80 Years On: Shattered Sword's Jonathan Parshall parses some Midway Counterfactuals

[Athelstane]

This weekend is the 80th anniversary of Midway, so it's not at all surprising to see Jonathan Parshall drop in today for another lengthy pod/video cast with Drachinifel.

But this one looked worth special mention here, because this time, unprompted, Parshall offers up some alternate history - or as he likes to call it, "counterfactuals." And these are the questions he looks at:

1. What would have happened if Midway had been 3 vs 5 carriers (the IJN brings Zuikaku)?
2. What would have happened if Midway had been 2 vs 5 carriers (the IJN brings Zuikaku, but the USN cannot bring Yorktown in time)?

And then, in turn, he also looks at a further variant of both, in which either a) the "good" Hornet shows up, or b) the "bad" Hornet shows up (i.e., whether Mitscher actually follows his orders in where he sends his strike groups out). First, though, he does model the existing battle (4 vs 3) with both Hornet variants, just to see what happens.

This gets modestly rigorous: Parshall employs a stochastic salvo combat model developed in cooperation with two Spanish scholars, Aneli Bongers and Jose L. Torres, in a 2017 article, "Revisiting the Battle of Midway: A counterfactual analysis."

So, here are the quick and dirty answers:

​

0(a): [OTL: 4 IJN vs 3 USN, good Hornet]: US wins a curbstomb, 4 IJN losses vs 0 US losses. In short, Parshall, concludes, "Mitscher sunk the Yorktown."​

0(b): [OTL: 4 IJN vs 3 USN, bad Hornet]: US wins as in OTL, 4 IJN losses vs 1 US losses. The model bears out history!​

​

1(a): [5 IJN vs. 3 USN, good Hornet]: 5 IJN losses vs. 1-2 USN losses​

1(b): [5 IJN vs. 3 USN, bad Hornet]: 3-4 IJN losses vs 2 USN losses​

​

2(a): [5 IJN vs. 2 USN, good Hornet]: 2-3 IJN losses vs. 2 USN losses​

2(b): [5 IJN vs. 2 USN, bad Hornet]: 2 IJN losses vs 2 USN losses​

This is interesting, because the results pretty resoundingly reject the the proposition, mooted over the years, that Yamamoto bringing Zuikaku along could have been a game changer. In fact, the Japanese still get smashed. The real difference maker, as it turns out, is Yorktown. As Parshall, Bongers, and Torres' analysis seems to consistently show, if the Americans bring all three carriers as they did OTL, the Japanese are going to get their asses kicked at Midway even if they bring all 5 intact KB fleet carriers. Even with Nimitz's crappy sub captains and Midway's crappy attack squadrons, American advantages are just too much for the Japanese to have much chance of overcoming.

The video is 3 hours long, but Parshall's actual discussion is only 7 minutes, starting at my timestamp:

Discuss amongst yourselves whether you find this analysis persuasive (though you really have to dig into the paper to see the math and the modeling).

 

[Athelstane]

P.S. No, the analysis doesn't look at variants where Saratoga shows up. But it's not hard to see that this would just amount to something approaching overkill - at least, if Yorktown is also present.

 

[Merrick]

Hmm... I haven't gone through the maths, but the analysis seems to assume that every attack is going to inflict significant losses, which puts an enormous premium on striking first. It then assumes (quite reasonably, given the superior American scouting and the initial Japanese strike on Midway) that it will be the Americans that strike first.
In fact, if you use the simple model:
- Every US carrier airgroup that strikes sinks 1.5 Japanese carriers
- Every Japanese carrier airgroup that strikes sinks 1 US carrier
- US strike first, followed by (surviving) Japanese
it matches the numbers above pretty closely

Parshall states outright that if Hornet's dive bombers & fighters had coordinated properly, they would have attacked simultaneously with the torpedo bombers "and by rights knocked out one or two Japanese carriers, leaving Enterprise and Yorktown to clean up by lunchtime".
The nearest parallel I can find OTL is Yorktown's attack (24 dive bombers, 9 torpedo planes) on Shokaku at Coral Sea (2 bomb hits, damaged but not sunk) and Hornet's attack (16 dive bombers) on Shokaku at Santa Cruz (3-6 bomb hits, severely damaged but not sunk). Assuming that a single airgroup at Midway - against a 4 carrier CAP - would be significantly more effective than either of those cases strike me as optimistic.

