A Game of Birds and Wolves: The Ingenious Young Women Whose Secret Board Game Helped Win World War II

by Simon Parkin

use against U-boats, encouraging escort ships to work together like team-mates, rather than individuals.

Around the edges of the room, Okell noticed great sheets of white canvas. They were arranged into enclosures, like voting booths, except each one seemed to have a peephole cut into it at eye level. The average visibility from the bridge of a warship is five miles, Roberts explained. The canvas sheets were positioned in such a way that, when a player peeked through the slit, he or she could see the equivalent of a five-mile view of the tiny wooden ships on the floor. Linen side wires, which could be bent to adjust visibility, depending on the game scenario that was being played, held the apertures open.

One team, positioned behind the canvas sheets at desks, played as the escort captains, he said. The other, usually captained by Roberts or his right-hand woman, Jean Laidlaw, played as the U-boats. As in the real Battle of the Atlantic, each side’s objective was focused on the convoy ships: the escort ships had to protect them, while the U-boats had to attack them. Each side had a secondary objective. For the escort ships, this was to sink as many U-boats as possible, while the U-boats hoped to avoid detection and exit the battlefield unharmed. The convoy ships, the prize in play for both sides, would automatically plod on at each turn of the game toward their destination, the battle raging around them, just as at sea. Next, Roberts explained the rules of the game. Players were given two minutes in which to submit their orders for the next ‘turn’, to replicate the urgency of a real battlefield. The movements of the U-boats were drawn in green5 chalk on the floor, a colour chosen as it was impossible to make out against the floor’s tint when viewed from an angle. This ensured the U-boat positions were undetectable to the players peering through the canvas screens. The escort ships’ movements would then be added to the floor in white chalk, which was, in contrast to

the green markings, legible to those peeking from the canvas holes. Turn by turn the pieces would move around the floor, as the escort ships dashed to the site of an explosion to drop depth charges, and the U-boats performed their feints and dodges in an effort to pick off convoy ships, while evading the escort. Okell’s role, Roberts explained, would be that of an umpire, measuring distances and marking movements in chalk to ensure that the game played out as accurately as possible. Finally, at the end of the game, the players would come together and, sitting around the board, now criss-crossed with chalk markings, Roberts would reveal how everyone had fared. More details needed to be worked out, Roberts explained, and by necessity the game was a simplification of real U-boat action, but there was enough here, he said, to begin staging recent sea battles. The objective was to experience the action from the perspective of the U-boats and, from that knowledge, assess what the escort commanders might have done differently to save ships, supplies and lives.

With the rudiments of the game in place, Roberts spent a great deal of his time studying after-action reports written by naval officers who had battled U-boats and survived, in search of clues to their tactics. Being ideally situated to meet and quiz any and every naval officer passing through Western Approaches Command, Roberts did not have to rely solely on the rather staid written testimony of sailors; he could also listen to first-hand accounts by interviewing men as they returned from sea. During the course of several interviews a chaotic picture emerged, not only of the sea battles themselves but also of the training process. Before deployment every sailor–an estimated 200,000 men by the end of the war8–underwent a two-week crash course in Tobermory harbour, on the Isle of Mull, under the gruelling tutelage of Sir Gilbert Stephenson (nicknamed, variously, ‘Puggy’, ‘Monkey’ and ‘The Terror’). Stephenson’s course was intended to build and improve efficiency among an individual ship’s crew, but it did not address the effectiveness

of ships working in company. Not only was there no universal set of tactics with which to fight U-boats, neither was there any training for how escort ships should work as a team. The destroyers and corvettes, it seemed to Roberts, were broadly free to direct their response according to individual whim or notion.

On the order ‘Buttercup’, he explained, all of the escort ships would turn outward from the convoy. They would accelerate to full speed, while letting loose star shells–an explosive that, like a dandelion puffball, released iridescent fragments that hung in the air on a parachute for up to sixty seconds and illuminated the ocean. If a U-boat was sighted, Walker would then mount a dogged pursuit, often ordering up to six of the nine ships in his group to stay with the vessel until it was destroyed.

