Surrendering the Initiative: A Command Decision

Predicting what the enemy is going to do makes our commanders reactive, not bold

Perhaps the most valuable training opportunities available to our land forces are the three combat training centers (CTCs).1 They are national military resources worthy of every attention. Each training center is built upon five pillars: adequate land, professional observer/controllers, instrumentation, a world-class opposing force, and a robust base support structure. The CTCs are a crucible where doctrine is developed and tested, providing unequalled feedback about the doctrine’s validity and application. Current training trends from our CTCs2 suggest problems exist in the way we execute portions of our Army-based Military Decision-Making Process (MDMP).3 Specifically, units place too much reliance on expected enemy actions during the MDMP, which makes the decision-making process more reactive. Does the way we practice our doctrine lead commanders to believe enemy actions and intent can be predicted with precision? Predicting an enemy commander’s most likely course of action (COA) is highly speculative at best, yet maneuver commanders rely heavily on the intelligence officer’s (S2) predictions of enemy tactical actions, rather than the commander forming his own estimate, based on his maneuver experience, knowledge, and training, that focuses on achieving his own mission. The mission statement, the most important information in the operations order (OPORD), follows the enemy situation. By placing the enemy first in the OPORD, we focus our efforts on reaction to the enemy, instead of action against the enemy. When does the enemy’s mission and objectives, if ever, take precedence over our own? Have we become reactive to enemy actions through the CTC experience? The training centers have developed highly trained opposing forces (OPFOR) that maintain a unique understanding of friendly force units. The OPFOR’s advantages typically cause friendly units to become largely reactive in the way they fight. Friendly units fight the CTC battle as a cohesive task force team only once, while the OPFOR fights dozens of times per year, on their own terrain. The limited land available at the CTCs also forces the OPFOR to become somewhat predictable. Land is a resource not likely to significantly increase,4 which limits the CTCs to executing a finite series of exercises. Battles or battle sequences are fairly constant and often repeat themselves, and terrain and the effects of terrain remain constant. After numerous battles, the OPFOR will settle into a standard process for attacking or defending a specific piece of terrain. At the NTC, for example, the OPFOR repeatedly attacks through the Brown-Debnam terrain complex. This fight occurs during every National Training Center (NTC) rotation, year after year, making Brown-Debnam perhaps the most fought-over piece of terrain in the world. Consequently, the OPFOR knows the terrain very well. Furthermore, the OPFOR are U. S. soldiers, and therefore share the same cultural biases, thought processes, institutional training, and ultimately similar conceptions of warfighting as the friendly unit. The OPFOR’s advantages encourage many friendly units to become reactive in their fight. This is certainly not an argument against the value of the CTCs. But commanders should realize that a “reactive dynamic” can permeate the command when fighting the OPFOR. This should force commanders to take afresh look at the way they plan their tactical warfighting operations. In the attack, the OPFOR uses a battlefield framework of see, shape, strike, and shield. This translates into establishing communications and reconnaissance, protecting the force on the approach, isolating the point of penetration, creating a penetration with fire and maneuver, exploiting the penetration, and blocking the enemy reserves. Reacting to these enemy actions permeates our mission planning and allows the enemy to make decisions for us. Thus we surrender the initiative before the battle begins. Battles become a matter of stopping the enemy from accomplishing his objectives rather than setting the conditions for us to accomplish ours. General U. S. Grant put it best by saying: “I am heartily tired of hearing what Lee is going to do. Some of you always seem to think he is going to turn a double somersault and land on our rear and on both our flanks at the same time. Go back to your command and try to think what we are going to do ourselves, instead of what Lee is going to do.”5 (author’s italics) The process that commanders and staffs use to “examine