Wargames

by Martin van Creveld

  Early in the twentieth century wargames were used to investigate, or help investigate, such issues as shifting from coal to oil and from ships with mixed armaments to all big guns vessels.83 The emphasis on planning did not last. When Admiral William Sims assumed the presidency of the College (for the second time) in 1918, he ruled that games would be held primarily for the purpose of providing players with decision-making experience. The games themselves were made much more sophisticated by adding submarines and aircraft to the order of battle. Also added were extremely detailed tables that reflected the characteristics of each class of vessel separately. Weather, visibility, the distance between ships, the angle at which shells struck their targets, etc. were all taken into account. The crown of creation was the so-called “War College Fire Effect System” by which the outcome of combat was calculated.84

  As Verdy du Vernois might have predicted, the desire to incorporate one factor after another in the name of “realism” soon caused it to become so complex as to be almost incomprehensible. Following a long tradition, results were presented in terms of the impact of the fourteen-inch shells fired by the battleships of the period: hence there was a clear bias in favor of guns and against more recent weapons such as submarine- and air-launched torpedoes. Another problem that afflicted the tactical games in particular were the three-minute turns each player was given to assess the situation, formulate his orders, and make his move or moves. Such a leisurely pace might have sufficed in an earlier age. However, as “fast” carriers – not to mention aircraft with speeds of several hundred miles an hour – entered service during the 1930s, it became hopelessly unrealistic. Perhaps worst of all, the system forced players to immerse themselves in a vast number of technical details. Straining to do so, they tended to lose any ability to see the engagement they were supposed to be leading as a whole.

  Of over 300 wargames known to have been played at the College during the interwar period, a little under half were of the operational and tactical level kind. Of those all but nine focused on a possible war with Japan. However, the US Navy during this period does not appear to have used wargaming in the planning process for any specific campaign as the Germans routinely did. What the players did do was work their way through numerous possible scenarios. Admiral Chester Nimitz, who commanded the US Navy in the Pacific during World War II, is credited with saying that, throughout that prolonged and enormously widespread conflict, “absolutely nothing” had come as a surprise except the kamikaze attacks. To that extent the tactical games did indeed prove very useful. This was especially true in working out and testing doctrines of ship movement – the idea that carriers and their supporting vessels should work in groups rather than individually – and the way they should deploy for battle.

  The more numerous strategic games were a different matter. Supported by a foreign intelligence department specifically created for the purpose, they were sufficiently sophisticated to take into account factors such as logistics and communications. Based on those, the navy concluded that operating in the western Pacific, where Japan had numerous bases whereas the US had few, would be difficult, time-consuming, and focused on seizing islands by means of amphibious landings. The Marine Corps did in fact start working in that direction from the late 1920s on, holding exercises and developing doctrine.85 The navy also concluded that modern fleets were so large and fast, and operated in such a dispersed manner, as to make a Tsushima or Jutland-style climactic battle between the main forces of both sides improbable if not impossible. The struggle, if and when it came, would be prolonged and consist mainly of mutual attrition. These conclusions proved broadly correct, except, of course, for the unimportant fact that the attack on Pearl Harbor, which appears to have been completely unforeseen, did take place and changed everything.86

  On the other side of the hill, or rather the ocean, the Japanese also engaged in wargaming. Modern wargaming, as distinct from Go, arrived in Japan almost as soon as the country started modernizing itself. By the end of the nineteenth century they seem to have learnt enough for the games to make a solid contribution to their success in the 1904–5 war against Russia, and in particular their victory in the great naval battle of Tsushima.87 Like their German mentors, the Japanese did not use the games merely for training purposes, but incorporated them into the planning and decision-making process. One series of table-top games was held at the Imperial Naval War College in 1937. It ended with the “defeat” of the British naval forces in the South China Sea.88

  As the clouds of war began gathering in the late summer of 1941, another and much more important series of games was held in Tokyo. Apparently the first to go was the army. It gamed the possibility of war with the US, but few details are known.89 Next, the Imperial Navy entered the picture. Normally the navy held its games in November−December, the season when storms made navigation over much of the Pacific difficult. This time, following the imposition of the US embargo on scrap iron and oil, they were pushed forward into the early days of September. The first round, again held by the General Staff, followed the 1937 pattern by focusing on operations in Southeastern Asia. The next step was to involve the Naval War College which gamed the protection of the sea lanes to that region once “Blue” forces had occupied it.