(Doenitz, who had also watched the encounter play out on his plot, albeit almost in real time, realised that Audacity’s scout planes had led to U-131’s downfall, and issued orders that in the event of sighting an aircraft carrier in any convoy, every wolfpack was to prioritise attacks on this ship above all others.19) Seeing the battle from a crow’s-nest perspective above the board, it became clear to

Roberts that this early success in the battle had been a direct result of the unusually large number of escort ships that Walker had at his disposal. This had freed him to pursue the spotted U-boat while leaving the convoy ships with adequate protection.

Then, as he examined the plot, a question formed in Roberts’ mind. If the U-boats were firing from outside the perimeter of the convoy, how had Annavore, which was in the centre of the convoy, been sunk? Might it be possible, he wondered, that the U-boat had attacked the ship from inside the columns of the convoy? There was, he reasoned, a simple way to prove his theory. ‘Hold everything,’ Roberts told his staff, as he rushed into his office to make a phone call. Roberts picked up the receiver and asked the operator to put him through to the Flag Officer Submarines in London, hoping to speak to its chief of staff, an old friend, Captain

Ian Macintyre.* To Roberts’ astonishment, the flag officer himself, Admiral Sir Max Horton, picked up.

On the phone, Roberts explained who he was and asked Horton if he might be permitted to ask a question. During the last war, Roberts asked, would you ever have crept among the ships of a convoy to fire a torpedo? ‘Of course,’ replied Horton. ‘It is the only way of pressing home an attack.’ And out of interest, Roberts continued, what is the range of a U-boat’s electric torpedo? ‘Five thousand four hundred yards,’ replied Horton without hesitation.

Between them, Roberts and the two Wrens began to plot different scenarios that might have enabled the U-boat to sneak into the convoy without being detected. Only one checked out: the U-boat had entered the columns of the convoy from behind. And it must have done so on the surface, where it was able to travel at a faster speed than the ships. By approaching from astern, where the lookouts rarely checked, the U-boat would be able to slip inside the convoy undetected, fire at close range, then submerge in order to get away.

The game had enabled the fledgling tacticians to think like U-boat captains, and from that perspective the answer suddenly seemed obvious: having made your attack, you would of course dive. Then you would sit and wait for the convoy to roll overhead. ‘Eventually,’ Roberts concluded, ‘I would emerge, deep, from the stern of the convoy.’21 With the U-boat tactic abruptly unveiled, Roberts wanted to try out some potential countermeasures that might foil the plan. The four returned to the game room. Roberts assumed the role of the U-boat captain, and Laidlaw and Okell played as Walker’s escort ships. The countermeasure revealed itself immediately. Rather than splay out from the convoy at speed, dropping depth charges at random, Laidlaw and Okell lined the escort ships up around the convoy. While the convoy continued on its way, each escort ship performed a triangular

sweep, listening for U-boats on the ASDIC. With a mounting sense of excitement, the team ran the procedure twice more. In both instances Roberts’ U-boat was detected and sunk.

For the officers playing as escort commanders behind the canvas peepholes the games were keenly intense. From his position behind the canvas sheet, the lieutenant would survey the ocean floor, then report back to his CO, who sat at a plotting table behind the screen, as though he was working in the charthouse up on his ship’s bridge. As well as listening to the situation report from his lieutenant, the senior officer would receive constant new information from his assigned Wren, designed to mimic the flow of information arriving from the ship’s ASDIC radar operator. ‘Star shell fired here,’ she would tell him. ‘Explosion heard there.’ She might explain that another ship was trying to make contact on the radio, or that a destroyer had opened fire but that the splashes were not observed. From this information, the CO would have to make quick judgement calls on what to do next, based on what he understood the other escort ships to be doing. Naval officers would then write their instructions

on a small square of paper known as a chit. They might choose to alter their vessel’s speed or direction, fire star shells or drop a depth charge–any of the actions that were open to them in an actual sea battle. The chit was posted into a little box and duly collected by one of the young women, who passed the instruction to her fellow Wrens, who would proceed to shunt the models around the make-believe Atlantic Ocean accordingly, while kneeling on the floor and carefully marking off each instruction in pencil. In an adjacent room Bernard Rayner and another clutch of Wrens would collect, decode, decipher and transmit signals to and from the naval officers behind the screens. These messages would be delivered to each officer, presenting an almost overwhelming amount of information on which instantaneous decisions had to be made. To mimic the chaos of battle, Roberts and the Wrens would introduce unexpected diversions to proceedings: a lone freighter, for example, spotted by an aircraft and reported to be blazing just over the horizon.2 In this instance, one of the Wrens would gingerly walk across the floor and place grey cotton wool around the miniature ship to indicate billowing smoke. Another Wren, meanwhile, might suggest to her captain that he should send an escort ship to investigate and search for survivors.

