Realistic Results: what are these things, what aren't they?

[Tristanjohn]

Now, what triggers that 'throwing the switch'?

Good question. This could either follow the historic timeline strictly (my preference without thinking, but then that poses the problem of knowing beforehand exactly when this stuff happens, not so good) or a random factor might be interjected--but then again, how random?

If there's a better method I'm open to suggestion.

[Drongo]

I am probably in the minority, but I do not take issue with the original IJN pilots have high experience levels.

I wouldn't make the assumption that you're in the minority. Perhaps a poll?

Regardless, I have yet to see any historical work dealing with the period (in detail) that does not credit the starting IJN aviators as getting real benefits from the areas of intense training and pre-war combat experience.

Everything I've seen (Lundstrom included) accepts that starting premise and then goes on to show how the IJN air arm was defeated in spite of it.

For example, at Guadalcanal where the overall kill ratio was close to equal, Lundstrom prioritised the major factors influencing the air combat results in the following order:

1) Altitude advantage (normally to the US)
2) Early warning (coastwatchers)
3) USN/USMC air combat training (specifically, deflection shooting)
4) The ability of the Wildcat to take enormous punishment (and the corresponding weakness in the A6M's armament).

Given these factors, you would wonder what the kill ratio would be if the Japanese naval aviators weren't of the initial quality that they were.

Kill ratios on their own shouldn't be used to pass judgement on just how good the starting group of IJN aviators were (or weren't).

[Tristanjohn]

I wouldn't make the assumption that you're in the minority. Perhaps a poll?

Regardless, I have yet to see any historical work dealing with the period (in detail) that does not credit the starting IJN aviators as getting real benefits from the areas of intense training and pre-war combat experience.

I'm sure they did get benefits from their training. No one said Japanese pilots were inadequately trained to fly.

No doubt the combat experience they gained had an effect, how great is the question and was it all good?

Everything I've seen (Lundstrom included) accepts that starting premise and then goes on to show how the IJN air arm was defeated in spite of it.

You imply that because everything you've read ("in detail" you noted above, whatever that means) agrees on a particular point then that point must be valid. That is an illogical conclusion.

For example, at Guadalcanal where the overall kill ratio was close to equal, Lundstrom prioritised the major factors influencing the air combat results in the following order:

1) Altitude advantage (normally to the US)

We would expect Allied pilots to seek an altitude over the enemy's bombers before attacking. Similarly their fighters mostly outperformed "Zeroes" when diving from above as opposed to swirling admist them dogfight fashion.

Still, was the latter an "advantage" or merely an effort to make the best of what they had?

Finally, as this is not currently modelled, yet would need to be if we followed your lead, how would you suggest we plug this function into the game?

2) Early warning (coastwatchers)

This is true though as far as I can tell not modelled, so yet again we find Gary's simulation does not give the Allies their just due.

3) USN/USMC air combat training (specifically, deflection shooting)

Wait a minute. Your thesis (Lundstrom's thesis) is:[indent]A) Japanese pilots were better trained than American pilots and this is demonstrated by the fact that

B) USN/USMC pilots were . . . better trained?

[center]..tap..tap..tap..[/center][/indent]Can anyone point out the flaw in that argument?


[left]

4) The ability of the Wildcat to take enormous punishment (and the corresponding weakness in the A6M's armament).

[/left][left]So much for the "Japanese 'Zeroes' were better planes" school of thought. (It seems to me you people want to have your cake and eat it too.)[/left][left]

Given these factors, you would wonder what the kill ratio would be if the Japanese naval aviators weren't of the initial quality that they were.

[/left][left]Different, perhaps. Which way different is unknown. But that's academic in light of the kill ratio we have for these encounters. What I'd like to know is where the model gets off with its premise of Japanese air superiority given that kill ratio.[/left][left]

Kill ratios on their own shouldn't be used to pass judgement on just how good the starting group of IJN aviators were (or weren't).

[/left][left]Then what should be added to the mix and how would you model it?[/left]

[Tristanjohn]

So what we are going to need is something like the following:

Japan( Aircraft metrics + pilot training + pilot ability + ground crew training + tactics + airfield condition + weather condition + supply status + maybe other factors ) =

Allied( ( Aircraft metrics + pilot training + pilot ability + ground crew training + tactics + airfield condition + weather condition + radar existance + radar operators training and experience + coastwatchers + maybe other factors )

The = will only be for the beginning of the war. As it progresses the Allies side of the equation gets much larger.

