The Agrinautica

   Its official. After growing rather suspicious over the last few months, I was becoming very concerned that the rumours were true -- that people don't read livejournal on the weekend.

   So this morning I participated in my favourite activity: getting up at 8am to smash statistics together from random livejournals while procrastinating doing something completely different (writing that long overdue OPEC paper for MUN)!!
   If you're curious about the mechanics behind my brilliantly conceived research project, you can read more about it here.

   The bottem line is that out of 25 randomly generated livejournals, 18 were too crappy to have enough entries / comments in the last two months to be of any value. But of those remaining seven, the HIGHEST proportion of average comments received on Saturday to those received during an average weekday was 73%. The median was 52%.
   So basically, your entries on the weekend will get half as many comments on average as those you make during the week (from which we extrapolate that half as many people will ever even see your entry on the weekend, since we have no reason to believe there's a disconnect between reading and commenting here).



   HOWEVER, then I ran myself. And _I_ have 30% more comments on average per entry on the WEEKEND than during weekdays!! ...but I think that maye be because my entry about Mackeson Triple Stout got the most comments of any entries I made in the last two months and fell on a Saturday.


Further Reading
More Detailed Write-Up of the Experiment
The Data
Other Similar Studies

   This entry is post-dated from Friday, in preparation for my expected death on 06-18-06.

   For 30 in 30 I in 2004 I conducting a groundbreaking research study in the field of blogology, compiling a range of statistics on a sample group of bloggers based on data available on the userinfo pages. A year later I did a follow up during 30 in 30 II. And now, now its time for this year's annual update on what the statistics say about our intrepid sample experimental group.

   Among the additions to the study group this time was hipstomp, who is one of (if not THE) most popular bloggers on livejournal (according to some meme that went around a bit ago).
   As I've discussed before, many people see the ratio of friends / friends-of as the primary gage of blogging superiority, but my research has shown that this is a far better indicator of who is an ogre that frequently friends & defriends people than anything else. Of the people with the highest such ratios, hipstomp does lead with a friends-of list of 832% the size of his friends list, but one legitimately couldn't expect him to keep up with his 1315 friends-of. He is followed however, by noted blogletists otimus & incomple, with f-o lists of 738% & 253%, respectively. Differences of less than 30% though are probably more likely to reflect something real, though many very qualified bloggists are around the 1:1 ratio or less due to adding people back.

   A particularly interesting series of numbers are changes of friends-of numbers from last year. ASUCD Vice President gennadotorg has had the greatest increase of 305%, which may be because her lj was only 3 months old when last year's numbers were compiled, and it still had most of its growth in front of it. Comparatively, emd who was new last year, and had a 423% increase from the last year, has levelled out to the middle of the pack with 107%. On the other side of the spectrum, formerly renowned livejournal blogging from Iraq, giantlaser has had a 5% decrease in friends-of since resuming a more boring life in England, incomple is down 8%, otimus has lost 20% of his friends-of since last year (2.7 defriendings per entry)(and see last year's report on this for how he got those friends), and roter_terror has lost 72% of his friends-of since this time last year (such sauce!).


   See for yourself how you stack up against my study group using the group averages: (out of 30 livejournals) - friends: 123; friends-of (FO): 147; FO/friends: 1.08:1; entries: 536; comments / entry: 12.2; comments received / cmmnts made: 1.11:1; lj age: 40mo; FO/mo 4.3; FO/entry: .55; entry/mo: 13.4; comment received / mo: 118.3
   Or leave me a comment and, if I turn up alive, I'll give you more analysis of how you compare to everyone else.


Picture of the Day

Once again what can happen when you drunkenly pass out at a party.

Previously on Emosnail
   Three Years Ago Today: Done With Last Paper of the Year - Finally. And an awesome picture of an anteater.
   Two Years Ago Today: 30 in 30 I - 6, 7 & 8 - 11 Hours on Amtrack, Controversial Political Statements about farm subsidies, & Interview: The Sexual Confessions of shid.
   Year Ago Today: 30 in 30 II - The Journey Home - An epic odyssey across California.. or something.

   A year ago in the study published in my 18th 30 in 30 entry I found that otimus had no less than 750 people on his friends list, with 143 friends-of, and 49 livejournal entries. This gave him the whopping ratio of 5.2 to 1 friends/friend-of. A year later I ran the numbers again, and this time otimus has 23 friends, 193 friends-of, and 37 entris. Whereas before he had had about five times the ratio of anyone else, now has by far the smallest ratio at 0.119 to 1.
   Also the deleted entries have caused his comments / entry to rocket up to about 88 / entry. Compare this to the next highest, apoplecticfittz with 48.6 and the average of 11. But did this craziness succeed in what I'd imagine its intended purpose was, to give him a throbbing friends-of list? Last year he had an average of 5.1 people friending him per month, now its 4.8.