The pressure of the two-minute intervals between turns mimicked the stress of action against U-boats at sea, and each officer would often be caught up in the fiction, no longer viewing the chalk lines and wooden models as game pieces, but as the real ships, wakes and explosions they represented. The game occupied an unusual position between reality and make-believe. No limbs or lives were lost here on the linoleum ocean. But neither was the game fully abstracted, in the way that Monopoly is based on, but distinct from, the property business. For the men who played WATU’s game, who had often returned from sea only a few days earlier, and who were often due to sail again a few days later, the game had an unsettling quality. The choices made on the floor reflected an of...

Each course, which lasted from Monday to Saturday*, and which ran weekly without interruption from the first week of February 1942 to the last week of July 1945, involved up to fifty officers at once. It consisted of four game scenarios, which each varied details such as the weather conditions, visibility, time of day and the size, speed and start point of the convoy. Finally, when the game finished the officers would step from behind the canvas screens and, along with the Wrens, sit in a square of chairs around the room. Then, with a ten-foot wooden pole Roberts would commentate on the preceding battle, blow by blow, like a sports pundit delivering a post-match verdict. He would draw attention to moments of particular brilliance, moments of particular disaster and the turning points of each game. These summations were, for Roberts, the most enjoyable aspect of the work. He relished the opportunity to create a picture of the battle far more vivid and engaging than the reports written by officers returning from earnest action at sea.

The U-boatman was describing an attack on an Allied convoy and used the words, in German, ‘And then no man may move, or anything.’ Roberts played these words over and over again. ‘What did it mean?’ he wrote in his diary. It could not be that the men had to be quiet while fleeing a battle on the ocean’s surface. ‘Oh no!’ Roberts wrote, his glee betrayed by every exclamation point. ‘It was when the U-boat had fired its torpedoes, gone deep, and was coming out of the stern of the convoy, deep and quiet in order to not be heard by the British ASDIC: what a proof for Raspberry!’

When he returned to Liverpool, Roberts commissioned the design and printing of these flick-books, which showed not only Raspberry, but all of the searches and operations designed by WATU.* These instructive booklets were routinely stolen and taken to sea by Officers undergoing the course.

‘Raspberry went like clockwork and whenever, during the night, the cry of “Tally-ho” was heard on the scram, I only had to check the bearings to know where a U-boat was being hunted.’

Still, the work was exhilarating, especially when the first fruits of WATU’s work began to be seen in summer 1942, when escort ships sank four times as many U-boats as the previous month,11 beginning an upward trend that would continue, broadly, for the rest of the year. The improvements in tactics were timely as they helped compensate for the fallout from an internal battle being waged between the navy and air force in London. The major source of contention between the Admiralty and the Air Ministry related to the deployment of long-range bombers. In Captain Walker’s battle of HG.76 in December 1941, a lone bomber had

played a decisive role in scattering a wolfpack. This led Walker to formally advise that aircraft were ‘absolutely invaluable’ to the protection of convoys. Following Walker’s report, the Admiralty repeatedly requested an allocation of bombers to provide air support (at one point the First Sea Lord Sir Dudley Pound requested 2,000 warplanes12). The Air Ministry repeatedly refused. In part, this was because Churchill had long asserted his belief that the only ‘sure path’ to victory lay in ‘an absolutely devastating, exterminating attack by very heavy bombers… on the Nazi homeland’.13 Knowing his rejection of the Admiralty’s request would be supported by the prime minister, Chief of the Air Staff Sir Charles Portal wrote: ‘To divert [the RAF’s bombers] to an uneconomical defensive role would be unsound at any time. It would be doubly so now when we are about to launch a bombing offensive… which will enable us to deliver a heavy and concentrated blow against Germany.’14