That's not bad. I'm not sure I'd want it that complicated but then again why not as long as we don't confuse our model in the process. The overriding importance of "equals" cannot be emphasized too strongly, however, for the beginning period of the study (early 1942). And, as you say, after that it's just a matter of dotting one's i's and crossing one's t's and mopping up afterward with the Japs.

[BillBrown]

Good question. This could either follow the historic timeline strictly (my preference without thinking, but then that poses the problem of knowing beforehand exactly when this stuff happens, not so good) or a random factor might be interjected--but then again, how random?

If there's a better method I'm open to suggestion.

The problem is that these types of things happened because of events that happened. How do you program that? You will never get exactly the same sequence of events in any game, as soon as turn one is done the whole war takes a different path from reality.

So what do you use a triggers since the exact events will not occur and without them the incentive for change is either not there or changes its timeframe.

[BillBrown]

That's not bad. I'm not sure I'd want it that complicated but then again why not as long as we don't confuse our model in the process. The overriding importance of "equals" cannot be emphasized too strongly, however, for the beginning period of the study (early 1942). And, as you say, after that it's just a matter of dotting one's i's and crossing one's t's and mopping up afterward with the Japs.

The title of this thread is "Realistic Results: what are these things, what aren't they?" In order to get absolutly realistic results, you will have to have all of the real data. The list I started with is only the beginning, many of the items would have to be broken down further. Without having every piece of relavent data for the sim to use, you will only get close to realistic results. Of course everyone will have to buy a Cray to run their copy of UV Unlimited on.

[Tristanjohn]

The problem is that these types of things happened because of events that happened. How do you program that? You will never get exactly the same sequence of events in any game, as soon as turn one is done the whole war takes a different path from reality.

So what do you use a triggers since the exact events will not occur and without them the incentive for change is either not there or changes its timeframe.

The "exact events" will surely occur if the model has it right (and the model currently does not within my experience--a flip of the coin seems enough for a USN SC TF to beat the bejesus out of the IJN in a night action, in fact I've seen this when the USN TF had far fewer assets), those "exact events" being "bad day at Black Rock" for the good guys, as in "Our guys lost three cruisers and a tincan inside eight minutes!" and "You know, I just had an idea . . . what if we made a hose nozzle that shot a fine spray instead of a stream?" or "How about if we had a space below decks set aside where a man could think and communicate with other thinkers during battle?"

Necessity is the mother of invention and the USN was certainly in need and so invention followed shortly. Gary's model doesn't get that truth at all.

The model also doesn't get ship "experience" right. It makes the bad assumption that all ships (and their abstracted crews) learn individually and only from personal experience in combat, as if these naval entities existed in some war vacuum out in the Pacific. The model does not allow for the fact the USN was not a collection of individuals in virtual isolation but a collective national service

In real life the word gets around, in war it gets around fast. Gary's model can't hear.

On top of this there appears to be no model function at all to simulate growth of experience/leadership ratings for naval commanders. Assuming I have that right, what's going on there?

Let's take a green lieutenant in command of a PT. He's 22, say, and learning the ropes. So he runs his boat out into the heat of the night and returns a week later dripping wet, saved by natives after being run over by a Japanese destroyer. Don't you suppose this youngster has learned something from that experience? How about the other skippers he discusses this stuff with? Doesn't this shared knowledge mean something in the greater scheme of things?

Of course it means something. I'm not sure how best to model it, doubt Gary's even tried--we wouldn't know that unless Gary spoke or at least published his formulas. Food for thought, though.

I don't pretend to have all the answers, or any answers when it boils down. I do have plenty of questions and I'd have less of those if Gary's model did better work.

[Tristanjohn]

The title of this thread is "Realistic Results: what are these things, what aren't they?" In order to get absolutly realistic results, you will have to have all of the real data. The list I started with is only the beginning, many of the items would have to be broken down further. Without having every piece of relavent data for the sim to use, you will only get close to realistic results. Of course everyone will have to buy a Cray to run their copy of UV Unlimited on.

Let's not get anal on ourselves. No abstract model can simulate war down to the blood, sweat and the tears. What we want is a model that feels right and gives acceptable results along the way. Gary's model just doesn't feel right in some important areas and its results often are, to quote a patient man . . .

[center]. . . off the "planet Xenon."[/center]

[Drongo]

I'm sure they did get benefits from their training. No one said Japanese pilots were inadequately trained to fly.

