public performace results

From: Bill Barowy (wbarowy@yahoo.com)
Date: Wed Aug 08 2001 - 07:54:22 PDT


The flurry of emails this morning was warming, a nice touch to the heat of
events here in the northeast corner of massachusetts. Here is my own brand of
mysticism, a graph of the data that Eva offered. In blue is the real data, and
in red is a straight line fit, which was calculated ignoring the contributions
of the top 6 authors.

http://www.lesley.edu/faculty/wbarowy/sr/xmca.8.8.gif

Why not fit all the data? Without insight into the data, one might insist on
doing so with mathematical dogma, a process not to be violated. I disagree.
For one, ignoring the top six contributions "looks best" for aligning to the
rest ( and one can interpret this mathematically with the terms for best fit --
i'm willing to do so if you find yourself in gear-head mode). Besides, rather
mystically, ignoring the numbers for Mike Cole, Paul Dillon, Diane Hodges, Bill
Barowy, Eric Ramberg, Judy Diamondstone leaves an "anomoly" for zipf's law,
which is what the straight line in red describes. Clearly, these top six
players are not posting enough, and it is great that Diane is working on making
up the difference. But can she really? Can Mike and Paul possibly post what
they must to conform to the pattern of zipf's law that the remainder of people
posting to xmca form? Here is what the red line indicates is necessary to do
so:

person #messages
_______________________
Mike Cole 3629
Paul Dillon 1039
Diane Hodges 500
Bill Barowy 297
Eric Ramberg 199
Judy Diamondstone 143

Quite interestingly, I think the anomaly has, in part, to do with the memory
discussion going on, and (relatedly) in part to do with the ecological
constraints on activity. After all, Mike would have to be posting 10 messages
a day to keep up with the pattern, and Diane about 2 per day. You might argue
that the ecological limits on time are sufficient to describe what is
constraining these top six. But "memory" may be important for two reasons, and
I'm interested in finding out to what extent it is involved. First, Eva's
diagrams of intermessage linking are a beautiful visual expression of "memory"
-- mapping how each message indexes those before it. Second, Zipf's law is
based upon the absence of memory -- that each contribution to the rank of a
category is completely unrelated to the contributions before it. This does not
seem reasonable for xmca, given Eva's work.

gotta go.
bb

=====
"One of life's quiet excitements is to stand somewhat apart from yourself and watch yourself softly become the author of something beautiful."
[Norman Maclean in "A river runs through it."]

__________________________________________________
Do You Yahoo!?
Make international calls for as low as $.04/minute with Yahoo! Messenger
http://phonecard.yahoo.com/



This archive was generated by hypermail 2b29 : Mon Oct 01 2001 - 01:02:02 PDT