a battlefield situation and reach logical decisions”6 is the MDMP. A continual seven-step process, MDMP never really ends but culminates once the objective is secured or the unit is issued another mission. The process is a lock-step method of defining our, and the enemy’s, objectives and allocating resources to achieve our ends. The staff officers construct a seemingly endless stream of estimates that build upon each other as battlefield influences change. But the estimates all have one thing in common. They begin with receipt (or prior to receipt) of the mission, and are not expected to reach an acceptable level of refinement until after the course of action analysis (wargame). That is, except for the S2’s estimate. This estimate is expected to have high resolution almost immediately, and is included as part of the mission analysis briefing to the commander. While this is doable given a somewhat predictable CTC enemy, it is by no means realistic in a fluid combat or rapid deployment environment. The mission analysis concludes with the staff delivering a briefing to the commander that provides him with the specified and implied tasks inherent in the mission. The S3 operations officer, however, does not begin the briefing. The first briefer is the S2, who provides the commander the expected battle effects of weather and terrain. At battalion level, the S2 is normally a junior officer with little or no maneuver experience. He is expected to be an expert in friendly and enemy doctrine in order to provide predictive analysis. Suddenly, the S2 has transitioned from his objective, scientifically-based terrain and weather analysis into a predictive role, assessing the enemy’s most likely and most dangerous COA. The S2’s ARMOR — 7 product now drives the planning process, because commanders and staff frequently weight the main effort based on a predicted most likely enemy COA. Overreliance on the S2’s product is compounded further when the MDMP process is time-constrained; the S2 may have only 30-50 minutes to prepare an enemy estimate. So, at the conclusion of the briefing, the commander issues guidance to his staff on fighting the battle, based in large part on the S2’s prediction of the enemy’s most likely COA. The military observer asks — how is it that the S2, prior to even establishing a solid reconnaissance plan, can predict the enemy’s most likely COA? Certainly the S2 must develop possible enemy courses of action, but can he really predict the most likely? Even the S2 “bible,” Field Manual (FM) 34-130 Intelligence Preparation of the Battlefield (IPB) states “History repeatedly demonstrates that those who predict only one COA are often surprised by the enemy.”7 Yet, at our CTCs, commanders repeatedly weight the friendly effort based on the enemy’s most likely COA, as predicted by the S2. The commander assumes significant risk in passing the initiative to the enemy when his staff develops a plan linked so decisively to supposed enemy actions. The question remains, why is the friendly effort weighted so heavily on the S2’s intelligence estimate, just one of many factors affecting the battle environment, instead of the commander’s estimate? These questions strike at the heart of our doctrine and military decision-making process. Even though CTC “train-ups” are usually a unit’s number one training priority, the same deficiencies emerge year in and year out in the after-action reviews of units training at our CTCs. These include S2s failing to predict the enemy’s intentions.8 It’s as if we never learn from our mistakes. The first problem is that the S2 is required to predict the enemy’s most likely and most dangerous COA early in the process, presumably so the staff can develop friendly courses of action. But perhaps the relevant question is not, “Why can’t the S2 predict the enemy’s intentions,” but, “Can we really predict the enemy’s intentions?” Probably not, since even at the CTCs, where enemy actions are generally finite, S2s have difficulty predicting enemy actions with any recurring accuracy. This dichotomy has created a debate in the military intelligence community between two very unique approaches to the intelligence analytical methodology, the so-called “capabilities” versus “intentions” schools of thought.9 The capabilities school of thought says the S2 should provide an estimate of what the enemy could do to keep us from accomplishing our mission. Conversely, the intentions school says the S2 must determine what the enemy will do. The capabilities school asserts it makes more sense for the S2 to present the commander with a set of enemy COAs. These COAs