"It's obvious that someone could read the encyclopedia and eventually get everything right about a target. It is also obvious that some information about a target, which correct, might have very little or no value (e.g. the grass is green). Therefore an appropriate algorithm needs to be created that gives a value to descriptive remote viewing that's better than a subjective value judgment. In the case of controlled targets this is easily done. When a target is placed within a target pool, there should be a prior and independent evaluation of the target to determine what could be said about the target or what one would naturally expect to be said about the target based on the method of targeting (e.g., target is a photograph - looking at the photograph what appears to be significant to the photograph, and what is important or significant to the description of the photograph.) These descriptions are written down and become part of the targeting package.

Once the remote viewing is done, the results or statements taken from the remote viewer are compared to what is already a part of the target package (the statements of expectation and import already written down and stored with the photograph). A determination is made of what percentage the remote viewer said that was expected as important and relevant to the targeted photograph, and what did the viewer say that was not relevant or pertinent to the targeted photo. One needs to also look at what was said that was wrong about the target or not material to the target as well. From this algorithm, one can derive a resultant value (we call it a Figure of Merit.) For obvious reasons, a figure of Merit of 60% or better would be considered a very good result. A figure of Merit of 95% would be considered exceptional and very rare. Statistically under double blind studies within laboratories, Figure of merits above 80% occur probably less than 25% of the time. "