 

[Spherical_light]

So @Athelstane , your first sentence in the OP is typo-ed. Should be 80th, not 8th.

I've only seen half of the video so far, but there are a great number of ideas for PoDs and alternate considerations presented in this video, certainly from what has already been mentioned to the implication of Nagumo choosing to not attack Midway so early.

Several implicative ponderances include:

Japan choosing to ignore I.Y.'s insistence of a Midway operation and ordering consolidation instead, thusly sparing the KB from attrition and allowing air groups to be replenished as well as repair and refit to properly take place.

One of the first few viewer questions asked was about Operation Z and Parshall's answer was very intriguing; saying that if the KB had had a bit farther, spotting the British. The RN presence in the Indian Ocean, at least for a time, would've been severely hampered. Parshall bases his answer off conventional wisdom at the time; certainly a must view for those who have worked on an RN timeline before.

Due to several of the viewer questions however, some of the answers given tend to be at odds with other answers given. Then again, Parshall is answering mostly hypothetical questions focusing on not only Midway, but the Pacific theatre in general. So while there are certainly ideas for divergences, a lot of them have to be accepted with a side order of salt and caution.

 

[Captain Seafort]

One of the first few viewer questions asked was about Operation Z and Parshall's answer was very intriguing; saying that if the KB had had a bit farther, spotting the British. The RN presence in the Indian Ocean, at least for a time, would've been severely hampered. Parshall bases his answer off conventional wisdom at the time; certainly a must view for those who have worked on an RN timeline before.

Drach's wargame that produced the possible outcome of "the world's first carrier battle ends with the opposing carriers fighting each other in a surface action" has to be one of the weirder ATLs.

 

[Spherical_light]

Indeed! That would be quite an interesting TL to read. I still haven't gotten through the last half of the video...

 

[173rd LRRP]

Indeed! That would be quite an interesting TL to read. I still haven't gotten through the last half of the video...

Might consider if damage control/avoidance had been better on Lexington and US had four carriers and three fully experienced air wings at Midway

 

[mattep74]

Might co spider if damage control/avoidance had been better on Lexington and US had four carriers at Midway

And if Saratoga arrived early five.

Then the USN might actually sink a battleship or two also

 

[Athelstane]

So @Athelstane , your first sentence in the OP is typo-ed. Should be 80th, not 8th.

I actually had TWO spectacular typos. The other one was "Mitscher sinks the Hornet." When in fact it should have been "Mitscher sinks the Yorktown." Maybe I should have gone off to mow the grass before hitting the "Post Reply" button. 😬

I've only seen half of the video so far, but there are a great number of ideas for PoDs and alternate considerations presented in this video, certainly from what has already been mentioned to the implication of Nagumo choosing to not attack Midway so early.

Yeah. As Parshall points out, that was really the decision that doomed the Kido Butai, regardless of whether Zuikaku was there or not. Because it took most of Nagumo's strike forces off the table while a major US carrier task force is present and already knows roughly where the KB is.

Notice that the IJN don't repeat this mistake in the subsequent three carrier battles. At Eastern Solomons Nagumo was careful to use only Ryujo to attack Henderson Field, while keeping Shokaku and Zuikaku back until he could be dead sure there were no US carriers present. And of course, Ryujo gets sunk for its trouble, but at least it's not fatal to Nagumo in the way that the Fatal Five Minutes was at Midway. At Santa Cruz and Philippine Sea, there's no attempt at all at a land raid - everything is focused on finding and destroying US carriers.

 

[Athelstane]

Might co spider if damage control/avoidance had been better on Lexington and US had four carriers at Midway

But I think that's the critical thing about Lexington: the USN learned some important damage control lessons from Lady Lex's sinking, as Parshall himself has pointed out. The first one was put into practice for Midway on board Yorktown: Oscar Myers' idea of purging the fuel lines with CO2 as soon as the strike packages were sent off. Had they not done that, Yorktown likely would have been lost in Hiryu's first attack, and then Hiryu's subsequent strike would have moved on and probably located TF 16.

