Apparently, I’m not the only person with absolutely nothing to do at 5:00 AM (EDST). Here be the stats, ranging from userid 10037 to 10049:
Terry Betts - 2
BlueCabbage - 6 (Complete Twit)
ok_guy - 0
chuteman - 0
Some Genius - 0 *
ctolbert6 - 2
Cerowyn - 255
Some Guy - 248
bluehobo - 0
Mr.Stefano - 8
drewcosten - 8
Qasper - 7
Melissa Jane - 0
In summary: 13 registrations, only two of which (Cerowyn and I) have any sort of active post count - though we both have <1 post per day; 1 person who is not merely banned, but a Complete Twit; and one surprising asterisk - an earlier failed registration of mine (the confirmation email never showed up, so I couldn’t activate that account - consequently, nobody’s more surprised than I to see it listed as active…)
Yanno, I’ll bet that Cerowyn’s never heard of me, either
Well I’ve entered all the data into Excel and the preliminary results are suggesting that the half-life is very well correlated exponentially to the “age” of the doper. The results trend beautifully. Man, you guys should see the graph (sniff).
This in itself means that the concept of “half-life” may be flawed when dealing with dopers, unfortunately.
One thing may be skewing the results however: back in 1999, the joiners must have ben pretty serious Dopers. The board was tiny compared to now, so the casual registers with 0 or 1 posts were far rarer. A higher percentage of serious dopers possibly means longer life. Certainly the joiners from 1999 have a “half-life” of about 15 months compared to 4-5 months for joiners a year or so back. If we remove this data, there is a better case for a static half-life. Better but not great - I still think that it trends even so.
Well I haven’t a CLUE whether anybody reading this is still interested, but I’ve done the first bit of the analysis:
Assuming n posters join at time t[sub]0[/sub], the number y expected to still be posting at time t years later is (drum roll please…)
n. 0.009134[sup]2.3[sup]-t[/sup].t[/sup]
Phew!
What that means is that we can say that t years after a poster starting, the probability that said poster is still active is: 0.009134[sup]t. 2.3[sup]-t[/sup][/sup]
This I declare the new kabbes’ hypothesis.
For stat-heads out there - the above shows a 99.8% [sym]C[/sym][sup]2[/sup] probability level, meaning it fits very well indeed!
I know there are some things that go beyond insults, and I shall understand if you never wish to speak to me again. I humbly beg forgiveness and sacrifice virgins to Great Alumni in the Sky.
Date registered: 4-21-00
user id 5872
24 total registered on that day
acitve:
# last post
Crown Prince of Irony 190 10-18-01
kiffa 687 10-29-01
dead0man 75 11-05-01
carnivorousplant 2031 11-05-01
This may be a bit of a hijack (I’m a newbie… please go easy on me), but I have a couple of closely related questions.
I wonder what percentage of registered users have fewer than, say, 10 posts? (from reading through all these “birthday” lists it seems a very high percentage of users have very few posts)
I wonder what percentage of registered users are currently active? (e.g.-- have posted at least 10 times in the last 90 days)
Anybody have any idea if there is a way to find out the answer to these questions? (without a lot of manual labor at least).
