During a typical ``day'' (10:00 PM to 4:30 AM), the Lansing State Journal
prints 76,000 newspapers.
The job of delivery begins with the presses since the product must
be available to customers early in the morning in the far corners of
the state.
This means that once salable papers appear at the end of the pressthey
are immediately loaded onto waiting trucks.
The process of printing, inserting (advertisements, etc.), bundling,
and loading continues throughout the night so that when the presses
finally stop the last truck is loading and just as quickly leaving.

During a typical ``day'', the Lansing State Journal sells 74,000
newspapers.

It would be reasonable to ask why 2,000 extra papers are printed.
The answer is more complex than one might imagine and partially appears
in a study of how a modern printing press operates.
The process of printing a newspaper is a continuous one, from largerolls
of ``white paper'' at one end to finished and folded papers at theother.
Starting the press typically produces 50 spoiled papers before the
first salable copy appears.
Stopping the press is even more costly, 75 to as many as 300 spoiled
papers result.

Depending upon the edition, day of the week, and other factors a press
may start and stop several times per ``day''.
These factors, as well as expected demand, are built into the
pressperson's daily ``manifest'', i.e. the total number and kind ofpapers
that are to be printed.
However, this is only the beginning of the story, since unpredictable
counting and loading errors, handling damage, etc. all add to the number
of papers that are needed.
Typically the latter problems result in several hundred --- up to 500
--- missing papers per day.

The project has two aspects.
The first is to determine if it is possible to reduce number of missing
papers --- and then describe how.
The second is to perform a statistical cost-benefit analysis of
printing the newspaper that seeks to minimize the manifest count (the
total number of papers to print).

The ideal completed project deliverable for the former is a set of
recommendations for new equipment and/or accounting techniques along
with evidence of their effectiveness. The latter deliverable couldbe a
parameterized production model that minimizes cost --- again,
along with the evidence of effectiveness.


This summary prepared with the assistance of M. A. Alford,K. Swanson, and
C. Syjud of the Lansing State Journal.