Where was I addressing the issue of whether or not they could adequately fly their aircraft?

You imply that because everything you've read ("in detail" you noted above, whatever that means) agrees on a particular point then that point must be valid. That is an illogical conclusion.

I didn't imply any such thing nor do I make any such conclusion.

It was simply an observation that those accounts that I've read are all firmly of the view that the Japanese pre-war trained IJN aviators were to be considered of very high quality in combat terms.

If I were to make a conclusion it would be that since I find no fault with the arguements that back their opinions, I therefore conclude that they have a valid point until disproven.

The implication was that there exists those who are well researched on the subject and that they were of a particular point of view which could be taken as supporting the concept that the IJN aviators in question would be considered of high experience.

"In detail" by the way means in detail, as in a detailed analysis of what happened and why, ie who shot who down and what were the circumstances.

Wait a minute. Your thesis (Lundstrom's thesis) is:[indent]A) Japanese pilots were better trained than American pilots and this is demonstrated by the fact that

B) USN/USMC pilots were . . . better trained?

[center]..tap..tap..tap..[/center][/indent]Can anyone point out the flaw in that argument?

Who said Japanese pilots were better trained than American???

If, on the other hand, Lundstrom makes the statement that:
A) The early war IJN aviators were to be considered of very high quality due to intense pre-war training and previous exposure to combat conditions

and

B) The USN/USMC emphasis on deflection shooting as part of their air combat training gave them a capability in combat not generally shared by the IJN pilots.

and the flaw is?
[/QUOTE]

[left]So much for the "Japanese 'Zeroes' were better planes" school of thought. (It seems to me you people want to have your cake and eat it too.)[/left][left]

You people???

[mdiehl]

mdiehl's point is accurate regarding the maneuvarability of the Zero at high speeds; however at dog-fighting speeds it had superior maneuvarbility. The plane was built to be a dog-fighter, since this was the perfered mode of engagement in the air war over China by the Japanese and was reflected in their designs, the Zero being the result of their obsession with ultra-maneuvarability.

The fundamental problem with your post is that there is no such thing as a single pervasive generally-everyone-fought-at "dog fighting speed." All speeds are suitable for dogfighting unless you're at speeds that you can only attain in a dive. The Zeke's optimum "dogfighting" speed was substantially lower than the P39s and P40s because the Zeke did not lose eneegy as fast as these other planes. Therefore, at slow speeds the Zeke could maneuver and maintain airspeed, whereas the P40/39 would lose airspeed faster when maneuvering. That is why the general admonition among US pilots was not "don't dogfight with a Zeke" (as some mistakenly assume) but rather, "never let your airspeed drop below 300 mph if you find yourself losing airspeed dive out, disengage and regain altitude and speed before attempting to re-engage."

I think the maneuvarbility statistic is really an abstraction of dog-fighting ability of the air craft.

This is a fundamental problem with the A/C combat model and illustrates TristanJohn's point about having a clear idea of what you are modeling before you abstract it down to its simplest components.

There is no such thing as "dogfighting ability." There are only roll rate (at which all single engined US a/c were faster than the Zeke and Oscar), time to turn through a given direction change (the US had the advantage above 300 mph the IJN generally at lower than 300 mph), climb rate (generally the advantage was with the Japanese until the high HP late war US a/c), dive rate (advantage universally with the Allies until the Tony came along ... it could about keep up in a dive with early P40s [through E series]), armament (advantage generally with the Allies although the Japanese a-a cannon was decent if the pilot could get close), and armor/self-sealing tanks (advantage Allies universally, all single-engined and most two-engined a.c. at all times throughout the war).

Thus, treating MVR as a proxy for something that really never existed in a simplistic form ("dogfighting ability") and then assuming that the Japanese a/c were better at it is frankly just plain wrong. Handling it in this matter is almost guaranteed to create a poor model.

If I recall correctly from the materials that I have read the perferred mode of attack for an Allied Pilot (P-40s) was to make a diving attack.

Not exactly. P40s and P39s were best at high speed attacks. These included diving attacks and level attacks at high speed. Both were considerably faster (at their optimum altitudes) than any model of the A6M or Oscar. A P39 at 6000 feet had a 50 mph max airspeed advantage over a fire-walled Zeke.

