Surrogation is a practice where we substitute one or more characteristic in place of one or more others.
The surrogation might consist of one or more entire topics, but is more commonly limited to parameterization or values thereof, where the conditions are changed in coordination with the value produced from a Measure1.
Substitutions are usually made for one or more of the following reasons:
(1) To avoid verification cost
(2) To reduce the order of the developmental problem
(3) To abstract peripheral issues away from the core development problem
(4) To benchmark related development programs over an extended period of time
Example: Probability of Hit (PH) against a specific Opposing Force (OPFOR) vehicle at some specific range2, is replaced by PH against a standardized target board at a standardized range3. The intrinsic ability of the projectile weapon is the same no matter which of the two Measures is chosen. The standardized pairing of range and target board permits ready comparison of many generations of weapons performing the same operational mission, and relieves developmental gun programs of the need to explicitly verify capability against individual OPFOR vehicles when using a legacy munition4.
Contrast the somewhat convoluted contrivance of the example’s PH-vs-Target-Board with the concept of a Limit Condition, which is of direct technical relevance5.
Footnotes- That is, the value under Condition Set A corresponds to a different value under Condition Set B.[↩]
- Which will have a unique value when integrated over all target aspects[↩]
- Which will have a different unique value, over just a single target board aspect[↩]
- For which the probability of kill given hit, against any particular OPFOR vehicle, is already known.[↩]
- Being merely a statistical data reduction as opposed to an inference.[↩]