And if Saratoga arrived early five.

Then the USN might actually sink a battleship or two also

Man, people are getting greedy here...

 

[e of pi]

The "good Hornet shows up to OTL Midway" alternative is one I've toyed with as a thought experiment. Like, say that Hornet's strike commander in the air has engine trouble and has to turn back early before Torpedo 8 broke off around 0825. Through whatever devolution of command happens, then, either Waldron is in command, or somebody willing to listen to Waldron is in command, and the whole air group (at that point the only American carrier airgroup that day in reasonably solid assembled mass!) turns onto the historic Torpedo 8 course, arriving over the Kido Butai as a collected mass around 0915-0930. The air group was a little less experienced, and there's plenty of room for mistakes like the two Enterprise bombing squadrons diving on the same carrier instead of distributing and underperformance of torpedo planes even with fighter escort and dive bombers making more of a mess. Still, it's possible that massive explosions from dive bomber attacks and such draw the attention of the parts of VF-6 that had ended up picking up on VT-8 historically but didn't follow them into the attack. Anyway, writing a giant furball at this point and rolling dice on hits left to the person who dares write the AH, but by the time the last stragglers exit the area or are shot down around 0940 to 0945, there's probably one or two Japanese carriers out of action and aflame. A towering smoke plume could help over-determine that Yorktown and Enterprise's squadrons know where to go. With a CAP degraded by engaging torpedo planes and dive bombers and two carriers unable to handle CAP refuel/rearm, there may even be a chance for VT-6 to torpedo one of the damaged carriers in the confusion when they arrive around 0940, and then it's just a question of if Hiryu or anyone else manages to accidentally hide in a squall line or something when three more American dive bomber squadrons arrive at 1020 (the historical hammer blow, but here just another hammer blow).

As Parshall puts it, it's a clean sweep by lunch and a lot of the drama then depends on what happens with afternoon strikes on cruisers, battleships, and changes in Nagumo, Fletcher and Spruance's strategy with the Japanese carriers decisively out of the fight by noon. Still, it could be dramatic to write for somebody who unlike me can actually write Thrilling Air/Naval Combat (tm).

 

[e of pi]

Additionally, Parshall's notion that Nimitz was initially shaping a two-day battle, with the first day being land-based attacks and submarines to attrit the Japanese and fix their location, and then the carriers moving in for the coup de grace if and only if the odds looked favorable is an interesting one. If they'd kept back, and the Kido Butai had spent two waves on Midway the first day, but then suffered minimal casualties from land based air and submarines as historically happened, what would decision-making have looked like overnight? Do you commit the carriers to battle knowing the attrition strategy totally failed, except in the sense of the number of Japanese planes shot down over Midway, or do you leave Midway to defend itself against an invasion without even trying a carrier strike?

 

[Athelstane]

Do you commit the carriers to battle knowing the attrition strategy totally failed, except in the sense of the number of Japanese planes shot down over Midway, or do you leave Midway to defend itself against an invasion without even trying a carrier strike?

Everything we know about Nimitz's and Fletcher's decision-making at that point seems to point to a very probable "yes." I think the failure of the subs and the Midway air garrison to make any impression on Nagumo (save for the squadrons attacking Midway - 11 were destroyed (including three that ditched), 14 were heavily damaged, and 29 with lesser damage, which would point to a sizable attrition of Nagumo's strike assets with additional attacks on June 4) WOULD be a disappointment, but they seemed willing to run some serious risks in this battle. And if Nagumo still hasn't detected TF 16 and TF 17 by the end of June 4, why not move west during the night and roll the dice?

But it would definitely be fun to game out.

 

[173rd LRRP]

But I think that's the critical thing about Lexington: the USN learned some important damage control lessons from Lady Lex's sinking, as Parshall himself has pointed out. The first one was put into practice for Midway on board Yorktown: Oscar Myers' idea of purging the fuel lines with CO2 as soon as the strike packages were sent off. Had they not done that, Yorktown likely would have been lost in Hiryu's first attack, and then Hiryu's subsequent strike would have moved on and probably located TF 16.



