The purpose of this section of our paper is to draw attention to the fact that the terms "POA," "POD," "POS" and "ROW" are used in different and often ambiguous ways. Extreme care must be exercised when using these expressions. We want you to realize that when you use a term one way, listeners may be interpreting it another.

POA - The Probability of Area

We have observed this expression used in at least two different ways:

1.	As the Mattson Consensus.
2.	As a measure of the likely whereabouts of the subject each day of the search.

To elaborate:

1.

POA is used in the Mattson Consensus, which is an estimated guess of where the subject is in some segment of the search area. In this way, each segment is assigned an initial POA. It is actually a number between 0 and 1, but is frequently expressed as a percentage. The assignment of initial POAs is usually done only once during the entire search, and then at the beginning. There has to be an extraordinary reason to assign initial POAs more than once in any one search. The higher this number for a segment, the more likely the subject is in that segment, according to the group that determined these numbers. At the beginning of the search, the current POA and the initial POA are the same.

2.

POA is used to indicate the likely whereabouts of the subject. This is sometimes called the shifted POA, but a more accurate description would be updated POA. These are the numbers that are updated mathematically from the current POAs each time a segment is searched and an actual POD for that segment is determined. The higher the POA for a segment, the more likely the subject is in that segment, based on the initial scenario. After the operational period, the updated POAs become the current POAs.

POD - The Probability of Detection

This expression seems to be used in at least five different contexts:

1.	Actual single resource search efficiency.
2.	Actual multiple resource search efficiency.
3.	Cumulative POD for a search segment.
4.	Predicted single resource search efficiency for planning purposes.
5.	Predicted multiple resource search efficiency for planning purposes.

To elaborate:

1.	POD is used to mean a measure of the efficiency of a single resource after the resource has searched a particular segment. It is obtained subjectively, by interviewing the search team. Questions similar to the following are asked: "If the subject or clues were in your search segment, what were your chances of finding them?", or "How well did you cover your segment?", or "If 10 clues of various sizes were in your segment, how many would you have located?" In this latter case, if the search team responds with "3," this is converted to a POD of .30 or 30%. The figures from the search team might be adjusted (usually downward) by the management team later. The decision to do this is usually based on past experience and on experiments that have been performed to measure the effectiveness of resources, such as with grid teams[1] and helicopters[2][3]. This is the actual POD for the resource in that segment. It is the number that is applied to the current POA to obtain an updated POA for that search segment.
2.	POD is used to indicate a measure of how well multiple resources have searched a particular segment during a single operational period. It is computed from each resource's POD by using the following formula: 1 - [(1 - POD 1)(1 - POD 2) ... (1 - POD r)] where r is the number of resources that have searched the particular segment during the same operational period. POD 1 is the actual POD for Resource 1, POD 2 is the actual POD for Resource 2, and so on. Note that all these are numbers between 0 and 1 - they are not percentages. This POD is sometimes called the cumulative POD, but it would be more accurate to call it the pool POD, for the resources used during that operational period in that particular segment. It is the number that can be applied to the current POAs to obtain updated POAs for the entire search area. (If a segment is searched by more than one resource, then there are two equivalent ways of updating POAs: either use the actual PODs one at a time, or use the pool POD once. Both answers should be the same.)
3.	POD is used to refer to how well a search segment has been covered by all resources since the search began. It is computed from the actual PODs per resource by the formula: 1 - [(1 - POD 1)(1 - POD 2) ... (1 - POD r)] where r is the number of resources that have searched that particular segment during the entire search, and POD 1 is the actual POD for Resource 1, POD 2 is the actual POD for Resource 2, and POD r is the actual POD for Resource r. Note that all these are numbers between 0 and 1 - they are not percentages. This POD is also sometimes called the cumulative POD for the segment, which is an accurate description. This POD is sometimes used to curtail a search or expand the search area if it is high enough in all segments.
4.	POD is used to mean the measure of the efficiency of a single resource before the resource has searched a particular segment. It is obtained either objectively or subjectively, depending on whether experiments have been conducted using that particular resource in similar terrain. It is a tentative POD. This is the figure that might be used by the management team in planning future operational periods, such as in CASIE's "Hypothetical Search," "Cumulative POD" and "Resource Allocation Advice" modules.
5.	POD is used to indicate a measure of how well multiple resources might search a segment during a single operational period. It is computed from the tentative PODs per resource by the formula: 1 - [(1 - POD 1)(1 - POD 2) ... (1 - POD r)] where r is the number of resources that have searched the particular segment during the same operational period, and POD 1 is the tentative POD for Resource 1, POD 2 is the tentative POD for Resource 2, and POD r is the tentative POD for Resource r. Note that all these are numbers between 0 and 1 - they are not percentages. This POD is sometimes called the cumulative POD, but it would be more appropriate to call it the tentative pool POD, for the resources for that operational period in that segment.

POS - The Probability of Success

This expression is used in at least three different ways. It is defined to be the product of POA and POD, but the problem is which POA and which POD?

1.	POS is a number obtained by multiplying two other numbers together. All computer programs and presumably people taught to do this by hand in search classes compute the POS for a segment by multiplying the updated POA by the cumulative POD for each segment. While there is nothing wrong with multiplying any two numbers together and calling them anything you like, using the word "success" in this regard has very misleading overtones. After all, the updated POA has already been obtained from the initial POA by using the cumulative POD, so this POS uses the cumulative POD twice. POS, used this way, does not contribute to the search. What is the point of computing the "probability that the last conducted search was a success"? If the subject was found, then it was a success, and, if the subject was not found, it was not a success no matter what the "after search" POS turns out to be! This POS is of no use after the search.
2.	POS is used as a planning tool. Here the segment POS is the product of the predicted POD and the current POA for that segment and therefore is the probability that, in a future search, a subject will be found in a given segment by a given resource. The POS for a segment really only has value in prediction.
3.	The overall probability of success, denoted earlier in this article by OPOS, is the probability of success (or probability the subject will be found) given a certain allocation of resources searching the current area segments with their current POAs. As mentioned before, the theoretical goal is to maximize this overall probability of success, and it can be shown that this is exactly the same as maximizing the POA of the ROW.

ROW - The Rest of the World

This expression is used in at least two different contexts, neither of which are likely to cause confusion. They are included here for completeness.

1.	ROW is used to describe the area outside the existing search area.
2.	ROW is used to refer to the current POA for the area outside the existing search area. Example: "The ROW was 14% at the end of yesterday's operational period."

References

[1] Wartes, J. An Experimental Analysis of Grid Sweep Searching, 1974.
[2] Bownds, J., Lovelock, D., McHugh, C., and Wright, L. Desert Searches: Effectiveness of Helicopters, 1981.
[3] Bownds, J., Harlan, A., Lovelock, D., and McHugh, C. Mountain Searches: Effectiveness of Helicopters, 1991.

Section I Section II Section IV