According to Wikipedia and some non-scholarly articles online, admiral Harry Yarnell pulled off an exercise eerily similar to the attack on Pearl Harbor 9 years earlier and was blown off by battleship adherents. Ernest King pulled off a similar attack in a 1938 exercise (see previous link).
My question(s) is/are about the other "side" of this story. What changes did Yarnell want the Navy to make and why were they rejected? Were there any legitimate reasons? How would one fairly evaluate the tradeoffs here without hindsight bias? If this really was a colossal screw-up, should we attribute it to bad conventional wisdom, to groupthink, or to something institutional (like Navy politics or misaligned incentives)?
Quote from Wikipedia, for context:
In February 1932, Yarnell pioneered carrier tactics in an exercise called Army/Navy Grand Joint Exercise 4. Rear Admiral Yarnell commanded the carriers Lexington and Saratoga in an effort to demonstrate that Hawaii was vulnerable to naval air power. The expectation was that Yarnell would attack with battleships, but instead he left his battleships behind and proceeded only with his carriers to the north of Hawaii where it was less likely he would be detected. With a storm as cover, at dawn on Sunday, 7 February, Yarnell’s 152 planes attacked the harbor from the northeast, just as the Japanese would ten years later. The army airfields were first put out of commission after which Battleship Row was attacked, with multiple hits on navy ships. No defending aircraft were able to launch. The Navy’s war-game umpires declared the attack a total success, prompting Yarnell to strenuously warn of the Japanese threat.[3]
The New York Times reported on the exercise, noting the defenders were unable to find the attacking fleet even after 24 hours had passed. U.S. intelligence knew Japanese writers had reported on the exercise. Ironically, in the U.S., the battleship admirals voted down a reassessment of naval tactics. The umpire's report did not even mention the stunning success of Yarnell's exercise. Instead they wrote, "It is doubtful if air attacks can be launched against Oahu in the face of strong defensive aviation without subjecting the attacking carriers to the danger of material damage and consequent great losses in the attack air force."[4]
This is a fun question, because it doesn't just involve the history and study of Pearl Harbor, but also touches on what wargaming is, and interwar naval doctrine as it was developed.
The short answer: Exercises like these are not necessarily meant to be predictive. Even if they were, there were vast improvements and changes in naval technology and strategy between 1932 and 1941 that would have made the results of 1932 hopelessly outdated by 1941.
The long answer: It is worth noting that there were many "predictions" of an attack on Pearl Harbor well in advance of 1941. There were multiple books that suggested that a war in the Pacific would begin with or include an attack on the US base in Hawaii; most notably Hector Bywater's "The Great Pacific War." It was not a new idea in 1940 or 1941, which is one reason why I am reluctant to give too much credit to the British attack at Taranto for "inspiring" the Japanese attack. According to some accounts, Japanese Admiral Isoroku Yamamoto, then just a captain in command of the aircraft carrier Akagi, lectured or discussed the possibility of an attack as early as 1928 or 1929. This was well before there was a serious chance of hostilities between the two countries.
It is also worth noting that Pearl Harbor was not the primary US navy base in the Pacific until about April of 1940. The ships had mostly been moored in San Diego or Los Angeles, but were moved to Pearl Harbor as a "deterrent" to check Japanese ambitions in the Pacific, with the approval of Franklin Roosevelt, likely upon the recommendation of the State Department. You could have attacked Pearl Harbor with the US ships on the West Coast, but the damage would have been much less.
To start with: The US Navy carried out annual "Fleet Problems" from 1923-1940, mostly in the Pacific (but sometimes in the Atlantic/Caribbean) and often around Hawaii (but also sometimes on the West Coast or around the Panama Canal, etc.). You might think of these as large-scale "wargames" that were planned well in advance with ships and personnel split up onto teams, with judges assigned, and scenarios or missions to play through. However, the objective of these games was not necessarily to test out individual tactics or to simulate exact scenarios, such as an attack on Pearl Harbor. Their purposes would have been mostly to provide training in simulated wartime conditions for officers and men, as well as to test new strategies, technologies, or doctrine.
Naval technology and doctrine in the 1920s, 30s, and 40s was progressing rapidly. For example, the US Navy immediately after World War I had scant experience operating submarines and basically no experience with aircraft. Those were entire new arms of the navy that had to be developed in peacetime. These large-scale exercises were used for working out some of these ideas. Should scouting be carried out by destroyers? Submarines? Aircraft? How far can submarines actually see through their periscopes, and how far away will the fleet be able to spot them and fire on them? How do you incorporate an aircraft carrier into a squadron of battleships? Should it sail out front to maximize scouting range, or behind to protect it? How many aircraft carriers do you need? No one knew the answers to these questions in the 1920s and 1930s. They had to learn them, and the Fleet Problems were where all of these ideas could be put to the test.