outline the courses the enemy could adopt to thwart our plans, rather than engaging in the highly speculative enterprise of predicting enemy COAs. It further asserts that the commander, not the S2, is the senior intelligence officer in the command, and that the commander’s estimate should form the basis of all planning. Intelligence is just one of many factors on the battlefield, and when the S2 provides the commander an enemy capabilities briefing, the commander must weigh intelligence with a myriad of other factors and form his own estimate. Additionally, emphasizing what the enemy could do to thwart friendly mission accomplishment allows units to focus more on the friendly mission, while emphasizing what the enemy will do tends to make units more reactive. Finally, by determining what the enemy could do while remaining focused on our mission forces the staff to create a highly flexible plan with realistic, executable branches and sequels throughout the battlespace. This is conducive to mission-oriented “task/purpose” instructions to subordinates while maximizing “reconnaissance pull.” History is replete with examples of experienced commanders, much less S2s, who could not predict what their opponent would do, supporting the capabilities approach. Could General Hancock have predicted that Robert E. Lee would direct Pickett to charge on July 3, 1863? Napoleon mused about how, at the battle of Waterloo, Wellington did the completely unexpected, yet both commanders knew their opponents well! Napoleon speaking of “the grand knowledge of warfare” stated, “There are no precise, determinate rules. Everything depends upon…a thousand circumstances which are never twice the same.”10 General Patton wrote that battles were “simply an agglomeration of numerous small actions and practically never develop according to preconceived notions.”11 According to FM 34-130, the IPB manual, in order to predict threat COAs, the S2 must have, among other things, “identified every characteristic of the battlefield environment that might affect the operation (step 1);” next, “identified the opportunities and constraints the battlefield environment offers to threat and friendly forces (step 2)”; and finally, “thoroughly considered what the threat is capable of and what he prefers to do in like situations (step 3).”12 The noted military writer Colonel DuPuy identified 73 variables impacting the outcome of battles, but ten of them, including intelligence, were intangible.13 U. S. Army Major (Ret.) Forrest Davis wrote that for the S2 to meet the first requirement in predicting the enemy’s most likely COA, he would have to “comprehend at least the majority of DuPuy’s variables, collect all the appropriate information, and place them in relational balance to each other.”14 This is a daunting task, to say the least. Perhaps equally daunting is step 2, which identifies “the opportunities and constraints the environment offers to both friendly and enemy forces.” Said another way, this requires the S2 and his staff to thoroughly understand the seven battlefield operating systems (BOS) of both the enemy and friendly forces, their current and expected relative combat power at each phase of the conflict, and the terrain and weather effects on soldiers, weapons systems, and each BOS. At a minimum, the S2 must be an experiential expert on friendly and enemy weapons and weapon support systems. Step 3, to “thoroughly consider what the threat is capable of and what he prefers to do in like situations,” is the most difficult task. Since some of the greatest generals the world has ever known have failed this task, it is probably asking too much for an intelligence officer, a captain or major, to master this step. Essentially, the S2 must “become” the enemy commander, placing the totality of the commander’s varied, lifelong experiences into a comprehensible mental model, then think, feel, and decide like an experienced, senior-ranked, foreign maneuver commander. Ultimately, the variables affecting how two opponents will act and react in a battle to the death are too complex for any analytical model, or even human comprehension. Richard Fox, in his archaeology and analysis of the Custer battlefield, uses historical examples to illustrate that battles are not precise models. Rather, they are extremely confusing experiences. Order is difficult to maintain. Events are often shaped by accident, and tactical disintegration can occur.15 Requiring an S2 to definitively state the enemy’s most likely COA and most dangerous COA, based on the previous steps, is nothing less than an extremely speculative enterprise.