   A year ago emd was new to the blogosphere and had a friends/friends-of ratio of 1.66, with comments per entry at 3.69. Since then she has quadrupled her friends-of list, and developed a bulging friends/friends-of ratio of 6.89, and not through any kind of trickery either -- her friends list itself has increased 176%, the highest increase of those studied, except for thetalesend who went from 6 to 13 friends and fucked up the numbers. Now emd gets 8.34 comments per entry

   The study was expanded from 20 to 25 livejournals studied, with three of the originals either inactive or no longer displaying their friends-of information (tingsquared!). Its important to note of course that these livejournals were not randomly selected, so comparative numbers might not be representative of the entire blogosphere. I bet someone with some computering skills could make a program or something that would run all these numbers rather than me being required to enter things by hand into excel. Then I could go on a statistical rampage.


   I already rather explained my thoughts on the statistics last year, but I'll go over them again in brief: my conclusions are that friends/friends-of ratio is a bogus indicator of blogging quality, because it encourages bloglitism (pronounced "blog-leet-ism") in keeping one's friends list low purely for the numbers. Many people who fancy themselves to be be kings or queens of the blogosphere and feel inadequate about the size of their friends-of will actually hide it from the world. Its perfectly healthy to have a ratio as high as the 1.2:1s, so don't be afraid, show off your blogsecurity and blogconfidence instead and whup out your friends-of list for all to see. If one is below .6(:1) though one is either awesome or bloglitest. (average of group studied: 0.91)
   Comments per entry is in my opinion the best indicator of someone who is an outstanding member of the blogosphere. High comment levels either indicate discussion inducing entries, or a livejournalist with great convsersational skills.... either that or they've been running around flaming people. 36% of those studied had less than four comments per entry, 56% had less than 6. Then there's a group with between 6 and 12 (24% of total), after which numbers hoot up with the top 20% averaging 45 comments per entry. (Average of group studied: 11.4, but the bottem 80% only averages 4.98)
   The other thing I think is a noteworthy indicator of megablogging skill is friends-of / month, which is kind of like looking at the friends-of list and then weighting it for time the livejournal has been existant. Top two in this category are giantlaser who was blogging from Iraq, and theuglyvolvo, with 16 and 13 respectively. In third we have emd with 8.7. I'd imagine this number starts out high with new livejournals and then decelerates, so far there have been only three livejournals studied at less than a year of age, two of which got exactly 7.00. (Group average: 4.42)


   In the past year my own comments have increased from an avg of 5.43 to 9.66 comments per entry over the last year. Friends list increased 136% while friends-of list increased 135%.


   And now I've probably bored you to tears. I'm sorry, I just wanted to see how things had changed since last year. Someone with skill, write me a program to do this. Otherwise, if you're interested in seeing how you compare to everyone else, ask me, and if you've already been included or ask really nicely, I'll tell you.


Previously on Emosnail
   Two Years Ago Today: Naught but a Meme - those things are only fun if you edit the results
   Year Ago Today: 4 of 30 - Discussion Topics - What a good idea. I should do that again. I posted a bunch of questions for the readers to answer, like which member of the US Executive Office would be most likely to grow an awesome mohawk, or which band would Saddam Hussain most likely to be a part of?

   In order to finish on schedule with the 30 in 30 thing I will need to make a post every 9.5 hours. I don't know how I got so far behind.

   Anyway, continuing my scientific investigation of the blogosphere, I've done a study of 20 livejournals on my friends list¹, based on numbers of entries, comments and other quantifiable information. The results are as follows:


( The Study )

   Also determined that this journal is currently averaging 26.2 entries per month.
   Next I think I'll look at a number of randomly selected livejournals. These ones were selected to take livejournals I hypothesized were good examples of a megablog or underblog, and as such are far from a random sampling of livejournals.
   Also I didn't use anyone who didn't have "friends-of" displayed in their userinfo. This prevented me from using several livejournals I was interested in looking at.


Related
   Interesting Livejournal of the Day: giantlaser in Iraq


¹actually 10% of them aren't on my friends list
²Yes I am aware that proper ratios should not have numbers less than one. For simplicity's sake however I decided not to convert numbers like 0.3 and 0.6 to 3:1 and 5:3.