The U-boat situation in the Atlantic was ‘so grave’, Pound wrote, that a stand had to be made, even if it led to ‘the extreme step of resignation’.15 Again his pleas were rejected. Despite Churchill’s personal ties to the navy, and belief that the safe passage of convoys was key to Britain’s survival, the prime minister was enamoured with Air Chief Marshal Sir Arthur Harris, the commander-in-chief of Bomber Command. Harris had a talent for public relations and, in modern parlance, optics. When he sent a fleet of bombers to attack Cologne on 30th May 1942, the figure of 1,000 aircraft was chosen not for any strategic benefit, but in order to capture the public imagination via newspaper headlines.

The Air Ministry’s political superiority was made clear in the autumn of 1942 when the Admiralty’s chief of operational research, Professor Patrick Blackett, a distinguished scientist who later won the Nobel Prize for Physics, presented a comparative statistical analysis of the situation. His research showed beyond all reasonable dispute that a force of 200 long-range bombers would make a decisive contribution to the Battle of the Atlantic. Moreover, using aircraft in this way

would have a far more meaningful effect on the broader war than in their current deployment, bombing German cities. Even with meticulous statistical analysis Blackett struggled, as he later put it, ‘to get the figures believed’. By January 1943, just one squadron of twelve bombers supported the escorting of convoys. As Churchill’s chief of military operations later concluded, the prime minister’s ‘obsession for bombing Germany’ resulted in ‘the navy being very short of long-range bombers’, which was ‘the only well-founded ground for criticism of our central war direction’.16 All of this was not lost on those who had to fight at sea. The convoy commander Peter Gretton, who became one of WATU’s most distinguished students, later said that ‘co-operation between the Navy and the Air Force in the field was very bad indeed, mainly due to stupid quarrels between senior officers in Whitehall.’ The bombers may have been absent, but WATU’s work was aided in other ways. There was the invention and introduction of the ship-mounted high-frequency direction finder (HF/DF), a piece of equipment known to sailors by the nickname ‘huff-duff’. Two ships equipped with this vessel-mounted direction-finding gear could pinpoint a U-boat location by detecting and triangulating the source of the high-frequency radio transmissions it was

sending, either to other U-boats or to U-boat headquarters.* Then there was the fact that, after September 1942, the Western Approaches Convoy Instructions, the bible issued to all British escort officers filled with WATU-coined tactics and signals, was published for Canadian and American ships too, under the new title Atlantic Convoy Instructions. This book ensured escort ships from different Allied navies now used the same anti-U-boat signals where, prior to this, the British were just as bemused by the American signal ‘Zombie Crack’ as the Americans were by the British signal ‘Pineapple’.17 Finally, there was the introduction of Hedgehog, a bow-mounted anti-submarine weapon that could spray a volley of mortar rounds directly ahead of the ship, toward a suspected U-boat location. The name derived from the weapon’s appearance: the twenty mortar rounds were bunched together at a near-perpendicular angle, giving the appearance of a hedgehog’s spikes. The projectiles were primed to explode not by fuse but on contact, and entered the war just as the U-boat captains had become adept at evading depth charges. By ...

while diving. During the time it took for the depth charge to fall, the U-boat would have disappeared. The Hedgehog enabled the explosives to be hurled ahead, while the U-boat was still within ASDIC contact, and before it could turn and dive.18 In the months that followed the development of Raspberry, using information gleaned via debriefs, Roberts and the Wrens developed numerous other tactical manoeuvres to suit the expanding variety of wolfpack attacks. Most of these manoeuvres, which involved the escort ships performing different shapes and varieties of coordinated sweeps to find and hunt lurking U-boats, were given the memorable names of fruit and vegetables: ‘Pineapple’, ‘Gooseberry’, ‘Strawberry’, ‘Artichoke’ and a modification to the original manoeuvre, known as a ‘Half-Raspberry’. These new manoeuvres provided escort ships with the tactics needed to go on the offensive, hunting U-boats before they made an attack rather than,...

