[Halld-jlab] Fwd: keeping track of shifts taken

[Mark Ito]

policy from last Spring run...


-------- Forwarded Message --------
Subject: 	keeping track of shifts taken
Date: 	Mon, 23 Feb 2015 16:46:19 -0500
From: 	Mark Ito <marki at jlab.org>
To: 	Alexander Somov <somov at jlab.org>, Alexandre Deur 
<deurpam at jlab.org>, Beni Zihlmann <zihlmann at jlab.org>, Dave Mack 
<mack at jlab.org>, David Lawrence <davidl at jlab.org>, Elton Smith 
<elton at jlab.org>, Eugene Chudakov <gen at jlab.org>, Hovanes Egiyan 
<hovanes at jlab.org>, Ilya larin <ilarin at jla.org>, Justin Stevens 
<jrsteven at MIT.EDU>, Lubomir Pentchev <pentchev at jlab.org>, Mark Dalton 
<dalton at jlab.org>, Mark Ito <marki at jlab.org>, Paul Mattione 
<pmatt at jlab.org>, Sergey Furletov <furletov at jlab.org>, Simon Taylor 
<staylor at jlab.org>



JLab Folks,

This is a proposal for keeping track of shifts and making sure
shift-taking loads are fairly distributed. There are three parts, (a) a
point system, (b) how we use the point system to assign shifts, and (c)
some necessary exceptions. Note that this is only for JLab people.

a) point system

We keep a record of shifts taken (shifts that are cancelled are
ignored). Each shift taken gives you one point.

Shifts that overlap the owl shift period (midnight to 8 am) are awarded
additional points as follows:

half overlap with owl period: 0.5 additional points
full overlap with owl period: 1.0 additional points

which gives the following point schedule:

owl, 00:00-0:800: 2 points
duck hunter, 04:00-12:00: 1.5 points
day, 08:00-16:00: 1 point
country club, 12:00-20:00: 1 point
evening, 16:00-24:00: 1 point
night club, 20:00-04:00: 1.5 points

If you are on stand-by for a shift that subsequently gets cancelled you
get 0.2 points independent of the time of the shift.

b) using the point system

We assign shifts randomly during a shift scheduling period with an
as-even-as-possible number of shift blocks assigned for each person.

If the number of shift blocks is not an exact multiple of the number of
shift takers (the usual case), then there will be some with one more
shift block than others. We apply the constraint that the
one-more-shift-block people come from those with the lowest number of
points.

c) exceptions:

1. If the run coordinator is a JLab person, he or she gets no assigned
shifts, but gets points equal to the group average for the period.

2. New group members are started with enough points to be tied for last
place in points

Finally, I propose we forget about shifts taken during Fall 2014 and
start keeping track for Spring 2015. Also note that if we neglect last
year's shifts, the shift assignments I've already made for Spring 2015
are not inconsistent with this system since everyone would start with
zero points.

   -- Mark



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-jlab/attachments/20151015/62afd2c5/attachment.html>