist, but they don't do what is described in the post, that IS better called positive feedback (and is what causes annoying squawking in speakers when you bring a mike close to them for example)

20. Meme Engineer on December 1, 2003 3:16 AM writes...

"If it has value & we think we can get away with it, we cheat. Expecting fairness in the absense of enforcement is unrealistic." - [snr]

This is insulting. Speak for yourself. Maybe _you_ cheat, but that has nothing to do with _other_ people who actually value precision and accuracy.

21. Frank Ruscica on December 1, 2003 8:00 AM writes...

One promising solution is to tie rankings to an objective measure. For example, if bloggers who are focused on the upcoming Presidential election were at the same time "investing" in the Iowa Electronic Market for the upcoming Presidential election, then blogs could be ranked by performance in the market.

Such rankings also add value because most A-list bloggers, as theory predicts, grow through first-mover advantage, but are actually not very interesting to a more mainstream audience (e.g. people who want news, not primers on social network theory, complexity theory, etc. -- for them, the network may be the computer, but it is not the news!).

Hmmmm...

This suggests that the A-list bloggers are ripe for disruption, via a shift in the basis of competition.

Objective metrics seem the likeliest enabler...

Where to get the metrics?

Investment portfolio simulation challenges for investment-focused bloggers; HSX for the entertainment-focused; Sourceforge; Slashdot; Amazon, DeanLink, ...

Hmmm...

This is really a textbook disruption waiting to happen. A-listers aren't thinking about innovations in assortive matching because they are busy serving their best customers: other A-listers. As such, the only innovations discussed are those that are sustaining in nature: RSS 2.0, Atom, services that priviledge the well-connected, etc.

But what about people who just want the best info in a given subject area!?

They are brutally underserved by the distortive nature of popularity derived from accidents of history and the politics of personality...

Hmmm...

22. Clay Shirky on December 1, 2003 8:02 AM writes...

Meme Engineer, I think s/n:r is is using we in the sense of people, as in "systems that create unpoliced value lead to cheating." He didn't put it in the passive tense, but I don't think he meant everyone cheats, but rather cheating in such situations is inevitable.

23. N.Z. Bear on December 1, 2003 9:05 AM writes...

Incidentally, folks, I'm also looking at the possibility of expanding my effort to allow "sub-ecosystems" focusing on individual communities to be formed by turning my Ecosystem into something resembling an open-source effort.

While this would not alter the fundamental nature of these systems (they'd still be ranked based on links, with the incentives that causes), I believe that a network of community-based listings would be far more interesting than my single undifferentiated ranking.

Details and a request for feedback / assistance here: http://www.truthlaidbear.com/archives/2003/11/27/let_a_thousand_ecosystems_bloom.php#001231

24. snr on December 1, 2003 10:11 AM writes...

"He didn’t put it in the passive tense, but I don’t think he meant everyone cheats, but rather cheating in such situations is inevitable."

Some people have a strong personal ethic and won't cheat under any circumstances. Some don't and will cheat at the drop of a hat. When you're working with enough people to exhibit a Power Law curve, you're sure to find some of both. me, my ethic's pretty strong. But we wouldn't be having this discussion if cheaters didn't exist.

s/n:r

25. Lawrence Krubner on December 7, 2003 2:16 PM writes...

You don't mean Faulkner, you mean Fitzgerald. It's uncertain that he ever said it, but Hemmingway claims Fitzgerald said, "The rich are different from you and me," to which Hemingway replied, "Yes, they have more money." It's possible that Hemingway invented the exchange in an attempt to injure Fitzgerald's reputation. It's also possible that it's true.

26. Lawrence Krubner on December 7, 2003 2:39 PM writes...

"who are, by tradition called the alpha bloggers"

Could we please, in the future, write this in this way:

"who are, by convention, called the alpha bloggers"

It seems like quite a stretch for the word "tradition" to refer to conventions that have existed for 9 or 12 months. It is a tradition to give gifts to friends and family on Christmas, it is a convention to refer to high-link bloggers as "alpha-bloggers". A tradition is a convention that's been around for a few centuries.