[mdiehl]

The bulk of the Japanese pilots lost at Midway were lost without being able to prove one way or the other their skill versus the USN in that battle. We are forced to judge them by what they had done prior to this.
If you compile the data you will find they outfought a RAAF unit at Darwin (A unit that had fought in BOB) Had inflicted considerable damage on enemy every where they went. While the IJA was having it's troubles up at Rangoon the IJN LBA units destroyed nearly 400 allied aircraft in Malaya. They then destroyed another 300or so in Java area.
On top of the PI airunits.
The A.V.G. maintains that they out performed all other allied units untill the Navy and USAAF adopted their tactic (under other names)

Almost all of that is at best a horrific oversimplification and at worst outright incorrect. The Japanese in the first half of 1942 rarely outfought a unit that had first line air craft in air to air combat. They shot up lots of stuff on the ground (PI), lots of Brewsters, Fokkers, and a handful of hurricrates in the Singapore area. In Indonesia, it is entirely possible that Dutch pilots in P40s shot down more Zekes than they lost in air-to-air combat. But the a/c did not last long ANYHOW because many of them were used in ground attack... too many jobs, not enough a/c.

The only places where I can find lopsided Japanese victories over P40 equipped units or USMC/USN pilots are single instances where the Japanese had overwhelming air superiority in numbers (Midway vs. VMF there) or by pure luck (18 August at Cactus, PM in March, Darwin raid and one instance in Indonesia) where the Japanese arrived suddenly and without warning. That tells me that any sustained advantage that the Japanese enjoyed in early 1942 had nothing to do with prior experience, training or aircraft. Instead, the primary factor was the quality (in many cases, absence) of Allied early-warning systems and the general advantage that goes with the person who has the initiative and is launching the attack (concentration and timing of effort).

[Tristanjohn]

Where was I addressing the issue of whether or not they could adequately fly their aircraft?


I didn't imply any such thing nor do I make any such conclusion.

It was simply an observation that those accounts that I've read are all firmly of the view that the Japanese pre-war trained IJN aviators were to be considered of very high quality in combat terms.

If I were to make a conclusion it would be that since I find no fault with the arguements that back their opinions, I therefore conclude that they have a valid point until disproven.

Until disproven, eh? You'll just take someone's word for something (if it suits your fancy) without any effort on your part to think about it independently, yet I'm supposed to believe that reason will touch your soul?

Why do I doubt this?

The implication was that there exists those who are well researched on the subject and that they were of a particular point of view which could be taken as supporting the concept that the IJN aviators in question would be considered of high experience.

"In detail" by the way means in detail, as in a detailed analysis of what happened and why, ie who shot who down and what were the circumstances.

Who said Japanese pilots were better trained than American???

If, on the other hand, Lundstrom makes the statement that:
A) The early war IJN aviators were to be considered of very high quality due to intense pre-war training and previous exposure to combat conditions

and

B) The USN/USMC emphasis on deflection shooting as part of their air combat training gave them a capability in combat not generally shared by the IJN pilots.

and the flaw is?

You people???

The flaw is the game doesn't try to model American strengths across the board as far as I can see, instead assumes an overall weakness, whereas it does attempt to model Japanese strengths (with a catchall quality called "Maneuver") while pretty much ignoring their weaknesses. That's what's known as either design bias and/or confusion, and it's found in every bad model around.

Let's cut through this rhetoric. Please tell me, in so many words, what you believe with regard to how the game models Japanese superiority in the air to begin the game, then I'll take a stab at dealing with that.

Sound fair?

[Snigbert]

"The flaw is the game doesn't try to model American strengths across the board as far as I can see, instead assumes an overall weakness, whereas it does attempt to model Japanese strengths (with a catchall quality called "Maneuver") while pretty much ignoring their weaknesses. That's what's known as either design bias and/or confusion, and it's found in every bad model around."

So if you include a characteristic of aircraft which, in this case, happens to favor the Japanese, you are showing bias...Well, from what I have read the Japanese Zero did have better turning at lower speeds than equivalent American aircraft...while American aircraft had advantages which made them more durable and better firepower (two characteristics of aircraft also included, which favor American aircraft). Why do you only point out the areas where it happens to favor Japan and ignore where it favors the US?

What weaknesses do you see the game assuming for Americans?

"Necessity is the mother of invention and the USN was certainly in need and so invention followed shortly. Gary's model doesn't get that truth at all."

How do you suggest modelling such an attribute as inventiveness?