During the next few days Roberts and Laidlaw began to develop a replacement tactic, which involved exploiting the U-boat captain’s natural caution by tricking him into ordering his vessel to dive to avoid an escort moving into potential spotting distance. Hearing the escort pass by and believing that he had not been seen, the submerged U-boat would then move to what it assumed was a safer location. Wargames and experience suggested this would be a slow turn designed to place the U-boat in a position parallel to the convoy while

conserving its batteries. The escort ships would dash towards its predicted position, using the rumble of the convoy ships’ propellers to mask their approach. The idea was that the U-boat, having realised that it was not under attack, would be too occupied with the wider convoy battle to notice the redirection of its adversary.

The course completed, Poole learned that her offshoot school was to teach the captains of escort ships how to make evasive manoeuvres to dodge torpedoes. One of these torpedoes had been captured by the Allies and dismantled.* Through this process, the Allies figured out that the German torpedoes had to keep within a certain angle of approach, or lose their

target. It was therefore possible to outrun a torpedo, providing the target executed a particular zigzag pattern. To teach the men this evasive manoeuvre, Poole was given a giant simulator. A ship’s bridge, almost full size, was built and installed in the warehouse at Gladstone Dock. It sat within a see-saw mechanism that allowed it to simulate the movement of the sea. Poole would run the game with a ship’s captain, flag lieutenant and navigator, who were used to working together at sea. ‘The [players] would all be swanning around doing nothing, then the ASDIC would pick up the torpedo,’ Poole recalled. ‘Then it was action stations, and all began to happen.’ Through a pair of speakers, the warning signal would sound and a pre-recorded ASDIC ‘hit’ would play, indicating the angle of the incoming torpedo. The crew would have to perform evasive manoeuvres using the giant apparatus. If they failed to outrun or dodge the torpedo, an obnoxiously loud banger known as a thunder-flash was let off by one of Poole’s fifteen staff members.* They would also release smoke capsules to mimic the disorientating effects of a direct hit, when the alarms howled, the floor tipped and, in the most extreme cases, visibility shrank to a few feet.

Poole, like Roberts, designed a clutch of different fake battle scenarios–eight in total, which lasted forty-five minutes each–and the ship’s crews would often return to replay through each variation. As with WATU, word of Poole’s game spread quickly. Soon she was training the majority of escort crews that came through Liverpool, including those of aircraft carriers and, much later in February 1944, the senior officers of the giant battleship HMS King George V, who took three lessons in total. Unlike WATU, however, Poole’s outfit was never given a formal title. To allocate a woman to such an operation was unprecedented and represented a significant victory for Laughton Mathews in her secondary campaign to loosen the tight constrictions on female opportunity. Poole was one of the many Wrens who, as Laughton Mathews later put it, ‘had started in very junior positions and had developed abil...

Elsewhere at the training facility, the men responsible for setting the timers on and firing depth charges were drilled till they could fire full patterns of explosives at fifteen-second intervals, building efficiency that would, at sea, often mean the difference between a hit and a miss on a U-boat. Teams responsible for operating the ASDIC sonar systems hunted imaginary U-boats on the synthetic attack teacher while, in a hut at Gladstone Dock, next to Poole’s torpedo-evasion trainer, communications staff were taught how to use a new cypher machine. According to Kretschmer’s captor Donald Macintyre, the ‘pleasantly secluded’ hut was run by a ‘very glamorous Wren officer’, and his younger officers made ‘remarkable… slow progress’, requiring ‘a great many lessons before the intricacies of the machine were mastered’.

Against all regulations the pair arranged to meet up. Roberts invited Handley to WATU’s confidential premises on a Sunday and explained the difficulty he was finding, as a ‘chairbound fellow’, in capturing the attention of his exhausted students who, after multiple trips across the Atlantic, were often in desperate need of rest and relaxation. Handley gave Roberts a number of ‘gimmicks’ to deploy with his hands. ‘He was immensely good at telling me how to talk and excite their interest’, Roberts later recalled.21 Comics transmute the angst and sorrow of their lives into comedy; they spill their guts on stage in the hope and belief that the specificity of their experience will prove, somehow, universal. Roberts was not telling jokes

on the radio, or in front of booze-loosened crowds, but there was a symmetry in his and Handley’s roles: to pull, from the darkness of the imagination, scenarios that brought the listener to attention. Handley’s advice proved effective. The Canadian officer A. F. C. Layard later described the WATU director as a ‘very good lecturer, very theatrical and, of course, would like you to know that he was seventy-five per cent responsible for the recent defeat of the U-boats in the North Atlantic. He’s probably right.’