  These were merely preliminary exercises. Much more important was the main round, held at the Naval War College on September 11–17. The participants were the principal commanders of the Combined Fleet. This time the assumption was that a Japanese offensive against Southeast Asia would not only involve the country in war against Britain and the Netherlands (Indo-China was already in Japanese hands) but lead to American intervention as well. Accordingly, in addition to rehearsing that offensive, the exercise dealt with the problem of defending against a US Navy offensive coming from Hawaii. During the first half of October yet another series of wargames took place. The objective was to familiarize the Japanese commanders with their missions, which included the occupation of the Philippines, Guam, Wake Island, Hong Kong, Malaya, Java, and Sumatra. Details such as the size and composition of the forces involved, ports of departure, courses, rendezvous points, and even the individual beaches on which the landings were to take place were studied in detail. Presumably to preserve secrecy, the games were held not at the College but aboard the flagship of the commander of the Combined Fleet, Admiral Isroroku Yamamoto, the battleship Nagato. They worked very well: when Japan joined World War II two months later, many of the operations it mounted followed rather closely in the games’ wake.

  The strike at Pearl Harbor itself was gamed for the first time in 1927. It was launched by two Japanese carriers, the only ones available, which were accompanied by destroyers, cruisers, and an advance guard of submarines. However, the umpires judged that the damage the Americans had suffered was minimal. Not only was the “Blue” commander, Lieutenant Commander Tagaki Sokichi, criticized for his “rashness,” but as the game developed the “Red” side went on to mount a two-carrier attack on Tokyo itself. As if to add offense to injury, the game ended with those carriers making their escape in spite of the Japanese attempts to locate and intercept them. At the time carriers and carrier aviation were still in their infancy and the technology in use was quite primitive. Perhaps this fact helps account for the outcome of the games.

  Not surprisingly, the idea of a strike at the main US naval base in the Pacific refused to die. It cropped up repeatedly in accounts of Japanese military planning during the 1930s, and in September 1941 it too was gamed. Two questions in particular had to be resolved: was the operation feasible at all, and could secrecy be maintained? After problems of selecting the correct route – the northern one – and refueling the warships en route had been solved to the satisfaction of the umpires, the time came to ensure that the attackers would not be detected by aircraft flying off Hawaii. The commander of the “Red” forces was a certain Captain Kanji Ogawa. During the 1930s he had witnessed at least one American exercise to defend Pearl Harbor (two were held, one in 1932 and
another one in 1938) against just the kind of attack that was now being planned. Both “attacks,” incidentally, had been judged successful.90 As the Japanese concluded their games, the outcome was completely different: Ogawa and his team inflicted unacceptable losses on the attackers. Thereupon the plan of attack was changed and it was decided to have the carriers execute the last part of their journey at high speed and at night. This time things went much better.

  Admiral Chuichi Nagumo, the officer whose ships were supposed to carry out the attack, remained skeptical. For one thing, attacking Pearl Harbor meant diverting forces away from the coming operations aimed at conquering Southeast Asia, which, he believed, were far more important. Besides, he disliked the choice of the northern route with its stormy waters and considered the plan as a whole too complicated. Others on the staff felt that the games were too theoretical and that too much depended on the umpires. The latter underestimated American strength, put various arbitrary restrictions on the actions of “Red,” and slanted their decisions in favor of the “Blue” teams. Some even considered that the games epitomized the Japanese penchant for short-sighted, self-centered thinking. Others still, while conceding that the plans might be a tactical success, considered them strategically wrong-headed.

