The problem of the exploration geologist is to predict accurately the state of nature at a concealed point within the earth. Management is committed to spend exploration money and has little choice but to accept the geologists' recommendations. They are the only people who have actually studied the data. Therefore the professional responsibility lies with the organization and analysis of information. Prediction requires an extension of knowledge from the known into the unknown areas. That is the trick. Two methods of increasing predictability are presented. One is concerned primarily with the sandstone reservoirs. Here the gross shapes of rock units are considered to be functions of the vectorial nature of the energy gradients present in the depositing currents. These vec orial directions can be deduced from the unda, clino, or fondo environmental position in which the rocks were deposited. The second method is of less general nature. It is concerned with predicting the loci of carbonate reef knolls from the available information on the submarine topography of deposition.

It is concluded that as the structural hunt in the unda beds with their ease of lateral predictability is brought to a close by the exhaustion of anticlinal possibilities the geologists will be forced to study the environment and topography of the clinoform. In this zone of unequal depositional rates, abrupt facies changes and few regional markers, the exploration tool will be one of logic. When economically feasible, drilling the first of a series of wildcat wells as a means of improving the distribution and quality of data will be a valid approach to later discovery.