by Attack on Pearl Harbor: Strategy, Combat, Myths, Deceptions
If 15% hits were expected, each AP bomb carried into the battle would have an expected value of one to two percent of a battleship, determined by multiplying the value of a hit by the hit probability. By this methodology, if all the hits were concentrated on a single battleship and no aircraft shot down before bomb release, the Japanese would need 50 to 100 bombers to sink a battleship.
The other model is stochastic. It assumes that a battleship is sunk or crippled by AP bombs not through cumulative damage, but by a catastrophic event such as a hit in a magazine or an engine room. Under this model each bomb has a chance to sink the ship. If luck is against the bombers, it is possible that the ship might take more than 18 hits without being sunk or crippled.
A stochastic model was developed by the US Navy Bureau of Ordnance during the war. A hit in an engineering space by an 800kg AP bomb could disable a battleship. If the expected hit percentage was 15%, and the engineering spaces constituted 20% of the target cross-section, then a single B5N Kate level bomber had a 3% chance of crippling a battleship. A hit in a main or secondary magazine or main battery turret could cause the magazine to explode and destroy the ship. Magazines constituted 23% of the target cross-section. Each bomber would have a 4% chance of hitting a magazine.
(7) AP bombing Anticipated Results
The chart above shows the result of a simulation of the AP bomb attack using the Bureau of Ordnance’s stochastic model, along with factors shown in the actual attack. In the simulation, ten groups of five bombers attack the three inboard battleships, three each against Maryland and Arizona and four against Tennessee. Each group has an 80% chance of scoring one hit on their target. If a salvo hit the inboard target, there was a 25% chance of also scoring a hit on the outboard battleship, based on historical results. Each hit had a 23% chance of detonating a magazine and a 20% chance of hitting an engineering space.
The results shown are an average over 1,000 repetitions of the attack. An average of 1.5 battleships were destroyed by magazine explosions (meaning that in about half the trials, one battleship blew up, and in the other half, two blew up, with a small number of cases where none or all three were blown up).
The number of crippled battleships might seem low considering the bombs had nearly an equal chance of hitting an engineering space as a magazine. The reason is that if a ship had both an engineering hit and a blown magazine, it was shown as blown up.
There was also about a 70% chance that one of the outboard battleships would be crippled or destroyed, important if the torpedo attack totally failed, and of little additional importance if the torpedoes were successfully delivered.
(8) Number of Inboard Battleships Crippled or Sunk
The chart above shows the data from the same 1,000 trials, this time showing the number of runs in which zero, one, two, or three inboard battleships were either crippled or sunk. Outboard or single-moored battleships are not included.
The chances that the level bombers would totally fail are vanishingly small. In over 80% of the trials, either two or three inboard battleships were either crippled or sunk.
The marginal value of an aircraft group in each role is revealing. A single group of five level bombers would have an expected return of one hit (80% hit on the target battleships, and if the salvo hits, another 25% chance of a hit on an outboard battleship). Each group would have a potential value of 0.43 of a battleship crippled or sunk.
The Japanese planners expected 67.5% hits with torpedoes, so five bombers would be expected to score 3.375 hits, sufficient to cripple or sink a battleship, or an expected value of 1.0.
So, even using the stochastic model, the expected return from a group armed with torpedoes was 2.3 times that of the same aircraft armed with bombs. The Japanese could have achieved better results if they had shifted several groups of level bombers to carry torpedoes, though this presupposes they had known that their technical solutions to the problem of launching torpedoes in shallow water would be successful.
Looking at the extreme case, the Japanese would have expected an additional 34 torpedo hits if they had armed all 50 of their level bombers with torpedoes. The three battleships double berthed would have been untouched, along with the one in drydock. In the actual battle, only one of these ships (Arizona) was crippled or sunk by AP bombs. In exchange, they would have had sufficient torpedoes to make all four outboard battleships nearly unsalvageable, with perhaps another 20 torpedo hits to distribute amongst cruisers and other fleet units. The damage to the fleet as a balanced force would have been much more extensive had those 50 Kate carrier attack bombers carried torpedoes, and if the Japanese were willing to expand their vision beyond battleships. An additional 20 hits could have sunk four cruisers and eight to twelve destroyers, sufficient ships to constitute a screen for a carrier task force in the early days of 1942.