"The model also doesn't get ship "experience" right. It makes the bad assumption that all ships (and their abstracted crews) learn individually and only from personal experience in combat, as if these naval entities existed in some war vacuum out in the Pacific. The model does not allow for the fact the USN was not a collection of individuals in virtual isolation but a collective national service"

That's another incorrect assumption on your part. Ships in TFs have the possibility to receive experience just for being at sea at the beginning of each turn. Shakedown cruises are a good idea. Unit cohesiveness does exist and is reflected, just not abundantly obvious to the player.

"In real life the word gets around, in war it gets around fast. Gary's model can't hear."

Inventiveness...hearing...if M.I.T. can't program a computer to have true human characteristics, I can't expect GG to be able to. Are these serious complaints or are you being facetious?

"Let's take a green lieutenant in command of a PT. He's 22, say, and learning the ropes. So he runs his boat out into the heat of the night and returns a week later dripping wet, saved by natives after being run over by a Japanese destroyer. Don't you suppose this youngster has learned something from that experience? How about the other skippers he discusses this stuff with? Doesn't this shared knowledge mean something in the greater scheme of things?

Of course it means something. I'm not sure how best to model it, doubt Gary's even tried--we wouldn't know that unless Gary spoke or at least published his formulas. Food for thought, though."

Have you looked at TF commanders ratings at the beginning and end of the game? I agree this should be modelled, I'm not sure how it's done now, if at all.

All I can tell you is that there are a LOT of computations taking place behind the scenes that the player isn't aware of. A good example is in ground combat, where it seems to go through a quick series of bombardments and assaults and then you get results. However behind the scenes many many quality/morale/supply/leadership/etc checks are being made for each unit. There was a design decision made at some point on how much information needs to be displayed and how long turn processing should take, where to cut down on time and all. I've been seeing you repeatedly mistake the absence of information being presented to the player as a lack of system being in place. Just because you dont like the model, doesn't mean it doesn't exist.

[Drongo]

Until disproven, eh? You'll just take someone's word for something (if it suits your fancy) without any effort on your part to think about it independently, yet I'm supposed to believe that reason will touch your soul?

Why do I doubt this?

How does "I find no fault with the arguements" come to mean that no effort was made to find fault with the arguements???

Why do you doubt this? Because you haven't a bloody clue.

Let's cut through this rhetoric. Please tell me, in so many words, what you believe with regard to how the game models Japanese superiority in the air to begin the game, then I'll take a stab at dealing with that.

Sound fair?

You're joking, right?

In case you're not, I'll lay down the rules of engagement:

1) Stop these mindless attempts at re-interpreting people's posts in an effort to score points in some imaginary game you play.

2) Stop the childish baiting that you feel obligated to incorporate into almost every post you make.

3) Stop being a boorish twat with delusions of granduer.

4) Be prepared to at least understand WTF happened (in an historical event or a game) before you go waving your conclusions around in other people's faces.

Sound fair?
Go on, give it a go.........you could be our next Mogami.

Till then, bye.

[Mike Scholl]

How does "I find no fault with the arguements" come to mean that no effort was made to find fault with the arguements???

Sound fair?
Go on, give it a go.........you could be our next Mogami.

Till then, bye.

I think you're being unfairly hard on MOGAMI. He rarely resorts to name-
calling unless really provoked---and he's generally willing to try and back
up his views with game examples. Plus he's very generous in sharing
information about the beta test games ow WITP. I find him obstinate
and irritating at times too, but usually at least restrained in his replies.

TRISTANJOHN, on the other hand, has strong opinions and is want to
try and beat down anyone who dissagrees with insults as well as
arguement. And while he is vociferous and even sulpherous in his
castigations of Gary's game model---I have yet to see him offer up
a workable alternative. I want to see what he proposes as an system
model. If he has a "better mousetrap", he should offer it up. A number
of his points have some validity, but so do many made against him. It's
time for the screaming to stop, and some workable suggestions to be
offered.

[Drongo]

I think you're being unfairly hard on MOGAMI. He rarely resorts to name-calling unless really provoked---and he's generally willing to try and backup his views with game examples. Plus he's very generous in sharing information about the beta test games ow WITP. I find him obstinate and irritating at times too, but usually at least restrained in his replies.

Maybe a bit of strine caused the confusion.

"Go on, give it a go.........you could be our next Mogami" means if TJ refrained from certain habits, he could one day grow up to be as respected as Mogami.

[frank1970]

Maybe a bit of strine caused the confusion.

"Go on, give it a go.........you could be our next Mogami" means if TJ refrained from certain habits, he could one day grow up to be as respected as Mogami.