The link you have provided is mostly correct but stretches the truth to the point of sensationalism (imagine finding sensationalism on the internet!). And unfortunately, the sources he links at the bottom of that post are dead, so I can't follow his trail. Here's a much drier source on the interwar "Fleet Problems," published in Naval Aviation News in 1962; you can find a pdf of the same article here.
A more thorough treatment of the subject can be found in Albert Nofi's "To Train the Fleet for War: US Navy Fleet Problems 1923-1940 " which was published in 2010 by the US Naval War College. You may be able to find a pdf version of it for free online.
But essentially, the situation that was set up in 1932 was not a precise analogue for the Pearl Harbor attack of 1941. In the first scenario in 1932, the Black (enemy) team had captured the Hawaiian islands and represented an occupying power, while Yarnell's Blue team was the American fleet. The Blue team successfully raided Black air bases on Hawaii and was able to move a landing force into position to "invade" Hawaii. In the next scenario in 1932, Yarnell's Blue force was based at Hawaii, meant to represent the US west coast, and supposed to raid Black team atolls that were represented by US navy bases on the west coast. This is another misunderstanding of people when they re-evaluate these interwar exercises; Hawaii wasn't even meant to represent Hawaii all the time in these attacks.
Yarnell's takeaway was not that Pearl Harbor was vulnerable to air attack. His lessons learned, according to Nofi, were that the Navy's two aircraft carriers at the time (Lexington and Saratoga) were inadequate and the Navy needed 6-8 carriers. Other lessons from 1932 were that the navy's airships might not be so useful in actual war, and that the heavy cruisers had trouble operating seaplanes because their designs allowed the ships to roll significantly. These are valuable lessons learned, but not especially sexy, especially knowing what we know now. But it's the kind of stuff the navy was working out at the time.
In the third exercise of the 1938 Fleet Problem you reference, a weakened American fleet was tasked with defending the French Frigate Shoals and Hawaii from an advancing Japanese fleet. Ernest King, leading the Japanese fleet, sailed in close to Hawaii and bombed the airfields, recovered his planes, and then was able to defend against a counterattack from PBY Catalinas operating from Oahu. One of the lessons the US Navy learned during the Fleet Problems was that the PBYs, while excellent scout craft, were slow and vulnerable if used in attack roles.
We should also discuss here that judges' rulings during wargames were often controversial and much-discussed. Army aviators, for example, argued throughout the interwar period that judges underrated the efficacy of high-level horizontal bombing of ships at sea. The Army pilots wanted credit for sinking more ships. The judges were skeptical. As it turned out, even skeptical judges had given the Army pilots too much credit; horizontal bombing of ships at sea during World War II was basically useless. The military did its best to learn from these interwar Fleet Problems and they are remembered as being hugely successful and influential on US Navy strategy and doctrine of the era, but they're not perfect.
Beyond that, if a judge had ruled in 1932 that carrier-borne aircraft were not capable of inflicting damage on battleships in Pearl Harbor, they would have been correct. The US Navy carrier aircraft of 1932 would have been much smaller, with weaker engines, and not capable of carrying the larger bombs needed to do anything more than superficial damage to the battleships of the era. The Navy had the Vindicator dive bomber and the Devastator torpedo bomber by 1938 but they still would have been teething. One key reason the American ships in Pearl Harbor were so vulnerable was because the Americans didn't believe the Japanese could construct a torpedo that could run in the shallow waters of Pearl Harbor. Normal torpedoes dropped from aircraft dove to 100-200 feet before rising to their run depths; Pearl Harbor was about 45 feet deep. The Americans also didn't know that Japan had spent 1941 training to perform at-sea refueling of its aircraft carriers and their smaller escorts, which couldn't carry enough fuel to make the trip to Hawaii and back on their own. The Pearl Harbor attack was controversial even in Japanese naval circles, because so few people believed it could be carried out successfully.
All of this is to say that yes, there were Fleet Problems that centered around an attack on Pearl Harbor, and yes, there were many similarities to the attack that the Japanese executed in 1941. However, to argue that these wargames should have provided warning for the US Navy, or that an attack was "predicted" is a vast overstatement and misunderstanding of what these Fleet Problems were intended to do. The interwar Fleet Problems are not remembered by the US Navy as missed opportunities, but as an excellent use of exercises to develop new strategies and doctrines in uncertain times. If you'd like to read more about professional wargaming and how it is used, I'd point you to "Successful Professional Wargames: A Practioner's Handbook" by Graham Longley-Brown, which does also reference US Navy interwar Fleet Problems.