Moreover, with the Reich now on the defensive on all other fronts, Doenitz was convinced that the U-boats represented the sole path to victory.

The timing was opportune. America’s entry to the war had exerted significant additional pressure on British shipping. Each American infantry division dispatched to Europe required 32,000 tons of shipping to cross the Atlantic.8 This used up space that might otherwise be used for civilian food supplies, refrigerated goods, oil and raw materials. As such, more than 150,000 cubic metres of frozen food destined for Britain was left rotting in American ports.9 The competing demands of troops, food, armaments and resources had ratcheted tension in the supply chain to an unprecedented degree.* ‘We are trying with the equivalent of about one third of normal fleet to feed this country and maintain it in full war production,’ wrote Lord Cherwell, Churchill’s scientific adviser. ‘With all the extra military demands that have emerged… it is not surprising that our imports… have suffered severely. But this cannot go on.’10 The Ministry of Food issued a reminder to the Cabinet that without sufficient imports of flour, meat and other key foodstuffs, Britain could survive only for between four and six months.11 In fact, by March 1943 Britain was consuming at least 750,000 tons more goods than it was importing. The country’s reserves would last for just eight weeks. Seven hundred thousand tons of shipping had been lost to U-boats in November 1942. It was time for Doenitz to press the advantage. In his first conference with the Führer as grand admiral, Doenitz convinced Hitler that no skill...

construction, in order to have total control over the allocation of steel, but had been denied this power by Raeder, Doenitz’s predecessor. By relinquishing control of naval construction in exchange for guarantees of U-boats and other assets, Doenitz could be certain of receiving the raw materials he needed. Finally, Doenitz would have his 300 U-boats. He would be able to destroy sufficient numbers of Allied merchant ships to remove food from British tables, fuel from British cars and heat from British homes.

The British were unaware that ONS.5’s route was already known to German intelligence, who had successfully intercepted their communications. At the height of his powers, and now, finally, with a sufficient number of U-boats to deploy his wolfpack strategy to its fullest, Grand Admiral Doenitz was poised for battle. In the days that followed, he would dispatch the greatest wolfpack ever assembled to attack a single convoy, one that would eventually consist of five times as many U-boats as escort ships. The odds against survival were great; ignorance may not have been bliss, but it at least allowed the convoy to set sail toward a battlefield from which anyone in possession of the facts would surely have fled.

At midday on 26th April, just as Gretton and ONS.5 entered the longitudes where the wolfpacks were assembled, the Germans made changes to the naval Enigma code settings that scrambled the messages sent between BdU, the U-boat headquarters, and the U-boats. In England, the Government Code and Cypher School, which intercepted and decoded these German messages, was abruptly blinded, unable to read the signals that, for months, had alerted ships to the whereabouts of U-boats and their plans. Without precise information, the best that the Submarine Tracking Room in London

could tell Horton, who was attempting to direct ONS.5 from his office at Derby House, was that the three U-boat groups were ‘in the general area off Newfoundland’.11 This blindness meant that nobody in the Royal Navy, least of all Gretton, knew that on 27th April, sixteen new U-boats had been instructed by BdU to assemble into a fourth wolfpack, Gruppe Star (Starling), east of Greenland, directly across the convoy’s path.