ARMOR —

A closer look at the capabilities versus intentions schools of thought reveals what may be the real problem, the apparent “disconnect” between the MI capstone manual, FM 34-1 Intelligence and Electronic Warfare (IEW), and FM 34-130 Intelligence Preparation of the Battlefield (IPB) FM. The IPB manual teaches S2s to predict threat COAs, rather than directing S2s to predict the enemy’s most likely and most dangerous courses of action in the absence of reconnaissance. The task, “to determine the enemy’s most likely COA,” is found in FM 34-1 and referenced only once, stating, “Intelligence should tell the commander his… (the enemy’s) most likely course of action.”16 The IPB manual, FM 34-130, does not teach S2s to predict a “most likely enemy COA,” and does not provide any tactics, techniques, and procedures on how to do so. Rather, it teaches predicting likely COAs and their order of probability. Unfortunately, we follow the IEW manual, which references only once that S2s should determine the most likely enemy COA. Now, certainly when an S2 ranks enemy COAs in order of probability, there will always be a most likely. But the problem is that we weight the enemy’s most likely COA in our MDMP to the exclusion of the other possible enemy COAs. In the book The Defence of Duffers Drift, the Boer “S2” (used as an example in the IPB manual) didn’t present his commander a most likely COA, but rather four enemy COAs “in order of probability as I gave them.”17 There is a significant difference in the degree and manner of emphasis between predicting a most likely enemy COA early in the planning process and allowing it to drive our planning, versus identifying four enemy COAs which can only be determined by thorough reconnaissance. The IPB manual states that the S2 should prepare “event templates and matrices that focus intelligence collection on identifying which COA the threat will execute.”18 We seem to ignore the IPB manual’s directive to not “overlook the less likely but still viable (author’s italics) COAs. Do not risk surprise by failing to consider all feasible COAs… Consider the following possibilities that might lead to ‘wildcard’ COAs.’”19 Rarely, if ever, do S2s articulate ‘wildcard’ COA factors in the estimate. And if they do, that COA normally correlates to the “throw-away” friendly COA. Further, we expect the S2 to tell us up front, in mission analysis, the enemy’s most dangerous COA, before we even develop our own friendly COA. FM 34-130 does not address enemy most dangerous courses of action. The S2 could say, “The enemy’s most dangerous COA is to air assault a battalion on top of our BSA” and be technically correct, but what does that really tell the commander? The S2 could predict any number of suitable enemy COAs which could be considered very dangerous. In actuality, the most dangerous enemy COA should be the one that makes us the most vulnerable when executing our own COA. Thus, it is impossible to predict prior to friendly COA development. Ultimately the enemy’s most dangerous COA is that which disrupts the friendly center of gravity when executing our COA. Therefore the enemy’s most dangerous COA should not be identified until late in the MDMP, at the later stages of the course of action analysis (wargame). In fact, the enemy’s most dangerous COA may be a branch or sequel rather than a “stand-alone” COA. Therefore, we question not only the S2’s capability to predict an enemy’s most likely COA during mission analysis, but why one sentence of doctrine from FM 34-1 drives the requirement for an S2 to do so. Worse, we require S2s to predict the enemy’s most dangerous COA, in the absence of written doctrine on how to do so. Why is it that we force our S2s to conduct intelligence activities not supported by doctrine in the IPB “bible,” FM 34-130? Early on in the planning process, there certainly has to be an intelligence focus, and S2s have an obligation to provide the commander probable enemy COAs, including objectives, and potential schemes of maneuver. But it is unreasonable to expect a captain or major S2, many of whom are not well founded in friendly maneuver doctrine, to predict during mission analysis the most likely COA with little more than a higher headquarters intelligence estimate. Ultimately, the commander must rely on his own insight and experience to determine the validity of enemy COAs and which he thinks are the most likely. If the S2 has constructed a robust reconnaissance plan to provide clear indications and early warning,20 he can evaluate incoming reports and provide some degree of predictive analysis. IPB is clear that in order to discern what COA the enemy has adopted, detailed, multi-source reconnaissance is required. In fact, the