Performance tracking: A historical background to promote learning

Abstract

This geologic note provides a lookback on lookbacks, hopefully to spur further interest by companies in generating learning from exploration experience, rather than being victimized by repeated mistakes to which exploration assessment may unknowingly be exposed. The lookback contribution to portfolio management is properly called performance tracking. Likely inspired by the after-action review process installed by the US Army approximately 50 yr ago, oil industry published results of tracking predictive performance began via a simple cross plot (predrill prospect resource size versus postdrill accumulation resource size) by US Geological Survey (USGS) workers documenting discovery sizes on the United States continental shelf. Whereas predrill estimates made by the USGS were relatively accurate, analogous publications by the industry have usually documented recurring optimistic bias. Other contributions have demonstrated the utility of determining (1) the efficiency of prospect ranking by various predrill estimation parameters (via an estimation efficiency plot), (2) when a drilling program falls outside of statistical control (via a sequential aggregation plot), and (3) where along a series of predrill forecast distributions persistent estimation bias has occurred (via a percentile histogram). These basics can serve as a starting point for any company to build their own baseline study to learn what aspects of their many estimates need improvement.