friedl@uunet.uu.net> (08/04/90)
This interesting tidbit from _Privileged Information_, 1 Sept 1990
Common sense says that when someone is waiting to use a
pay phone, the person using it will hurry up. Reality:
While the average pay phone call without a waiting line
lasted only a minute and a half, when someone waited
behind the person using the pay phone... the caller's
conversation lasted *four* minutes. Why? People using
pay phones become territorial when someone else wants
to move in... study by Dr. Barry Ruback of Georgia State
University.
Stephen J. Friedl, KA8CMY / Software Consultant / Tustin, CA / 3B2-kind-of-guy
+1 714 544 6561 / friedl@mtndew.Tustin.CA.US / {uunet,attmail}!mtndew!friedlclj@ksr.com (Chris Jones) (08/06/90)
Alternative explanation: the reason there are lines behind the people
using the payphone for longer calls is that the people in front are
using the phones longer. I have a gift, on the order of being able to
turn gold into lead, of being able to study a set of lines in, e.g., a
grocery store, and, with ridiculously high probability, pick out the
one wihich will cause me to wait the longest.
Chris Jones clj@ksr.com {world,uunet,harvard}!ksr!cljcosell@bbn.com (Bernie Cosell) (08/06/90)
mtndew!friedl@uunet.uu.net (Stephen J. Friedl) writes: }This interesting tidbit from _Privileged Information_, 1 Sept 1990 }Common sense says that when someone is waiting to use a }pay phone, the person using it will hurry up. Reality: }While the average pay phone call without a waiting line }lasted only a minute and a half, when someone waited }behind the person using the pay phone... the caller's }conversation lasted *four* minutes. Why? ... I cannot assert that his data is really 'cooked', but there is a statistical oversight in his reasoning that can best be explained by looking at lines-of-cars trapped in no passing zones. If you look at the longest lines, you find the slowest drivers at the head of them -- should you then conclude that having drivers piling up behind one tends to make drivers slow down? or is the more reasonable observation that the slower you drive the more *opportunity* you have to get folks to pile up behind yuou. Similarly, if one looks at some of the underlying queueing theory one will see that if the server availability [i.e., the number of phones] pretty closes matches the client demand [i.e., the number of people that want to make calls], VERY small changes in the duration of a call will make a BIG difference in the length of the resulting queue. /Bernie\