Project: Application of Linear Programming in Professional Sports
Evaluating Baseball Player Ratings
Option III
Models and Methods of Optimization, Fall 1999
Overview of Project
Each student will utilize the LP model described in
the article,"Using Linear Programming to Determine Post-Facto
Consistency in Performance Evaluations of Major League
Baseball Players" to estimate the relative rating
of a baseball player who does not appear on the top 10 list in the article.
The
student will then write a report discussing the estimation procedure.
The project will be graded primarily on
the final project report, due on Monday, November 29, 1999 at the beginning of
lecture. Details of the
requirements of the final project report and grading criteria are described
below.
The article,
"Using Linear Programming to Determine Post-Facto Consistency in Performance
Evaluations of Major League Baseball Players"
by
Chistopher Zappe, William Webster and Ira Horowitz,
appearing in Interfaces Vol. 23, No. 6,
November-December 1993 (pp. 107-113), is on reserve in Hunt
Library (behind the circulation counter on the 1st floor).
The journal also is located in the Bounded Journals section of the library
(on the 3rd floor).
The article abstract:
We use linear programming to judge the consistency of the performance
appraisals that Bill James has made of major league baseball players. In
particular, we use a linear programming model to determine whether
weights for the set of performance measures can be found such that the players'
ordinal ratings can be reconstructed. This approach can be used to judge
the consistency of performance appraisals for groups of employees.
Project Timeline
- Read the assigned article, "Using Linear Programming to Determine
Post-Facto Consistency in Performance
Evaluations of Major League Baseball Players"
- Player Choice: The student shall choose a baseball player that
does not appear on the list. The student is responsible for finding
the required data for estimating the players relative ranking..
- Drafting the LP: The student shall derive a linear program
to estimate his/her player's standing using the data from the article.
- Tuesday, November 9, 1999 and Thursday, November 11, 1999: Optional Consultation with Instructor: The student shall
set up a 10 minute meeting with the instructor to discuss his/her
solutions to the problem. Sign up for a timeslot on the sheet outside of
7213 Wean Hall. The student should have the LP prepared
and solutions found by the
time of the consultation.
- Friday, November 19, 1999: Optional Rough draft of Analysis Section of report is due.
Comments will be made on the rough draft and returned to the student
on Wednesday, November 24, 1999. No late rough drafts will
be accepted.
- Monday, November 29, 1999: Final Project Report is due.
Note: Aside from the Optional Consultation with Instructor, any discussion
of individual projects with the instructor shall be done during the
instructor's office hours.
Final Project Report Requirements
The final project report is due at the beginning of lecture on Monday,
November 29, 1999.
The typed reports should consist of the following sections:
- Problem Statement:
This section of the paper should describe the problem at hand. The
student should include solution strategies that were used in the past
and reasons why a new strategy might be warranted. The student's choice
of ball player and relavent data should be included here. One source
for possible data would be
totalbaseball.com under people and legendary players.
- Solution Strategy:
The student should outline the solution strategy employed and
give some justification why this strategy might be more desirable than
the other strategies listed in the Problem Statement. If variables are
used, explanations of what the variables represent should be given. (This
information should be provided once variables are introduced, either in
this section or the next).
- Solution:
This section of the report should include the student's solution to the
LP described in the Solution Strategy section. If LINDO was used, the
computer output should be included here.
The solutions should be organized and
annotated
so that a reader will be able to relate to the solutions steps
described in the solution strategy.
- Analysis: This section of the report will
act as a conclusion. The student should not only state the final solution
to the problem, but the student should discuss the result. He/She should
include the following:
- Advantages/Disadvantages to the solution strategy
- A list assumptions that were made in the formulation
and solution to the
problem and a discussion of
how reasonable the assumptions were, how one might
eliminate some of the assumptions, etc.
- Variations of the solution strategy.
- Possible extensions or variations of the problem.
- Importance of the problem and the importance of extensions or
variations of the
problem.
Project Grading Criteria
Each student is responsible for turning in a final project report which
shall be solely the work of him/herself. No collaboration with anyone
other than the instructor or one of the TAs is allowed.
Any violation of this shall be dealt with according to university regulations.
The project grade will be calculated from the final project
report, worth 100 points, described below. For each day the project is
late, 10 points will be deducted from the project
score.
- (15 pts) Problem Statement: These points will be awarded
for a concise statement of the problem, organized presentation of
data and discussion of the points detailed in the Problem Statement
Requirements above.
- (15 pts) Solution Strategy: These points will be awarded for
a clear description of the LP and the method of estimating other players
baseball rank.
- (10 pts) Solution: These points will be based on clear
organization and annotation of the solution, as well as a correct
solution.
- (50 pts) Analysis: These points will be awarded for
a well-organized and thoughtful discussion of the
project, including (but not restricted to) the items listed in the
analysis section requirements above.
- (10 pts) Presentation: These points will be deducted for
spelling and grammatical errors in the report.
Return to
Optimization Page.