Favstar.fm

jorshuwah:

I took this screenshot yesterday of the top 98 ‘Funny, Entertaining or Insightful’ tweeters as determined by tim’s favstar formula. I was up there for about 6 months or something.

I just thought I’d post this to preserve that moment for time memorial and write a little about favstar - a site that attempts to show the best tweets and the best tweeters. Favstar does this by recording how many stars a tweet receives.

It’s an incredibly daunting task that Tim (the creator of favstar) has: how do you create a formula based on amount of stars received that allows for the best quality tweets to rise to the top?

There are three principal challenges that I see:

• It is a quid pro quo world. People give stars when they receive stars. You give out a lot of stars you will receive a lot of stars back (the exception to this is the celebrity factor which I will discuss as my next point). The quid pro quo reality goes fundamentally against achieving a system where the best rise to the top because it is not the best that rise to the top but rather the people who give out the most stars that rise to the top.
• The celebrity factor that I spoke about in my previous point refers to the situation whereby you have achieved an element of notoriety such that the world is no longer a quid pro quo world. That is, people do not give you stars because you gave them stars, they give you stars because you are a celebrity. I am generalising but this is only because I want to express the challenges that are faced in simple terms. This challenge is real - how do you get an audience not to base their adulation not on the person who wrote it but the quality of the content? (This is further complicated because you don’t want to separate the identity of the author from their content because knowledge of the author can of course add to the content in terms of depth and playing off assumptions).
• The third problem is that people have a different amount of followers. Where someone has over a thousand followers who are aware of favstar and star accordingly and another person who writes a similar calibre of content and has a hundred followers who are aware of favstar and star accordingly it is simply a much more difficult playing field the less followers you have. Although, that being said, all you really need are about 50 dedicated quid pro quo followers and you can still prosper. 

 My suggestions are as follows:

• If a tweet is incredibly similar or identical to a tweet that has already been done then it should immediately be removed from the site. This should help originality to flourish. And in any case if someone has already made a tweet that is identical or so similar to yours that you aren’t adding anything to the pyramid of “humour” why would you want people to waste their time?
• There should be two types of stars. One is a quid pro quo star, the equivalent of saying - thank you for starring me. The other is a merit star - the equivalent of saying this content is worthwhile and I’d like others to see it. (This sounds like a retweet.)

Do you have any ideas to help reduce or avoid the challenges that a site like favstar faces?