January 1942, Derby House, Liverpool, England, Western Approaches Command. Britain is losing the Atlantic. In 1941 alone, German U-boats sink 1,299 Allied merchant vessels.
The country imports 70% of its food and 95% of its fuel. Every ton of it arrives by sea.
Admiral Karl Dönitz’s wolf packs are not simply sinking ships. They are sinking Britain’s ability to continue fighting at rate of loss, Churchill’s planners calculate the country cannot sustain the war beyond 1943.
The Royal Navy has a response, convoy. Group the merchant ships together, surround them with escort vessels, make the U-boats work harder for every kill.
It is not working. Through the summer of 1942, U-boats are hitting as many as 146 Allied ships in a single month.
Escort commanders are following doctrine. Lookouts are watching. Depth charges are being dropped. And yet, the merchant ships keep dying.
Often inside convoys that should have been protected. The Navy knows the attacks happen at night.
They know the submarines submerge before escorts can respond. What they do not know is something more fundamental.
They do not know where the U-boats are firing from. The institutional answer to the U-boat problem in 1942 pointed outward away from the convoy toward the ocean perimeter.
Since the First World War, Royal Navy doctrine held that submarines attacked from outside a convoy’s edges.
They would approach submerged, identify a target through a periscope, fire from distance, and withdraw.
Everything about British anti-submarine procedure was built around this assumption. Escort ships patrolled the flanks.
Sonar operators listened outward. When an attack occurred, the standard response was to fan out, spread the escorts away from the convoy in the direction the torpedo had come from, drop depth charges, and sweep back.
It was logical. It was coherent. And in 1942, it was consistently failing. The loss figures were not marginal.
Between January and November 1942, U-boats sank more than 1,000 Allied vessels in the Atlantic theater.
Convoy SC 107, sailing in November 1942, lost 15 of its 42 ships over 5 days, a 36% loss rate in a single sailing.
Convoy ON 127 in September 1942 lost seven ships in three nights. The escorts performed their doctrine correctly.
The doctrine was wrong. Commander Roger Winn, who ran the Admiralty’s submarine tracking room in London, was one of the sharpest analytical minds in British naval intelligence.
He tracked U-boat positions through radio direction finding, dead reckoning, and when available, decrypted Enigma traffic.
Winn understood the scale of the problem better than almost anyone. His conclusion, shared by the Admiralty’s senior staff, was that the solution lay in faster intelligence.
Break the ciphers quicker, route convoys around the wolf packs, deny the submarines contact entirely.
Meanwhile, Admiral Sir Max Horton, who took command of Western Approaches in November 1942, was a decorated veteran submariner from the First World War.
He was respected, experienced, and openly skeptical that any exercise conducted on a floor in Liverpool could tell him something the sea had not.
When first presented with the wargaming unit operating at Derby House, Horton considered it a diversion from real operational work.
He was not being irrational. He was applying a lifetime of experience to a problem that required something his experience had not yet encountered.
What nobody in the institutional structure had yet done was stand directly above an accurate reconstruction of an attack and watch it happen from the beginning.
Captain Gilbert Roberts arrived at Derby House in January 1942 with orders from Churchill himself, find out what is happening and sink the U-boats.
Roberts was a Royal Navy officer who had been placed on shore duty while recovering from tuberculosis.
He was not a current fleet commander. He was, in the Navy’s internal accounting, a convalescent.
What he had was a method. Roberts established the Western Approaches Tactical Unit, WATU, on the top floor of Derby House.
His staff was drawn almost entirely from the Women’s Royal Naval Service, the Wrens. Most had never served at sea.
Some were recent school leavers. Several were still teenagers. One of them was Janet Okell.
Okell joined WATU as a tactical analyst and wargaming specialist. She had trained in mathematics and had the specific habit of mind common among the Wrens Roberts selected of treating a tactical problem as a spatial and statistical one rather than an intuitive one.
She did not bring combat experience. She brought something the escort commanders lacked. She had no prior doctrine to protect.
The floor of the WATU game room was covered in linoleum. Roberts and his team drew convoy formations on it in chalk, placed wooden ship models in their correct positions, and reenacted actual battles from the official records, movement by movement, 2-minute intervals at a simulated 10 knots.
Jean Laidlaw, 21 years old, was responsible for statistical analysis. She tracked outcomes, logged variables, and over dozens of replays began to identify what the official reports alone could not show.
The chalk and the models were not decoration. They were the instrument. The discovery came from a single anomaly in a reconstructed battle.
In late 1941, Convoy HG 76 had been attacked with heavy losses. Roberts and his team laid out the entire engagement on the game floor, 48 ships in 12 columns.
