Date: Sat, 2 Apr 2005 03:52:27 -0500 (EST)
From: "Keith F. Lynch" <kfl at KeithLynch.net>
To: WSFA members <WSFAlist at WSFA.org>
Subject: [WSFA] Predicting WSFA attendance
Reply-To: WSFA members <WSFAlist at WSFA.org>

As some of you noticed, I had the minutes of tonight's meeting in the
WSFA Journal that was distributed before the meeting (and that was
available online for five days before that).  The minutes were a
deadpan mixture of the best prediction I could make and the most
off-the-wall things I could come up with, sometimes in the same
paragraph.

It included an attendance list -- my best guess at who would be there.
Or rather the best guess of the program I wrote which prints the
sign-in sheets.  The sign-in sheets have rooms for lot of names, so
it's fairly rare for anyone who shows up except a first time visitor
to not find their name already on it.  But for the meeting minutes, I
made the cutoff 50%.

So how well did I do?

I predicted that there would be 35 people at the meeting, plus two who
would show up afterwards: Ted White and Bill Squire.

In fact, there were 35 people at the meeting, plus one who showed up
afterwards: Ted White.

I got 29 of the 35 (83%) correct.  Six people were present whom I
didn't expect: Adele Tyhurst, Dan Joy, Drew Bittner, James Uba, Lance
Oszko, and Tamara Griesel.  (All were on the sign-in sheet, however.
Nobody had to write their name in.)  Six people were absent whom I had
expected: Adrienne Ertman, Cat Meier, Houston Westfall, Judy Kindell,
Judy Newton, and Michael Nelson.

I also predicted that someone would check off Jim Kling and Ivy Yap,
even though neither would be present.  And I was right.  (No, I don't
know who does it, or why.)

I will continue to attempt to improve my predictive algorithm.  I'm
always open to suggestions for ways I can do this.

For two of the unexpected six, it was their second meeting.  Most
people (53%) who show up for one meeting never return.  As for the
expected six who didn't show, perhaps the rainy weather was partly
to blame.