Man, people are getting greedy here...

As I have noted on earlier threads, Dad as boiler officer on Yorktown at Coral Sea and Midway and it was his job to get the boilers operating again once most had been knocked out. He was in groups that was aboard to salvage when she was torpedoed by submarine.

5 carriers would have been possible if Saratoga was ready earlier. Filling BG 5 Air Wings with experienced crews might have been difficult: 350 air crews or so

 

[Athelstane]

As I have noted on earlier threads, Dad as boiler officer on Yorktown at Coral Sea and Midway and it was his job to get the boilers operating again once most had been knocked out. He was in groups that was aboard to salvage when she was torpedoed by submarine.

And your dad's job would have been met with success if I-168 hadn't shown up and gotten lucky. For all the beating Yorktown had taken, she was definitely salvageable at the end of the battle, and the salvage teams were making solid progress when the torpedoes hit. The Yorktowns were very tough ships.

Did your dad make it off the ship?

5 carriers would have been possible if Saratoga was ready earlier. Filling BG 5 Air Wings with experienced crews might have been difficult: 350 air crews or so

Another beloved counterfactual! Since Sara showed up at TF 17 on June 8, just a few days too late for the battle . . .

Nimitz had lit as much fire as he could under Saratoga's teams to get her underway, to the point of wiring Ramsay on May 30 to leave San Diego immediately even if Admiral Fitch hadn't shown up yet . . . I think what really made it hard to do was that Nimitz and King weren't convinced of the Midway operation intel until May 16, and two weeks just wasn't enough time to speed it up sufficiently. Had they had more warning, I think Saratoga could have been gotten out there sooner.

 

[mattep74]

The "good Hornet shows up to OTL Midway" alternative is one I've toyed with as a thought experiment. Like, say that Hornet's strike commander in the air has engine trouble and has to turn back early before Torpedo 8 broke off around 0825. Through whatever devolution of command happens, then, either Waldron is in command, or somebody willing to listen to Waldron is in command, and the whole air group (at that point the only American carrier airgroup that day in reasonably solid assembled mass!) turns onto the historic Torpedo 8 course, arriving over the Kido Butai as a collected mass around 0915-0930. The air group was a little less experienced, and there's plenty of room for mistakes like the two Enterprise bombing squadrons diving on the same carrier instead of distributing and underperformance of torpedo planes even with fighter escort and dive bombers making more of a mess. Still, it's possible that massive explosions from dive bomber attacks and such draw the attention of the parts of VF-6 that had ended up picking up on VT-8 historically but didn't follow them into the attack. Anyway, writing a giant furball at this point and rolling dice on hits left to the person who dares write the AH, but by the time the last stragglers exit the area or are shot down around 0940 to 0945, there's probably one or two Japanese carriers out of action and aflame. A towering smoke plume could help over-determine that Yorktown and Enterprise's squadrons know where to go. With a CAP degraded by engaging torpedo planes and dive bombers and two carriers unable to handle CAP refuel/rearm, there may even be a chance for VT-6 to torpedo one of the damaged carriers in the confusion when they arrive around 0940, and then it's just a question of if Hiryu or anyone else manages to accidentally hide in a squall line or something when three more American dive bomber squadrons arrive at 1020 (the historical hammer blow, but here just another hammer blow).

As Parshall puts it, it's a clean sweep by lunch and a lot of the drama then depends on what happens with afternoon strikes on cruisers, battleships, and changes in Nagumo, Fletcher and Spruance's strategy with the Japanese carriers decisively out of the fight by noon. Still, it could be dramatic to write for somebody who unlike me can actually write Thrilling Air/Naval Combat (tm).

Arent the US torpedoes still bad?

 

[Athelstane]

BTW, one interesting question that came up in the Parshall Q&A (42:30) was just what ships in Fletcher's fleet actually had any radar at the time. Parshall wasn't sure, beyond the carriers . . . so, digging around . . .

But most of the crusiers present had radar by that point, even if only of the primitive CXAM and SC-1 sort.

