follower counts have no purpose here. If Hubski is about intelligent back-and-forth, it doesn't matter that a user has 10 or 2000 followers-- the concern should be on the content they're putting out there.

'Follower count' isn't a meaningless score, like Karma or something, where we can say, "You have 2000? Not gonna affect anything." High numbers of followers drastically change how information spreads in Hubski.

Posts are shared by two things: number of people who see it, and percentage of those people who think it's a good idea to share it.

Let's say I have 10 followers who each have 10 followers, and I post something that 50% of people will think is worth sharing (which would be an absolutely insane popularity level). The same level of popularity holds true for each level of sharing. I will have: 1 original poster 5 first sharers 25 second sharers. So a total of 31 shares. Cool!

Now, I'm sure you can imagine where this will go if I have 2000 people following me, so let's use a real number. mk has 676 followers, and for the sake of simplicity we'll say that each of his followers have 10 followers. If I knew the average I would use that.

Let's say mk posts something which only 10% of people will like- it should be obvious that this is "lower quality" than the 50% shared thing that I posted. If only 10% shares at each level, then mk will have: 1 original sharer 67 first sharers 67 secondary sharers For a total of 135 shares. Wow.

So, in this example mk posted something which is strictly less popular and less likely to be enjoyed/shareable than what I did, and came out way ahead. In fact, even though he made something which was 1/5 as likable, he got 5 times the number of shares!

The kleinbl00 scenario you describe is, I think we can agree, much more of an edge case than what I'm saying. Sure, if we want to continue this to an arbitrary number of users I will overtake mk. But I think that we can both see that having something 50% of people will share is just not going to happen, and I am much more likely to get my 4-5 shares and be happy for it.

And how many of the algorithms on this site are based on popularity? The feed is a combination of popularity and time; the popular posts/comments, obviously; the order of comments in a thread seems to be based on popularity. And there's even more happening here- because as I get more followers, the percentage of followers who think that something is badgeworthy stays the same. If I had 2000 followers, and 500 of them followed me to this point, and then 6 of them thought that this comment was well thought out, I would instantly reach the top 5 of all badged people in Hubski. This would be listed as the most badged content in the site's history (to the best of my knowledge). Even if, percentage wise, the same number of people would badge this as would badge one of the posts in the global unshared page.

Do I want people to have arbitrary scores that they're keeping track of? Not really, no. We're after the same thing. But having a high follower count is not just having a high score- it is a self-fueling love-machine which can overstep community bounds, interest bounds, and drastically change the user experience for everyone on the site. Short of ignoring them, I cannot actually escape the incredible pull of TNG, kbl00 and mk.

I like hubski, but I don't want people to think that it automatically is balanced and perfect. The superusers pose a huge problem to the rest of the site's architecture, and in fact go against the design by their nature. They push too hard and in too many places at the moment, and they cannot be caught up to using traditional 'post good content' means. Even if one of them never posted again, and only shared other people's content, it would take a whole hell of a lot to get ahead of them by putting up good links or good comments.