take a look at what the numbers say

June 3, 2001 2:43 PM

Considering all the hub-bub we keep getting about vets VS oldies, signal-to-noise, volume posters, etc, I just finished wasting a huge volume of time crunching the may stats based on individual posters - all 310 of 'em. For whatever it's worth, take a look at what the numbers say about who's posting what and how successfully. info inside
posted by Perigee to MetaFilter-related at 2:43 PM (15 comments total)

In May 2001, MeFi had 9078 unique possible users;
An unkown number of active users
there were 310 unique thread posters - 3.4% of possible users - who created 608 threads

Of those 310 posters:
1 posted 16 links
1 posted 11 links
1 posted 10 links
1 posted 9 links
5 posted 8 links
2 posted 7 links
5 posted 6 links
8 posted 5 links
11 posted 4 links
25 posted 3 links
61 posted 2 links

The poster of 16 links joined Mefi 3/2000;
(364 total/16 = 23 responses per post average.)

the poster of 11 links joined Mefi 4/2001;
(221 total/11 = 20 responses per post average)

the poster of 10 links joined mefi 11/2000;
(117 total/10 = 11.7 responses per post average)

the poster of 9 links joined mefi 4/2001;
(92 total/9 = 10.2 responses per post average)

poster 1 of 8 links joined mefi 3/2000;
(133 total/8 = 16 responses per post average)

poster 2 of 8 links joined mefi 7/2000;
(94 total/8 = 12 responses per post average)

poster 3 of 8 links joined mefi 2/2001;
(139 total/8 = 17 responses per post average)

poster 4 of 8 links joined mefi 2/2000;
(133 total = 17 responses per post average)

poster 5 of 8 links joined mefi 8/2000;
(156 total/8 = 20 responses per post average)

posted by Perigee at 2:44 PM on June 3, 2001

In total,121 posters set 411 links, leaving 197 single link posters in the month of May.
posted by Perigee at 2:48 PM on June 3, 2001

The top 9 posters account for 1558 responses, or 13% of total comment traffic
posted by Perigee at 2:53 PM on June 3, 2001

poster 1 of 7 links joined mefi 5/2001;
(109 total/7 = 16 responses per post average)

poster 2 of 7 links joined mefi 2/2000;
(194 total/7 = 28 responses per post average)

poster 1 of 6 links joined mefi 1/2001;
(186 total/6 = 31 responses per post average)

poster 2 of 6 links joined mefi 9/2000;
(96 total/6 = 16 responses per post average)

poster 3 of 6 links joined mefi 5/2001;
(60 total/6 = 10 responses per post average)

poster 4 of 6 links joined mefi 4/2001;
(108 total/6 = 18 responses per post average)

poster 5 of 6 links joined mefi 4/2001;
(199 total/6 = 33 responses per post average)
posted by Perigee at 3:33 PM on June 3, 2001

Benchmark for postings -
Average response per post on 5/14 - 21
Average response per post on 5/15 - 19


posted by Perigee at 3:44 PM on June 3, 2001

What does all of this mean - apart from the fact that I really need something better to do with my Sundays?

It means what it is; at least at this point you have some solid numbers to argue your conclusions on all the basic mefi us VS them issues.

Up to 6 link posters, 2000 and 2001 members are of equal number - 8 per side.

all taken together, 2000-joined multiple link posters average 18 responses per post - just below the imaginary benchmark. 2001-joined multiple link posters average 19 responses per post; nearly a dead draw.
posted by Perigee at 3:56 PM on June 3, 2001

Good work, Perigee (I was going to ask about your slow Sunday but you beat me to it).

My only caveat is your definition of a "successful" post. Not all subjects are likely to draw the same number of comments. Some topics are "hot", others are just mildly of interest. Hot topics are not necessarily always good however, I believe there needs to be a mix.

Invoking the 80-20 rule, I would say that on Mefi, 20% of topics attract 80% of the comments. I could be wrong but to be sure but you would need to do an analysis of topic key words.

Things like:
G.W. Bush, DUI, Survivor, McDonalds, Pyra/blogger, movies, anti-globalization, sexual preference, death penalty, Kaycee, kottke, TiVo, Salon, SUVs, blogs in the media, Metafilter, gun control, the U.N., a-list, Pepsi vs Coke, Jakob Nielson, Neale...

I'll stand up and defend my right to post only mildly interesting threads!

posted by lagado at 5:15 PM on June 3, 2001

Sure, Lagado, I'll give you that right. Most of the majorly interesting threads are news stories, and aren't truly what the web ... and mefi ... are about.
posted by SpecialK at 6:13 PM on June 3, 2001

I agree with the poster's moral contention. ;-)
posted by bradlands at 7:19 PM on June 3, 2001

I kept the spreadsheet out, and depending on my ability to manufacture a life, I may take a crack at taking all this deeper next weekend; if I find myself doing it the week after that, I'll make sure Quirked posts my suicide obit. ~grin~

The numbers are what they are, lagado; but what I personally took away from this is the incredible stability of the system so far.

People have argued that Mefi is cliquish; but just as many newbies as vets are sharing the spotlight. One myth debunked, for me.

People have argued that Mefi Newbies are bringing down the quality of links; but Newbie posts and Vet posts are practically identical in user interest. So, another myth down, for me.

People have argued that quantity posters are ruining Mefi; well, quantity posters are right there in the average interest range - some hit, some miss, but overall, there's really no horror in quantity posting. So, another one bites the dust.

As for myself - 3 posts last month, and although I didn't check my numbers, memory suggests that my average was probably somewhere around 5 responses per thread. I am the weakest link, but I'm too damned stubborn to leave the building.

Overall, the study (so far) proved out to me, at least, what I hoped it would; despite the growth in potential users, Mefi is really doing pretty well. All of the us VS. them and witch hunting seems to come under the category of 'monsters under the bed'.
posted by Perigee at 5:17 AM on June 4, 2001

One thing that should be noted is that number of comments doesn't necessarily equate to user interest. There have been a number of posts recently that I've really appreciated... but no discussion ensued because (ostensibly) none was needed. Some posts just "are".

What would be of interest to me, would be the number of times a link was clicked on... I bet that a lot the posts without comment still generate a lot of traffic for the referenced site; likewise I bet that some of our most commented threads generate relatively little traffic for the referenced site. I, for one (and I'm baring it all, here, folks... so be kind) don't always read the link before joining a thread: I have a lot of sources for my information, and a lot of the "hot" threads on MeFi are of a current events nature... so sometimes I just skip the read, and jump right in. I don't do that too often, but I do it. (Please tell me I'm not alone on this!)
posted by silusGROK at 1:03 PM on June 4, 2001

One thing that should be noted is that number of comments doesn't necessarily equate to user interest. There have been a number of posts recently that I've really appreciated... but no discussion ensued because (ostensibly) none was needed. Some posts just "are".

sure, vis10n, but that's the beauty of statistics. Those interesting posts with few comments are balanced by lousy posts with a bunch of "this post sucks" comments. In the end, it all works out.
posted by jpoulos at 1:19 PM on June 4, 2001

We hope it works out, at least. : )

I'm of the opinion that it does... and BTW, perigree, great work!


posted by silusGROK at 2:08 PM on June 4, 2001

Considering all the hub-bub we keep getting about vets VS oldies

Hmm, which is better, vets or oldies. . . ? ;|
posted by rodii at 11:01 AM on June 5, 2001

and what about old vets?
posted by lagado at 5:25 AM on June 6, 2001