In other words, the planners valued the chance of getting one or two inboard battleships over a higher probability of getting many cruisers, destroyers, and auxiliaries.
Attacking other ships was not in the Japanese mindset. They were not attacking the Pacific Fleet as a material force, but as a symbol of American naval power. Their audience was not the Pacific Fleet Staff, but the American people. They were taking a spiritual approach toward attacking the enemy’s morale and will to fight. The Japanese did not think in terms of maximizing the physical damage to the Pacific Fleet as a whole, but maximizing the damage against the popular symbols of American naval power.
Arguments could be made to either decrease or increase the allocation of level bombers. A decrease would free more bombers to carry the more lethal torpedoes. An increase would give the level bombers the potential to cripple or sink four battleships (the three inboard battleships and the one in drydock), thus accomplishing Yamamoto’s objective without requiring any contribution from the torpedo bombers. The level bomber attack was less uncertain than launching torpedoes in terms of weapons delivery and the effects of AA fire, or the possible interference from enemy fighters.
Shifting carrier attack plane allocation between AP bombs and torpedoes was not an option available to the Japanese on 7 December 1941. The allocations had already been made and the crews trained in one role and one role only. There was no flexibility to change crews from one mission to the other.
The allocation decision, made in late September, represented Japanese expectations at that snapshot in time. As the training process unfolded, at times the Japanese seemed to expect that they could sink every capital ship in the harbor with the combined level bomber and torpedo bomber attack, and that everything would go well; at other times, if Fuchida is to be believed, they worried that they might not sink even one battleship, and that the torpedo attacks would achieve nothing.
These bipolar mood swings were apparently communicated to the aviators, and could have affected their decision making during the attack.
A Shortage of Torpedoes?
The decision to include level bombers was not made because of any shortages of the specially modified shallow water torpedoes. It was not recognized until late in the training process that the torpedoes needed modification, well after the aircraft weapon allocations had been made.
On 17 November 1941, Kaga loaded 100 shallow-water torpedoes at Saeki Bay, before departing for the rendezvous at the Kurile Islands. There the torpedoes were distributed. The striking force departed for Pearl Harbor on 26 November. There were sufficient extra torpedoes to arm the 50 level bombers, if desired. But this was not a planned contingency, or even possible, since the 50 carrier attack planes assigned as level bombers had not trained for the tricky release conditions needed for a successful launch in the shallow, confined waters of Pearl Harbor.
Planning Inflexibility
The Japanese committed to their distribution of torpedo and level bombers very early in the process. It would have seemed logical to allocate weapons to the carrier attack bombers only after there was a better idea of the hit percentages that could be expected.
Mos
t of the B5N Kates could have initially trained in the torpedo attack role, the most difficult method. Sufficient lead bombardiers could have been trained separately to accommodate an expansion of the level bombing group. If the numbers indicated that fewer torpedo bombers and more level bombers were needed, torpedo bombers could have been shifted to the level bombing role as wingmen, since all the wingmen had to do was maintain formation and release their bomb when the leader dropped, a skill that did not require extensive training. A last minute allocation of weapons could have been possible, had the planners made such a provision. The Japanese had the intelligence support, but they elected not to use it to its full potential.
If the aircrews were capable of delivering either torpedoes or high-altitude bombs, the proper allocation decision point would have been the day before the attack, based on last-minute intelligence on torpedo nets, the number of battleships and carriers in port, and the number of double-berthed battleships. Instead, the Japanese made their major allocation decisions three months prior to the attack and held to them, regardless of emergent information that ought to have prompted changes in their plan. This could be another manifestation of Japanese inflexibility, or it could be because Genda thought the aviators could not become sufficiently skilled in both roles to allow last-minute shifts.