IPB manual says we must “identify those areas and activities that, when observed (author’s italics), will discern which COA the enemy has adopted.”21 The Boer “S2” even stated that the four COAs he thought the enemy would take were merely guesses: “We need to conduct reconnaissance of the river bed and the Kraal in order to find out which of these courses of action he has chosen.”22 Note that the entire Boer plan was in no way hinged on a speculative most likely COA. Rather, only through thorough reconnaissance would the COA be determined. While predictive intelligence may be what today’s commanders expect, they must understand it is a very risky and highly speculative enterprise. It is even riskier, and perhaps in no way practical, to expect the S2 to predict the enemy’s most likely COA, especially in the early stages of MDMP. Commanders must take more “ownership” in assessing enemy intentions, place greater emphasis on friendly mission accomplishment, and form their own estimate. BG Richard Quirk, G2 of the 24th Infantry Division (Mechanized) during Desert Storm, wrote that while Army doctrine states within a division the G2 is the senior intelligence officer, the doctrine is wrong. “It is the commander who is the senior intelligence officer in any command.”23 BG Quirk reminded himself of COL E. C. Townsend’s dictum that, “In any command, there should only be one estimate – the Commander’s Estimate.” Further, that, “The Intelligence Officer should not be permitted to publish his personal opinions to a command.”24 COL Townsend’sand BG Quirk’s assertions are supported in current Army doctrine. FM 100-5 Operations clearly states that intelligence is the commander’s responsibility.25 Intelligence failures at our CTCs begin with the commander not identifying his intelligence needs, and his failure to provide detailed, focused guidance to the S2. Therefore, commanders fail to form a viable, relevant command estimate. FM 100-5 states “The commander drives the intelligence effort.” This clearly means the commander, not the S2, is responsible for the intelligence effort. Next, that “He must ask the right questions and focus the intelligence work.” Thus, the commander must provide specific guidance to his S2 defining his intelligence needs. Additionally, “He must know the enemy; the commander’s personal involvement and knowledge have no substitute.” This implies the commander as the senior intelligence officer in the command. If his guidance to the S2 is focused and clear, the S2’s estimate will remain relevant throughout mission planning. Finally, “He helps his intelli-

ARMOR — 9 gence system work effectively by stating his intent and decisively designating his priority intelligence requirements.” Ultimately, the intelligence system belongs to the commander, not the S2. To remedy current tactical trends, we must radically change our thinking. First, change the format of the five paragraph OPORD and place the friendly mission statement, commander’s intent, and the task/purpose for subordinate units, prior to the enemy situation.26 This will reinforce to commanders and staff that the primary focus of our efforts should be our mission, not what the enemy is expected to do, and will consequently serve to restrict a reactive dynamic. Second, understand that predicting enemy actions and intentions is highly speculative and cannot even begin to be accomplished until thorough reconnaissance is conducted. Because the commander drives the intelligence effort, he is responsible for training (through the chief of staff or XO) the S2 as a functional member of the battle staff. The S2’s success or failure is a direct result of the commander’s action or inaction. Third, commanders must form their own estimate, based in part on the enemy situation, and clearly articulate to the S2 what intelligence he requires to form the command estimate. It is difficult (and in a 96-hour deployable Army, nearly impossible) to predict what a real-world enemy will do, thus the S2’s estimate on the enemy’s intent should not drive the mission process. The commander must understand that Army doctrine clearly establishes intelligence as his responsibility, and he not only relinquishes significant authority by overreliance on an S2 estimate, but assumes significant mission risk if he does so. The U. S. Army is highly agile, technologically advanced, and remarkably lethal. The ability of our maneuver formations to close with and destroy the enemy is unsurpassed among modern armies. The resources the U. S. Army brings to the fight today is unequalled in human history. We will certainly get there with “the fustest with the mostest,” but all this is moot if we keep allowing ourselves to surrender the initiative.