Task Force 17
USS Yorktown
USS Astoria (Fletcher's temporary flagship after first strike on Yorktown)

Task Force 16
USS Enterprise
USS Hornet
USS Pensacola
USS Northampton
USS New Orleans
USS Vincennes
USS Atlanta

The Japanese, of course, had no radar on any of the ships involved in Operation MI. Which only made it easier for Fletcher and Spruance to surprise him.

 

[naraic]

Arent the US torpedoes still bad?

Yes. Parshall makes the point that a US doctrine stated that a carrier strike was considered likely to take out a pair of carriers with dive bombings alone.

The carriers usually had a pair of dive bomber squadrons (18 planes each) and dive bombers should take out a carrier with three bombs. US doctrine claimed that one in six dive bombers would get hits.

Its also worth noting that the main problem with US torpedoes was that the air dropped torpedoes were slow. The mark 13 air dropped toredoes were 30 knot torpedoes.

If a carrier is damaged and it can't turn away the torpedoes were capable of blowing up.

Furthermore torpedo bombers showing up could distract the Japanese fighter screens which was already overwhelmed.

 

[marathag]

But most of the crusiers present had radar by that point, even if only of the primitive CXAM

In gunnery trials, New York and the prototype XAF, was able to detect the flight of outgoing 14" shells on their 'A' Scope. 100 miles air, 15 surface.
Not bad, considering it was somewhat low mounted, ontop the pilot house, displacing an optical rangefinder. Only 15kW output, at 200mHz, 'P' Band.
That's some good operators, and it allowed shot tracking. The improved CXAM-1 was most for reliability and lower cost to manufacture, not new features or more power.

 

[e of pi]

Arent the US torpedoes still bad?

Yes. Parshall makes the point that a US doctrine stated that a carrier strike was considered likely to take out a pair of carriers with dive bombings alone.

The carriers usually had a pair of dive bomber squadrons (18 planes each) and dive bombers should take out a carrier with three bombs. US doctrine claimed that one in six dive bombers would get hits.

Its also worth noting that the main problem with US torpedoes was that the air dropped torpedoes were slow. The mark 13 air dropped toredoes were 30 knot torpedoes.

If a carrier is damaged and it can't turn away the torpedoes were capable of blowing up.

Furthermore torpedo bombers showing up could distract the Japanese fighter screens which was already overwhelmed.

Like @naraic says, the problem with the aerial torpedos was less that they didn't work than that the planes were slow, the torpedos had to be dropped even lower and slower, and then once in the water the torpedos were slower than a Japanese carrier. With a hammer-and-anvil attack coming in two angles from the bow on a distracted carrier that may already have dive bomber damage or simply be distracted by other attacks, VT-6 might have had some luck. It'd definitely be luck, though and depend on the situation the Good Hornet leaves behind exiting the zone--call it an assist or a run batted in, not a home run.

Also as @naraic says, US doctrine figured one trained dive bomber squadron was enough to bag a carrier...which at Midway proved basically correct. While Midway's squadrons had minimal dive bombing experience and did not press their dives (which both makes a harder target and for better targeting), the Yorktown and Enterprise squadrons indeed proved three bombs was enough to leave a ship to brew up into a conflagration beyond all control even if it didn't flatten the carrier outright, and they scored someplace between one hit in four to one hit in six, so 18 planes was definitely enough to put 3 bombs into a carrier. You could be Hiryu, or Soryu, or Kaga, but you were still going to be rendered a total loss with 3 US bombs and Japanese damage control (or lack thereof). The likelihood of anyone pulling of a re-roll of Dick Best's "why'd you use only one bomb, well they only have one carrier, right?" gambit is more a matter of debate, but if one was enough by luck, three was enough by total overkill. The only question about whether Hornet's air group (having stumbled by luck as a group over the Kido Butai) will prove more competent to be there, like the other carriers, or less trained like Marines and unable to push their attack in successfully. Given Waldron' aggressiveness with VT-8 compared well to Enterprise's VT-6, I'm inclined to guess they'd have given a decent show, so one or two Japanese carriers would probably have Bombing 8 and Scouting 8's name on them here.