Whichever the case, the result was inflexibility, and that inflexibility limited what the plan could achieve.
The Japanese did show some flexibility with their level bombers. Their doctrine was to employ level bombers in a formation of nine aircraft (a chutai), dropping a pattern of bombs that could compensate for target motion when attacking an underway target. They broke doctrine by going to a smaller formation of five bombers. This was a vote of confidence in the bombardiers, that they could hit a stationary target without needing the larger bomb pattern. Since the distribution of the bombs in the pattern allow for only one hit per group per target, doubling the number of groups provided twice the chances to hit. By breaking doctrine the Japanese went from five salvoes (five chotai with a total of 45 aircraft) to ten salvoes (ten formations with 50 aircraft).
Breaking up the established chutai organization shows that the planners were not totally enslaved by doctrine. At the same time it raises the question why such flexibility was not demonstrated elsewhere.
Dive-Bombers
Eighty-one D3A Vals of the second wave were assigned to strike fleet targets. Most histories imply that this force was to finish the destruction of ships only damaged in the first wave. This is incorrect.
The dive-bombers’ first objective was the carriers. They were to sink them if they survived the torpedoes delivered by the first wave; if they were sunk they were still to bomb the hulks to make them unsalvageable. Lieutenant Commander Abe Zenji confirmed postwar that “the original plan for the dive-bombers was to attack the carriers.”
Next they were to go after cruisers. Only after the cruisers were destroyed were they to direct attention to battleships.59 This conformed with the capabilities of the 250kg GP bomb.
The dive-bombers trained to attack as a shotai or chutai. The lead bomber would attack in a 55-degree dive directed at a predetermined aimpoint, usually the ship’s mast. The other aircraft would follow in 20-to 30-second intervals. If the leader hit, the following aircraft would maintain the predetermined aim point; if he missed (likely deflected by the wind or target motion), his followers would adjust their aimpoints, just as a ship would adjust its gunfire based on the fall of shot. In practice runs against the battleship-sized target ship Settsu, an attack by a shotai of three dive-bombers “almost always guaranteed at least one successful hit on target.”60
The 1941 Combined Fleet Exercises saw dive-bomber hit rates improve to 55%.61 For the Pearl Harbor attack, the lead dive-bomber pilot, Lieutenant Commander Egusa Takashige, trained his crews. The release altitude was reduced from 600 meters to 400 meters for better accuracy. Lives were lost during training when aircraft did not recover from their dives in time. At 55% accuracy, the 81 dive-bombers going against the fleet would expect 44 hits.
The wargames provide an indication of Japanese expectations for 250kg GP bombs against carriers and cruisers. In the September wargame, air groups from four Japanese carriers sank two carriers and severely damaged one, and sank three cruisers and severely damaged three others. If we assume that the carriers and cruisers were all the targets of the dive-bombers, and that each carrier was hit by at least four bombs and each cruiser by three, that would yield a total of 30 bomb hits, or approximately what would be expected from four carriers’ dive bombers—if six carriers would score 44 hits, 4 carriers would score 29.
This corresponds to 1941 National War College lethality estimates. A carrier had a life of 6.8 equivalent penetrating 14-inch shell hits; cruisers 4.2 to 4.6. A 500-pound bomb had a value of 2.0, requiring four hits to sink a carrier and three to sink a cruiser. Given these assumptions, Egusa’s dive-bombers would be expected to have the firepower to sink up to 14 cruisers, or three carriers and ten cruisers. There were eight cruisers in Pearl Harbor at the time of the attack. The dive-bomber allocation appeared to be a good match for what would be required to kill the Pacific Fleet’s carriers and cruisers, with some overkill (as most military planners prefer). In the event, with no carriers in port, there would be extra dive-bombers available for other targets.
