In May 1942, the Second World War was expanding across the Pacific.
An Imperial Japanese navy invasion fleet was steaming toward Port Moresby in Papua New Guinea. To protect that amphibious force, Japan sent the aircraft carriers Shokaku and Zuikaku to patrol the adjacent Coral Sea. The United States navy carriers Lexington and Yorktown sailed out to meet them.
The opposing carriers fought each other on May 8, 1942 in the Battle of the Coral Sea. American bomber aircraft heavily damaged Shokaku. The Japanese planes were more successful: Yorktown was disabled and Lexington eventually sank. (Its wreck was recently discovered.) However, the fierce fighting caused the Japanese invasion fleet to turn around. That saved Port Moresby.
This was the first carrier-versus-carrier battle in history. Tactically, it was a Japanese win. But preventing the invasion was a key success for the American-Australian alliance. The outcome also kept Shokaku and Zuikaku out of the decisive Battle of Midway one month later.
Consequently, the Battle of the Coral Sea is still commemorated in Australia and the U.S. each year.
But were those results certain to happen? If either side had made different decisions, could the battle have turned out differently? If so, how?
I investigated those historical questions using two mathematical studies. I began each one by calibrating a math model with historical data to reproduce the battle’s results. (The model was originally created for naval missile combat.) I then adjusted the model’s inputs to reflect different decisions by the U.S. and Japanese navies. The model’s outputs estimated the battle’s new outcome.
My research asked several “counterfactual” or “what if?” questions. For example, one-quarter of American aircraft at Coral Sea were fighters. (The rest were dive bombers and torpedo bombers.) The fighter proportion grew to almost two-thirds by 1944. That trend gave carriers stronger defences but less offensive power.
What if U.S. carriers at Coral Sea had likewise carried two-thirds fighters and one-third bombers?
The model suggests this would have left the Americans worse off. They would’ve suffered slightly less damage themselves but inflicted much less damage on the Japanese. That would have hindered the Americans during any subsequent exchange of airstrikes.
The U.S. carriers sailed together during the battle. But what if they had separated, as some American officers had advised? Travelling together, carriers can share fighter cover for mutual protection. But sailing separately, one might avoid detection if the other is attacked.
My calculations indicate that separating the carriers would have slightly improved the Americans’ results: One carrier sunk, but one left unharmed. (This assumes the separated carriers could have still coordinated their attacks.) However, this tactic condemns the detected carrier to certain destruction. Sailing together gave a small chance of both surviving.
What if the U.S. had added another carrier? Two were busy executing the Doolittle Raid on Tokyo, led by Lt.-Col. James “Jimmy” Doolittle of the U.S. air force. By cancelling that raid, America could have sent one or both to the Coral Sea.
My model estimates a third U.S. carrier would have boosted Japanese losses. But it wouldn’t have reduced American losses much. In fact, it would have exposed more U.S. ships to attack.
Four carriers would have given the U.S. a 95 per cent chance of destroying both Japanese ships. But that would have tied up its entire Pacific carrier fleet in one place. And it still wouldn’t have reduced American losses much.
On the morning of the battle, each navy frantically searched for its opponent. Since they found each other around the same time, their attacks occurred simultaneously. But if one side had remained undetected, it would have struck first. Only the survivors on the other side could have counterattacked, much like at Midway. What if that had happened at Coral Sea?
Calculations show striking first would have been decisive for either side. It would have been even better than having an extra carrier. The other side’s survivors would have been too weak to counter-attack effectively. Both sides were right to pursue that goal.
This research examined several plausible alternative decisions for the Coral Sea fight. It confirmed the battle easily could have turned out differently. The alternate outcome would have had an impact on the larger war.
For example, suppose the Americans had succeeded in attacking first. Lexington and Yorktown likely would have taken minimal damage. That could have given the U.S. a larger margin of safety defending Midway one month later. Shokaku and Zuikaku likely would have sunk, putting the Japanese farther behind in carriers after Midway. That might have shortened the Pacific war.
Conversely, a Japanese first strike would have been disastrous for the United States and Australia. With both American carriers neutralized, Japan would have captured Port Moresby. Both Shokaku and Zuikaku then could have supported the Midway invasion, giving Japan an overwhelming advantage there.
These results remind us that aircraft carriers are both powerful and fragile. Analysts have discussed that contradiction since carriers first arrived. They continue to discuss their pros and cons now in the era of anti-ship cruise missiles.
Michael J. Armstrong, Associate professor of operations research, Goodman School of Business, Brock University