Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#1

Post by ajbatac » Mon, Jul 07, 2008, 1:34 AM UTC

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#2

Post by ojbyrne » Mon, Jul 07, 2008, 1:58 AM UTC

On the scale of bitching about digg (which I'm known to do :-)) this one doesn't rank. Given the various power laws out there, I'd say that it's actually pretty diverse.

The people at the top, for whatever reason, work very hard to get there.

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#3

Post by aswanson » Mon, Jul 07, 2008, 1:59 AM UTC

Nothing new: http://en.wikipedia.org/wiki/Power_law

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#4

Post by pg » Mon, Jul 07, 2008, 2:06 AM UTC

It looks like the number on News.YC is about 26%.

  > (let bys (map [_ 'by] (beststories nil 1000))
      (/ (apply + (map cadr 
                       (firstn 10       
                               (sort (compare > cadr)
                                     (dedup (map [list _ (count _ bys)] 
                                                 bys))))))
         1000))

  .258

This is the fraction of the 1000 top-scoring stories in the past month or so (http://news.ycombinator.com/best) submitted by the top 10 submitters. Who incidentally are:

  nickb    77
  edw519   34
  markbao  26
  bdfh42   20
  nreece   19
  ajbatac  18
  wumi     17
  prakash  16
  smanek   16
  ilamont  15

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#5

Post by dhotson » Mon, Jul 07, 2008, 2:07 AM UTC

This isn't all that surprising to me.

I would guess that there a few groups of users, with approximate proportions:

Lurkers: 70% Diggers: 20% Submitters: 10% Regular Submitters: .. is there anyone from Digg here that can confirm this? :)

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#6

Post by bdotdub » Mon, Jul 07, 2008, 2:11 AM UTC

aswanson wrote: post #3

Nothing new: http://en.wikipedia.org/wiki/Power_law

yup, nothing new. That might even be a good percentage :) http://www.useit.com/alertbox/participation_inequality.html

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#7

Post by aswanson » Mon, Jul 07, 2008, 2:22 AM UTC

pg wrote: post #4

It looks like the number on News.YC is about 26%. > (let bys (map [_ 'by] (beststories nil 1000)) (/ (apply + (map cadr (firstn 10 (sort (compare > cadr) (dedup (map [list _ (count _ bys)] bys)))))) 1000)) .258 This is the fraction of the 1000 top-scoring stories in the past month or so ( http://news.ycombinator.com/best ) submitted by the top 10 submitters. Who incidentally are: nickb 77 edw519 34 markbao 26 bdfh42…

For some reason, that code sample makes me want to investigate lisp.

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#8

Post by froo » Mon, Jul 07, 2008, 2:39 AM UTC

pg wrote: post #4

It looks like the number on News.YC is about 26%. > (let bys (map [_ 'by] (beststories nil 1000)) (/ (apply + (map cadr (firstn 10 (sort (compare > cadr) (dedup (map [list _ (count _ bys)] bys)))))) 1000)) .258 This is the fraction of the 1000 top-scoring stories in the past month or so ( http://news.ycombinator.com/best ) submitted by the top 10 submitters. Who incidentally are: nickb 77 edw519 34 markbao 26 bdfh42…

Hey Paul,

I speak to JD all the time (he wrote the article) and we've been looking at these numbers again every couple of hours for the last couple days, and the 31% that he wrote about (it hovers around the 30% mark) is fairly consistent when looking at the last 500 frontpaged stories.

One thing he didn't write about was that at the time if you expanded it out to top 28 users, that number was around 51% of Digg's frontpage stories.

Around 5am this morning (for me, about 5 hours ago), we ran the numbers again and at that particular time, one user was accounting for 53 of the last 500 frontpaged stories on his own.

Not going to say the data is conclusive or anything, because I frankly just dont know, but when you consider Digg says they have something like 23 million uniques a month and well over 1 million registered users, 50% of their frontpage traffic being dictated by less than 30 users is a little strange.

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#9

Post by wallflower » Mon, Jul 07, 2008, 2:46 AM UTC

Related reading: "The murky demimonde of Amazon's top reviewers"

"Amazon's rankings establish a formal, public competition for power or its online equivalent, recognition...

To the extent that competitive energies drive Top Reviewers and their nemeses to generate content, and to spend time on and publicize Amazon.com, the chief beneficiary of misuse of Amazon.com's ranking system is Amazon itself."

http://www.slate.com/id/2182002/pagenum/all

Re: Recommended Flaw: 31% of Digg Homepage submitted by 10 Users

#10

Post by pg » Mon, Jul 07, 2008, 2:47 AM UTC

froo wrote: post #8

pg wrote: post #4

It looks like the number on News.YC is about 26%. > (let bys (map [_ 'by] (beststories nil 1000)) (/ (apply + (map cadr (firstn 10 (sort (compare > cadr) (dedup (map [list _ (count _ bys)] bys)))))) 1000)) .258 This is the fraction of the 1000 top-scoring stories in the past month or so ( http://news.ycombinator.com/best ) submitted by the top 10 submitters. Who incidentally are: nickb 77 edw519 34 markbao 26 bdfh42…

Hey Paul, I speak to JD all the time (he wrote the article) and we've been looking at these numbers again every couple of hours for the last couple days, and the 31% that he wrote about (it hovers around the 30% mark) is fairly consistent when looking at the last 500 frontpaged stories. One thing he didn't write about was that at the time if you expanded it out to top 28 users, that number was around 51% of Digg's fr…

It doesn't seem strange to me. The top submitters aren't dictating the stories. They're just submitting them first. And it doesn't seem at all unlikely that a few dedicated users could be the first to find a large fraction of the stories that ultimately became popular.