Then, in the early evening, Duncan picked up a U-boat close on the port bow. He made chase, ordering Tay to make a parallel search to port. Forty minutes later Duncan’s bridge spotted a telltale cloud of spray, where the waves had struck a U-boat conning tower. As Duncan approached, the U-boat dived. After dropping a pattern of ten depth charges, Gretton and Robert Sherwood, captain of Tay, executed the first WATU operation of the journey, Observant, a search of the area in a square shape, each ‘side’ two miles long, with the U-boat

boat’s last known position at its centre. As the weather worked against the convoy, so too it worked against the U-boats. Despite Gretton’s fear that the convoy was currently surrounded by a flock of attackers, only four of Starling’s U-boats had managed to rally to U-650. It was enough to unsettle Gretton. That evening, he listened in dismay as each member of the wolfpack made its evening report to BdU. The sound on the huff-duff was, as one report put it, like ‘a chattering of magpies’. Darkness fell and with it the likelihood of an attack rose. The air escort that ONS.5 had enjoyed from 24th April had been discontinued at midnight on the 27th, since, up until that point, there had been no Admiralty reports of U-boats in the vicinity. For now, Gretton and his handful of escort ships would face the U-boats alone.

Lamb’s ship was a member of the newly formed support groups, a rather plain name for highly trained, quick-moving flotillas that were sent into the Atlantic with the express purpose of providing impromptu assistance wherever it was required. The support group would, essentially, loiter in the ocean, waiting to be dispatched to the nearest emergency. The Oribi was currently escorting the convoy SC.127, en route from Sydney to Halifax. Horton estimated that it could be spared and, if it made good time, could be with Gretton within a day or two. The idea for these elite naval groups was devised months earlier by Sir Percy Noble, but he never had control of a sufficient number of spare ships to implement the plan. When Horton took over from Noble, he was equally convinced that this kind of flexible, hard-punching force would be decisive in the Battle of the Atlantic. On his first day at Derby House, Horton wrote to the Admiralty to press the ‘urgent need’ for these support groups. His appeal was rejected: there simply weren’t enough quick-footed warships available. Horton changed tack. He reasoned that since convoys still had to cross the Gap in the middle of the Atlantic unaccompanied, why not reduce the number of ships in the escorts either side of the area by one, and form the spare ships into support groups that could operate with more flexibility inside the Gap? After all, this was the only area of the ocean where the U-boats could still act with impunity and so represent...

SC.112, each headed toward three wolfpacks totalling thirty-eight U-boats. Concurrently, one of Roberts’ colleagues, Captain Neville Lake, added three fictional support groups to the wall, shadowing the real convoys. Lake hoped to estimate what effect such reinforcements might have on the looming battle. He populated his support groups with markers to represent real warships that were currently under construction, or nearing operational capability, and used his ghost escorts to support the real convoys, fending off wolfpacks while the battle played out in almost real time on the map. As well as adhering to the current weather conditions, Lake took into account fuel consumption and the depletion of depth charges to ensure that his exercise, overlaid on the real battle, was as accurate as possible. Although many of the Wrens forced to play out Lake’s game while simultaneously trying to plot the real battle were frustrated (they described it as ‘Lake’s folly’14), the final results were clear and, for Horton, conclusive. At the end of the battle for HX.229 and SC.112, which German radio reported as ‘Die grösste Geleitzugschlacht aller Zeiten’–‘the greatest convoy battle of all time’15–twenty-one Allied ships lay on the ocean floor. Lake estimated that, had his fictional support groups been real, at least eight ships would have been saved.

By the final week of March 1943, Churchill had made good on his promise: Horton had more than twenty new destroyers at his disposal, which he divided into five support groups ready to dash to wherever assistance was required.

An alarm bell woke Gretton, sleeping after the night’s work, and, at 07:30, one minute after the torpedo struck, he ordered the next WATU-coined manoeuvre of the journey: Artichoke. As one destroyer powered at maximum ASDIC speed toward the wounded ship, all others in the escort turned outward and performed a fifteen-knot sweep in line. U-258 managed to slip away. The McKeesport, listing to one side, managed to maintain convoy speed for almost fifty minutes while taking water, before the order was finally given to abandon ship at 08:15. The sixty-eight men on board were picked up by the crew of the Northern Gem, one of

Surrounded by ice and, having entered the Greenland air gap, now sailing under skies empty of air support, Convoy ONS.5 proceeded through the night with a woefully diminished force of protectors. As the merchant ships entered the most perilous leg of the journey, just three of its former seven-strong posse of bodyguard destroyers remained. At Derby House, recognising the convoy’s vulnerability, Max Horton ordered another