  Another round of wargames was held in October and resulted in several changes to the plans. On the one hand, the number of participating submarines was increased. On the other, the number of carriers was reduced from four to three; since all of these were long-range vessels, not only was the problem of refueling them made much easier, but more ships were left to participate in the southern operation. Eventually this particular change was cancelled, and the number of carriers attacking Pearl Harbor increased to no fewer than six as new solutions to the refueling problem were sought and found. Yet another round of games, the third in the series, was held. This time they suggested that two out of the six carriers would be lost. This, however, was considered an acceptable risk. Some officers also continued to express their doubts concerning the attack’s long-term effect, but they were silenced by Yamamoto.91

  What actually happened went only so far to confirm the games. First, if secrecy was indeed preserved and surprise achieved, then this was due primarily not to anything the Japanese had done or left undone but to the fact that their approaching aircraft were mistaken for a flight of B-17s due to arrive in Hawaii the same morning. Second, the attackers missed their most important target, i.e. the American carriers, which happened to be at sea on exercises. However, strictly speaking that failure was the result of faulty intelligence, not of a wargame that was improperly set up and played. Third, the attackers did much better than the wargames had led them to expect. Not one of the precious Japanese carriers was lost or even damaged: in other words, American land-based aircraft operating from Hawaii had proved completely ineffective. The only losses were those inflicted by anti-aircraft fire from the guns surrounding the base, and they took a greater toll of the second wave than of the first.92

  This success, so unexpected in some ways, in turn formed the background to the events that took place before and during the battle of Midway in June 1942. This was not Pearl Harbor. Japan and the US were in the midst of a ferocious war, meaning that achieving strategic surprise of the kind that had largely accounted for the success in the previous year was ruled out. Still the Imperial Naval Command hoped to achieve at least operational surprise concerning the time and place of their next attack. Their main problem was that they did not know the whereabouts of the American carriers; to ensure they would not be present, Yamamoto dispersed part of his forces in the hope of making his enemies follow suit.93 As in the previous year, the Japanese planners wargamed their offensive not once but twice. The setting was the newly commissioned battleship Yamato where the officer in charge, Yamamoto’s chief of staff Rear Admiral Matome Ugaki, had set up a large map table on which all moves were shown. The remaining participants had their maps spread out in surrounding cabins, using runners in order to simulate shore-to-ship and ship-to-ship radio communications.

  The scenario assumed that the deception had succeeded and that the US carriers would be absent from the scene of action. Accordingly, the main opposition was expected to come from the land-based bombers the US Army Air Force had stationed on Midway itself. Once again, Nagumo had his doubts. He asked what would happen in case the American carriers appeared out of nowhere and took the Japanese in the flank and rear. However, his position was weak – at Pearl Harbor he had erred on the side of caution. Consequently his objections were brushed aside. As it turned out, when the dice were cast to determine the outcome of combat during the second game the Japanese carriers suffered nine hits. Two, the Akagi and the Kaga, were listed as sunk. At this point Ugaki intervened, arbitrarily reducing the number of hits to three. He also re-floated the Akagi and ruled that it had only been slightly damaged. Later in the games the Kaga too was restored to the fleet, but that is a different story.94

  Ugaki and Yamamoto have often been criticized for the rigidity of their plans and their unwillingness to listen to objections. Yet it is necessary to make two points. First, the use of dice inevitably implies the possibility of obtaining statistically very unlikely results. Such results are quite capable of upsetting any kind of human reason and putting it on its head. In fact it is one of the umpire’s main tasks to cancel such results if, in his view, they occur;95 does anyone really think the fate of states and nations should be determined by the accidental throw of a dice? Second, Ugaki’s guess that Midway’s land-based bombers would be no more effective than those at Pearl Harbor had turned out to be correct. Though the aircraft in question launched several attacks and dropped many bombs, they failed to register a single hit; earlier, the same had happened in the Mediterranean where, until the German dive-bombers came to the rescue, high-level Italian bombing of Royal Navy vessels had been ineffective. What proved decisive was not Ugaki’s somewhat high-handed action but the above-mentioned fact that the Japanese deception plan had not worked. It did not work because the Americans had broken the Japanese codes;96 they were like a man with good eyesight attacking a blind one in the rear while he was otherwise occupied.