Ähmmmm, no.

[frank1970]

What I like of TJs posts is his unability to see his own flaws. Example:

American pilots were as good as Japanese.
American planes were better than Japanese.
> The war ended after two weeks because all our superheroes in superplanes had cleared the sky of all Japanese planes.

And that is exactly what NOT happened. The USA got a draw in the early war, so at least one of the two assumptions of TJ must be wrong.

I am all for changing the air battle calculations, if there is a founded reason to do so. Until now I have heard two persons claiming the modell is absolutely wrong. One is Mdiehl, whom I respect because he is a) very busy in bringing data to support his claims ( although he sometimes exagerates IMHO) and b) he really tries to improve the modell and the second is TJ, who aparently has some feud with G. Grigsby about whatever and therefore has a "It (HE) is wrong because I know better" stance.
TJ, why don´t you write a mail to G. Grigsby and discuss all the ugly problems you had with him and then come back and do some constructive critics on the game?

[mdiehl]

I've been trying to resolve in a flowchart kind of way this disconnect between MVR ratings that give the Zeke the edge and the real world. I shake my head when I read, for example:

So if you include a characteristic of aircraft which, in this case, happens to favor the Japanese, you are showing bias...Well, from what I have read the Japanese Zero did have better turning at lower speeds than equivalent American aircraft...

OK. So the MVR rating is an "all-speed" rating. Why then give the Japanese a/c better MVR ratings if they were less maneuverable at high speeds?

To reconcile this, I will try to think of MVR not as the literal "maneuverability" rating of the a/c. For one thing, Erik Shilling made the argument that roll rate was the most important thing in the maneuverability of an a/c and every US single-engined fighter-interceptor could out roll the Zeke. Instead, I will think of MVR as something like a derivative in a complicated differential equation involving airpseed, weight, roll rate, turn rate, and level flight acceleration. In a sustained turning-engagement type combat, such an equation would eventually favor the Zeke. It loses energy less quickly.

Now, I know that UV has multiple air combat rounds. I also suspect that there are situational modifiers. Someone gets positional advantage ... "the bounce" .. at times. Assuming for a moment that there is no bounce, one can expect all Wallied pilots to accelerate and climb upon knowing of an enemy in their area (pre-engagement positioning, that is). IN that event, the P40s and what not will be flying at their optimum speeds. In the first combat round, the Allied planes should have a short term MVR bonus that makes them slightly BETTER dogfighters than the Zeke. (About 1 point for example for a P39 and 2 points for a P40).

Following this round, one could have something like an "Experience Check." If by random determination a pair of pilots (it should be pairs, that is how US pilots operated) pass the Experience Check, in the second combat round they continue to receive the bonus. It is assumed that they blew through the enemy formation and re-engaged at high speed.

On failing an Experience Check, the bonus goes away, and the Zeke because of its higher basic MVR rating (I'm treating this as an established fact of the game whether or not one thinks that the Zeke merits it) is higher. The P40/39/F4F/whatever has lost sufficient energy (because of its persistence in following through in this turning engagement) that the Zeke has the advantage. With the F4F, this mechanismm reflects the tendency for Wildcat pilots to fly intensive mutual support. A failure of the EXP check means that the beam defense and its predecessors was not prioperly executed.

This model will adequately work for the whole war because. Better US aircraft like the F6F F4U and the advanced USAAF will have better MVR ratings owing to their MUCH greater speed. I have assumed that the MVR rating reflects a consideration of the maximum airspeed attainable by an a/c in addition to it's characteristics in re pitch, yaw, and roll.

It should also make some of the folks who think "experience ought to count for something" happy. It will give the early-war Japanese pilots an edge because early war Allied pilots (army pilots in particular) will be less likely to pass the EXP check than later Allied pilots (who may have learned via the "information virus" or via training not to lose energy in a-a combat).

[mdiehl]

BTW, that all presumes that anyone thinks the model is really broken. Mogami seems consistently to get something approaching 1:1 losses in the early war. It makes me wonder whether a fix is necessarily needed. The biggest problem I've seen lately is the huge masses of Japanese a/c operating from New Guinea and New Ireland. Strikes escorted by fifty Zekes should be an extreme rarity. If they're not, UVs problem may not be the a/c model at all, but rather a flawed supply and logistics model that lets the Japanese keep that many a/c flying at the end of a very narrow and overstretched logistical pipe.