Copyright (c) 2013 John L. Jerz

Decisions with Multiple Objectives by Keeney and Raiffa

Home
A Proposed Heuristic for a Computer Chess Program (John L. Jerz)
Problem Solving and the Gathering of Diagnostic Information (John L. Jerz)
A Concept of Strategy (John L. Jerz)
Books/Articles I am Reading
Quotes from References of Interest
Satire/ Play
Viva La Vida
Quotes on Thinking
Quotes on Planning
Quotes on Strategy
Quotes Concerning Problem Solving
Computer Chess
Chess Analysis
Early Computers/ New Computers
Problem Solving/ Creativity
Game Theory
Favorite Links
About Me
Additional Notes
The Case for Using Probabilistic Knowledge in a Computer Chess Program (John L. Jerz)
Resilience in Man and Machine

Decisions with Multiple Objectives by Keeney and Raiffa

Briefly, what is the difference between a goal and an objective?
 
p.34"An objective generally indicates the 'direction' in which we should strive to do better... A goal is different from an objective in that it is either achieved or not.
   Goals are useful for clearly identifying a level of achievement to strive toward."
 
When we select attributes which tell us to what degree we are accomplishing our objectives, we should select cues which are measurable and which address the full scope of the issues - real and strategic - which stand between us and our objectives.
 
p.38"To be useful to the decision maker, an attribute should be both comprehensive and measurable."
 
Keeney and Raiffa have this to say about the set of attributes which we select in order to make our decision: they should be complete, operational, decomposable, nonredundant and minimal. The magazine Consumer Reports is a great way to see this in action.
 
p.50"It is important in any decision problem that the set of attributes be complete, so that it covers all the important aspects of the problem; operational, so that it can be meaningfully used in the analysis; decomposable, so that aspects of the evaluation process can be simplified by breaking it down into parts; nonredundant, so that double counting of impacts can be avoided; and minimal, so that the problem dimension is kept as small as possible."
 
Could it be that the traditional methods that chess programs have been using in their evaluation function, such as mobility and nearness to the center of the board, are just proxy attributes - that is, an indirect  measurement of an associated objective? If so, then what are the real attributes that help us reach our objectives?
 
p.55"A proxy attribute is one that reflects the degree to which an associated objective is met but does not directly measure the objective. Thus, proxy attributes indirectly measure the achievement on a stated objective. It could be argued that essentially all attributes are proxy attributes because nothing can be absolutely measured. There are just varying degrees to which an objective is directly measured."
 
We need to spend some time just thinking about our objectives and the decision making attributes which help us make progress towards these objectives.
 
p.64"To attempt any formal analysis of a complex decision problem requires an articulation of the decision maker's objectives and an identification of attributes useful for indicating the extent that these objectives are achieved."

Enter supporting content here