Notes 1The three are: The National Training Center at Ft. Irwin, Calif., The Joint Readiness Training Center at Ft. Polk, La., and the Combat Maneuver Training Center at Hohenfels, Germany. 2The Center for Lessons Learned (CALL) “CTC Trends,” published quarterly, continually points to recurring deficiencies in the S2’s Enemy Estimate. 3MDMP is a model found in FM 101-5 Staff Organization and Operations consisting of Receipt of Mission, Mission Analysis, Course of Action Development, Course of Action Analysis, Course of Action Comparison, Course of Action Approval, and Orders Production. 4NTC has been negotiating for land to expand the battlespace. 5Shelby Foote, quoted in the PBS Series The Civil War, and Horace Porter’s, Campaigning with Grant, 1897. 6FM 101-5, Staff Organization and Operations, p. 5-1 7FM 34-130, Intelligence Preparation of the Battlefield, p. 2-43 8Recurring deficiencies are even acknowledged by Military Intelligence professionals. LTC Mike Flynn, formerly senior observer/controller for intelligence at the Joint Readiness Training Center and G2 for the 82nd Airborne Division, states, “Too often, S2s are not able to produce this predictive intelligence for a number of reasons.” “Intelligence Must Drive Operations: How Intelligence Can Clear the Fog of War,” Military Intelligence Professional Bulletin (MIPB), Jan-Mar 2000. 9Arguments outlined in Colonel Elias Townsend’s, Risk: The Key to Combat Intelligence, (The Military Service Publishing Company, Harrisburg Pa., 1955). 10Herold, J. Christopher, editor and translator of The Mind of Napoleon, A Selection from His Written and Spoken Words, Columbia University Press, N. Y., 1955, p. 223. 11Blumenson, Martin, The Patton Papers, 1940-1945, Houghton, Miflin, and Co., Boston, 1957, p. 436. 12FM 34-130, p. 1-3 13Colonel Trevor N. Dupuy, Numbers, Predictions, and War, Hero Books, Fairfax, Va., 1985, p. 33. 14Major Forrest L. Davis, Predictive Intelligence: Do We Really Need It?, MIPB, Apr-Jun 1997. I have borrowed heavily from Major Davis’s arguments. 15Fox, Richard A., Archaeology, History, and Custer’s Last Battle, University of Oklahoma Press, 1993. Fox cites other studies, as well as the archaeology of the Custer battlefield, to show the unpredictability of battle, as well as how units can fall victim to tactical disintegration. A few of the books he cites include Du Picq, A., Battle Studies, Military Service Publishing Company, 1946; Keegan, J., The Face of Battle, Biddles Publishing, 1978; Marshal, S. L. A., Men Against Fire, Peter Smith Publishing, 1978. 16FM 34-1, Intelligence and Electronic Warfare, p. 2-7. 17FM 34-130, p. 2-39. 18Ibid., p. 1-3. 19Ibid., p. 2-43. 20FM 34-1, Indications and Early Warning is one of six Primary Intelligence Tasks, the other five being Perform IPB, Perform Situation Development, Perform Target Development and Support to Targeting, Support Force Protection, and Perform Battle Damage Assessment, p. 2-8. 21FM 34-130, p. 2-39. 22Ibid., p. 2-39. 23Colonel Richard J. Quirk III, “Intelligence for the Division – A G2 Perspective,” U. S. Army War College, p. 165. 24Ibid., quoting pg. 29 of Colonel Townsend’s work. 25FM 100-5, Operations, p. 2-12. 26Perhaps two OPORD formats are required. One for the offense that places friendly forces, mission, intent and task/purpose first, and one for the defense, which states the friendly forces and mission, followed by the enemy situation, and then the friendly force intent and task/purpose.

MAJ Kenneth Deal was commissioned Armor from the University of Idaho in 1986 and is currently the XO for the Training Support Battalion supporting 278th ACR, Tennessee Army National Guard. Previously, he served in tank and scout platoon leader positions in both Korea and with the OPFOR at Ft. Irwin, Calif. He has served in battle staff positions at brigade/regimental and TF level. He commanded in 3rd ACR and held various observer/ controller assignments at the Combat Maneuver Training Center, Hohenfels, Germany, culminating as a TF S3 trainer. He has previously published articles in Military Review and ARMOR. He will report to the Defense Threat Reduction Agency to serve in Functional Area 52 (Nuclear Weapons Research and Operations).

CPT Paul Carter was commissioned Military Intelligence from Officer Candidate School, Ft. Benning, Ga., in 1991, and is currently the Military Intelligence team chief for the Training Support Battalion supporting 278th ACR, Tennessee Army National Guard. Among other assignments, he served as an enlisted intelligence analyst in the 7th Infantry Division (Light) and the 703d Military Intelligence Brigade, Schofield Barracks, Hawaii, and as a battalion S2 and G2 operations officer in the 82d Airborne Division.

ARMOR —