The calculations seem to confirm that the Japanese expected approximately 55% hits, matching their training results. This was optimistic. Over all of World War II in the Pacific, dive-bombers averaged about 20% hits against stationary battleship-size targets.62
Japanese dive-bombers delivered some of the very best and worst performances of the war. There were cases where every bomb from a chutai missed slow-moving merchantmen.63 On the other hand there is the case of the heavy cruisers Cornwall and Dorsetshire, caught in the Indian Ocean by a strike of D3A Vals that scored on the order of 90% hits and damaging near misses (DNM), sinking both ships in short order. The success against Cornwall and Dorsetshire did much to establish the D3A Val’s fearsome reputation in the history books.
The root reason for the wide variation in results is the Japanese divebombing technique. Wingmen followed as nearly as possible in the path of the leader, and corrected their bomb aim point based on the striking location of the leader’s bomb. If the leader had a good sight picture and good aim and a steady release, there was a good likelihood that the wingmen would hit. If the leader was off, then the following aircraft would be adjusting their aim point based off the results of a bad attack, making it very unlikely that following aircraft in the chutai would hit. If the leader’s aim was good but the wind strong it could be difficult for the wingmen to judge large corrections and apply them in the course of their dive. Any problem with the leader’s attack and the entire formation could end up scattering their bombs over the ocean. In analytic terms, each Japanese dive-bomber was not an independent event with a 55% probability of hit, but had their probability of hit dependent upon the (improved or degraded) skill of the lead bomber. The Japanese technique could be devastating, or it could be ineffective, with little middle ground.
Airspace control and target selection also were issues. With so many bombers from different formations going after targets in a restricted airspace, the likelihood of mutual interference was high unless tight controls were maintained. The dive-bombers had the same targeting problem as the torpedo bombers—concentrate to sink as many cruisers as possible, but with a minimum of overkill.
The Japanese plan metered the bombers in to the attack, with each shotai attacking only when the previous shotai completed their runs. With nine chutai in the attack, one chutai should have been allocated per cruiser, with one spare. The attack ought to have been a process of the nine chutai leaders coordinating their target selection, and the rest of the aviators just following their leaders down. Problems could arise with smoke from fires caused by the first wave, especially where targets were close together, such as the four cruisers tied
up at the shipyard piers—an early hit on one might obscure all four.
Nine aircraft at 20-second intervals would make for a three-minute attack. If four chutai went in one at a time against the four shipyard targets (the longest queue, the other cruiser targets being far enough apart that they could be attacked simultaneously), the attack would have lasted 12 minutes. This would mean a relatively drawn-out attack, and would allow the defenders to concentrate their defensive fire against each succeeding attacker, but it would minimize mutual interference in the bomb runs and allow some command and control over the distribution of aircraft to each target.
Approach: Time on Target (TOT)
(9) Japanese Aircraft Approach Routes, From Campaigns of the Pacific War
The first and most elementary planning consideration ought to have been simultaneity. In the first wave, while torpedo bombers were attacking the fleet at Pearl Harbor, dive-bombers and fighters were to attack Hickam, Wheeler, and Ewa Fields. It was within the technology of the time for the first-wave aircraft to arrive over every target area simultaneously, called a Time on Target (TOT) strike. This would eliminate any chance that one defending area could flash a warning to other areas.
If simultaneous arrival over all the target areas could not be achieved, the attackers would want the Pearl Harbor group to go in first to allow the torpedo bombers to attack before the defenses were manned.
The approach was planned to come down the center of the island. North of Pearl Harbor the two torpedo groups would split to go to their initial points (IP) east and west of the harbor. The distance to their IPs would be nearly identical, so a “nearly simultaneous” attack by the two torpedo groups was planned.
A simultaneous TOT for all formations would not have been hard to plan or achieve. There were several prominent geographic points on the northern coast of Oahu that could have provided navigational starting points. The formations that had the furthest to go would fly directly to their target, while the other formations could orbit in place and depart timed so that all the formations would arrive at their targets simultaneously. This “time and distance” exercise is one routinely inflicted on pilot and navigator trainees on the way to their wings, and was within the technology of the period.