The tracks of three confirmed U-boats plotted against the escort positions from the official record.
Roberts began moving the convoy forward in 2-minute steps, exactly replicating the documented action. Then, he stopped.
If the U-boats were firing from outside the convoy perimeter, as doctrine assumed, it was geometrically impossible for HMS San Flaviano, positioned near the center of the formation, to have been hit.
The torpedo track didn’t work from the outside. The angles didn’t fit. Roberts called the Flag Officer Submarines in London.
Admiral Sir Max Horton himself answered. Roberts asked him a direct question. In the last war, would you ever have entered the columns of a convoy to fire a torpedo?
Horton’s answer was immediate. Of course, it is the only way of pressing home an attack.
Roberts reset the game. He placed a U-boat model inside the convoy columns, approaching from astern on the surface, where diesel engines gave it a speed advantage over the merchant ships, and where lookouts almost never watched.
The geometry resolved instantly. The attacks made sense. The U-boats had been hiding in plain sight inside the convoys they were destroying.
Laidlaw named the countermeasure they developed Raspberry. When any ship in the convoy was hit, any officer of any rank could immediately initiate the procedure without waiting for command authorization.
Escorts fired star shells to illuminate the area. One escort swept directly astern. Others converged on the convoy perimeter, then turned outward in coordinated sweeping arcs, combing the entire surrounding water surface and blocking escape routes simultaneously.
A surfaced U-boat caught by Raspberry either had to remain on the surface and be attacked, or dive and lose its speed advantage against depth charges.
The Navy’s senior staff received the findings with institutional caution. Horton, who had been contemptuous of the wargaming unit since its establishment, was invited to play the game himself.
He played five rounds as the U-boat commander, opposing an escort commander he could not see, separated by a canvas screen with small holes cut in it, replicating the limited visibility of night operations at sea.
Horton lost all five of his submarines. When the canvas screen came down, his opponent was Janet Okell.
She was 20 years old and had never been on a boat. Horton, by all accounts, was furious.
Then, he ordered WATU’s recommendations implemented across the fleet. May 1943, North Atlantic, multiple convoy operations.
WATU begins training escort commanders in early 1943. The sessions run continuously at Derby House.
Convoy after convoy of escort officers cycling through the game room, playing both sides of the board, learning to recognize the attack pattern from inside the formation, drilling raspberry until the response is automatic.
The results are not gradual. They are immediate. Through the summer of 1942, before WATO’s tactics reach the fleet, U-boats are hitting an average of 96 Allied vessels per month.
By March 1943, following initial implementation, that figure begins to fall. By May 1943, the number of ships sunk drops to 49.
By June, it reaches 27. The shift on the other side of the ledger is more severe.
In May 1943 alone, the Kriegsmarine loses 41 U-boats. It is the worst single month of losses the German submarine service has suffered in the entire war.
Dönitz calls it Schwarzer Mai, Black May. On May 24th, 1943, he issues the order that ends large-scale wolfpack operations in the North Atlantic.
He withdraws his submarines. The battle that Churchill had called the dominating factor of the whole war is, in operational terms, over.
The human cost on the German side was not abstract. Over the course of the war, 28,000 of the 40,000 men who served in the U-boat arm did not return.
May 1943 was the month that rate became unsustainable, even by Dönitz’s own calculation. He wrote in his war diary that the losses were too high to continue, a phrase that, in the context of a man who had spent 3 years accepting catastrophic attrition, carried specific weight.
On the Allied side, the merchant mariners who had been dying in the Atlantic since 1939, men whose service branch had the highest casualty rate of any in the British Armed Forces, continued sailing.
They had no way of knowing yet that the worst was over. WATO trained 5,000 officers before the war ended.
The course Roberts and the Wrens developed was eventually adopted by the United States Navy as well, running equivalent sessions for American escort commanders.
Janet Okell continued at WATO through the end of the war. Jean Laidlaw remained in her statistical analysis role.
Roberts was awarded the OBE. The Wrens who built and ran the tactical system, Okell, Laidlaw, and 63 others, received no equivalent public recognition.
Much of their specific work remained classified until the 1990s and was not fully documented in accessible historical records until the 21st century.
The WATO game floor, the chalk markings, the wooden ship models, none of it survived the war.
Here is the fact that reframes everything you just watched. The U-boat tactic that was killing Britain, submarines entering convoy columns on the surface at night, firing at close range from inside the formation, had been standard German practice since at least December 1941.
Allied escort commanders had been fighting it unknowingly for over a year before WATO identified it.
The Royal Navy did not lack the firepower to stop it, or the courage, or the ships.
It lacked someone willing to stand above the problem and draw it out in chalk.
