REPORT ON PRE-WAR FORECASTING For Information and Comment Accuracy of Pre-Kuwait War Forecasts by T.N. Dupuy and HERO-TNDA 18 March, 1991 Introduction Before the outbreak of the Kuwait War T.N. Dupuy, President of HER0-TNDA, Publishers and Researchers, Inc., and his organization had three opportunities to forecast the likely results, losses, and duration of a war between Iraq and United Nations coalition forces (particularly those of the United States). The first of these opportunities was in a book by Dupuy, entitled: Attrition; Forecasting Battle Casualties and Equipment Losses in Modern War (HERO Books, 1 December, 1990). In that book Dupuy discussed in considerable detail the historical attrition data that was used in the development the Tactical Numerical Deterministic Model (TNDM), a proprietary computerized simulation of combat produced by HERO-TNDA. He then described a methodology for forecasting casualties based upon the TNDM. Included in that book was a simple scenario of a possible conflict between US and allied forces against Iraqi forces in Kuwait. Quoted below is the last paragraph of that book. "...The reader should know that some analysts who have seen the forecast casualties for the ten-day Kuwait operation scenario, shown above, have expressed concern that these rates seem to be low. There is a reason for this. These analysts have been exposed to prospective attrition rates calculated on some of the major models now in use by the Department of Defense. Until or unless the projected war actually occurs, there can be no certainty that the attrition forecasting method proposed in this chapter is better than those contained in the Defense Department models which have been used over many years to forecast casualty rates in the event of a war in central Europe. The author of this book has long asserted that the rates forecast by those models are -- on the basis of projections of historical trends -- far too high. Sadly, it is possible that we may soon learn who is right!" Dupuy's second pre-war forecast was made on December 13, 1990, in testimony before the Armed Services Committee of the U.S. House of Representatives. In that presentation he gave forecasts of the likely casualty costs of a variety of potential scenarios; all of the results were consistent with the earlier forecast in his book on Attrition. The third pre-war forecasting opportunity was in another book, entitled If War Comes, How to Defeat Saddam Hussein (HERO Books, 13 January, 1991). That book was published four days before the outbreak of the war. Summarized below are some of those pre-war forecasts, produced by Dupuy and his colleagues of HERO-TNDA. No other organization or individual made forecasts before the war -- or even before its last week -- as prescient, or as close to accurate, as those of Dupuy and HERO-TNDA. A. U.S. Casualty Cost of the War Dupuy Estimate (preferred scenario, 30-42 days): KIA: 1154; WIA: 4605; Total: 5759 Typical pre-war Estimate: Jack Anderson, quoting "Pentagon sources" (apparently correctly): 20,000 to 30,000 killed in two weeks. Other sources were hardly better. Actual Casualties (as of 14 March, 1991): KIA: 125; WIA: 357; MIA: 21; Total: 503 B. A Post-War Casualty Re-Assessment On the basis of what is now known of the effects of the air campaign upon the Iraqi troops in and near Kuwait, the effect of surprise, and a reassessment of their combat effectiveness (to the value which Dupuy assessed for the 1973 October War), it is possible to recalculate the casualties of the four-day ground war*, as follows: Estimated Ground War Casualties: KIA: 79; WIA: 417; Total: 496 From Pentagon information and newspaper reports we can estimate the actual ground war casualties approximately as follows: KIA: 95; WIA: 368; Total: 423 It will be seen that the re-estimated Dupuy casualties are 17% higher than the actual casualties, although the Dupuy estimates for killed in action are 20% lower than the actual figures. These discrepancies, however, are well within the 50% margin for error postulated by the methodology. ___________________________________________________________ * Calculation details will be supplied by HERO-TNDA on request. B. Duration of War Dupuy Estimate: 30-42 days; after the outbreak of the war Dupuy publicly predicted on national television and radio that the war would be over before the end of February, 1991. Typical Estimate of other sources: Six months or more. Actual Duration: 43 days; ended 27 February, 1991. C. Duration of Ground War Dupuy Estimate: 10-14 days. Typical Estimate: Many months Actual Duration: 5 days (4 days and 4 hours) D. Relative Effectiveness of US and Iraqi Forces Dupuy asserted: (1) The basic relative combat effectiveness superiority of US forces with respect to Iraqi forces was about 2 to 1; (2) The airpower superiority of US forces would multiply basic effectiveness of US forces by a factor of about 3; i.e., with airpower considered, the effectiveness ratio of US forces to Iraqi forces would be about 6 times the numerical ratio. Other authorities spoke fearfully of the defensive capability of "battle-hardened Iraqi troops," and respectfully of the still greater effectiveness of "the elite Republican Guard." No one else looked at the history of Iraqi military performance since 1941, or the relevance of the historical example of the effect of unopposed airpower on ground troops in the 1967 Arab-Israeli War. E. Strategic Concept of War Dupuy forecast that the main UN effort would be a wide strategic envelopment by US armored and mechanized forces through the Iraqi desert west of Kuwait, combined with a holding attack against the Iraqi fortifications in southern Kuwait. Other authorities warned of the difficulties and costs of finding or opening avenues of approach through the "fearsome" Iraqi fortifications in southern Kuwait. Aside from Dupuy/HERO-TNDA, it seems that only General Schwarzkopf and his staff saw the Iraqi deployment as an opportunity instead of a problem.