  Since a fair amount is known about the games in question and also about the way they interacted with reality, this might be a convenient point at which to stop and take stock. Spreading from Germany, where it was invented in the 1820s, military-type wargaming differed from almost all the rest in that it did not try to create conditions where each side would have an equal chance at gaining a victory. Instead players were presented with scenarios supposedly rooted in reality and encouraged to explore what could happen, what would happen, and what they themselves would do in case something similar actually did happen. The wargames were used primarily for two purposes, i.e. planning and preparation on the one hand and training on the other. Often players included not just individual officers but entire teams which thus learnt to work together. Though tending to be complicated and time-consuming, the games could also be used, and were used, for entertainment. But that is a topic we have already discussed.

  In McCarty Little’s words, when it was a question of training, wargames represented not just the best but practically the only method of simulating a conflict against a thinking, acting, and re-acting opponent. When the editors of Military Review published von Cochenhausen’s game rules in March 1941 they made exactly the same point. Provided the games are carefully handled, arguably this fact compensates for any shortcomings they may have. The games took two basic forms, “rigid” and “free.” In practice, as many of the above examples show, there was always a tendency to mix the two in order to rule out unlikely results and/or make the game faster and easier to play.97 So complex is the reality of war that devising rules to cover all possibilities, as they do in simpler games such as chess, was and remains next to impossible. As Clausewitz rightly points out, too often the results will be pure nonsense.98

  In all probability, rare was the wargame in which
the rules of play were not modified to some extent or another. One could even argue that, since almost all training has to be based on past experience, in games used for that purpose the players’ ability to gain insight into the how and why of things by examining the rules and modifying them, if doing so is considered appropriate, represents a bonus and not a problem. Relaxing the rules made an umpire necessary. Often the greatest difficulty consisted precisely in finding one who was both capable and objective. Conversely, it could be argued that, had such an all-knowing, all understanding person been available in reality, and assuming his authority was unquestioned, then wargames would have become superfluous, and suitable at best for training second-rate operatives. Wargames, in other words, are just as susceptible to human frailties as war itself, and indeed had this not been the case most of them would have been completely useless.

  Like all other methods used for the purpose, wargames provide only a very partial solution to our basic inability to foresee the future. For every scenario that was wargamed, probably fifty other possible ones never were. (When Nimitz said that every possibility was gamed, it meant that, over a generation or so, every sort of problem was encountered by somebody at some moment in some game. It did not mean that everybody was now prepared for anything, let alone that the navy as such had gamed everything and consequently knew how to cope with any problem that might arise.) Conversely, out of every fifty scenarios that were wargamed probably only one turned out to be correct in the sense that the events it foresaw actually took place in anything like the form that had been foreseen. The random numbers used to determine the outcome of combat, whether produced by dice or by more sophisticated methods, are a very inadequate substitute for extremely complex chains of causality into the dynamics of which we humans only have very limited insight.99 To be sure some wargames, German ones in particular, played an important role in campaign planning. On some occasions they provided warnings, on others they turned out to predict the future rather accurately, though what percentage of the total they may have formed is impossible to say. But does it matter? When everything is said and done, there appear to be worse ways to try and peer into the future than by doing the best one can to comprehend all of the relevant factors, establish the way they relate to each other as accurately as one can, and play them out, perhaps repeatedly and with variations, against an able and imaginative opponent who has done exactly the same.