Experience vs Machine Performance

Gary Grigsby's strategic level wargame covering the entire War in the Pacific from 1941 to 1945 or beyond.

Moderators: Joel Billings, wdolson, Don Bowen, mogami

HMSWarspite
Posts: 1405
Joined: Fri Apr 12, 2002 10:38 pm
Location: Bristol, UK

Post by HMSWarspite »

mdiehl wrote:I do not agree. Here's how I see it:

Corsair (20) vs A6M3 (80): a couple of F4Us shot down, high operational losses for the F4Us, low operational losses for the A6M2s, maybe a Corsair gets lucky and takes down a Zeke.
I did not cover operational losses, so will ignore those comments (although it is pretty easy, op losses is inversely proportional - or something - to exp). I think the fundamental issue is whether markedly inferior a/c should shoot down better ones when the exp is skewed in their favour. I flatly disagree with the 'couple of F4U shot down' version. This is of course well outside RL, since no US 20 exp group got anywhere near a Jn unit, but 20 exp pilots struggle to do ANYTHING but fly the plane. 80 exp are aces, and know how to shoot aircraft down, therefore there must be a good chance that they bounce the F4U's and shoot some down. The only exception would be guns vs durabilty, in which case they might damage lots
Corsair (20) vs A6M3 (20): A few Zekes shot down. No F4Us shot down unless a Zeke gets very lucky. High operational losses all around.
Same problem - the F4U pilots struggle to use their a/c - loses should be random (first person to get behind the enemy unseen - no SA remember - gets the kill, but very few do.
Corsair (50) Zero (50): Zeroes take a pasting, losing three for every Corsair brought down. Operational losses "normal" if "50 EXP" is assumed to be "normal training."
You give the 50 exp pilot too much skill - I agree losses in US favour, but not 3:1 (unless IJN really goofs up). 50 exp should be still underachieving.
Corsair (50) Zero (70): Slightly more Zeroes shot down than Corsairs. Operational losses for the Corsairs normal. Slightly lower op losses for the Zekes.
We more or less agree here, but I put overall losses low (usually)
Corsair (70) Zero (70): Zeroes losing big time. Five or six to one destroyed for each F4U brought down. Operational losses low.
Now this I would put at 3:1 US favour (but I do not have records of engagements of these a/c. a 70 exp Zero group is probably MUCH better than anything that met F4U in history!
Corsair (80) Zero (30): Zeroes horridly trounced in a bloody 1-sided, no-contest engagement, losing dozens of A6M2s for each F4U brought down. Post combat, most of the remaining damaged Zekes are lost operationally. Japanese ground crews weep as strikes of hundreds of aircraft see fewer than a score returning to their base(s). Air combats see American pilots making ace in a single combat, sometimes seeing interceptions where modest numbers of Corsairs destroy several times their number in enemy fighters, break through all escorts, shoot down most of the bombers, and for the most part return safely to their bases. At night, on Japanese airfields, the ice-weasels come and finish off whatever's left.
I more or less agree hear, but dozens of zero's shot down presumes a reasonably big, VERY one sided fight. I would say losses >5:1 sure enough. 'Dozens' (plural) is OTT.
I have a cunning plan, My Lord
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

fighter combat

Post by mogami »

Hi, While there is the picture of fighters eyeing one another and moving to gain some advantage before attacking many Air to Air between fighters began with
one side shooting down enemy aircraft that had been up to that moment unaware they were about to be in combat. The low experience pilots are the ones most likely to be shot down before they are aware the enemy is at hand.
The aircraft do not matter in this case except for how much damage they take to knock out. Really good pilots know that the best target to aim for in a surprise attack is the enemy cockpit. All WW2 fighters had enough firepower to down any other fighter if they could get a good shot into the cockpit.

(Jeez don't you guys watch WW2 movies? The pilot always gets a face full of lead)

I don't think 20s' versus 80's have much of a chance no matter what they are flying.

What was the Turkey Shoot. 240-350 Japanese aircraft shot down for 29 USN fighters lost. (How they lose 29?) Must have been a few non 20 Japanese pilots in the batch.
Image




I'm not retreating, I'm attacking in a different direction!
mdiehl
Posts: 3969
Joined: Sat Oct 21, 2000 8:00 am

Post by mdiehl »

Let me explain where I'm coming from in re the extreme results. Caveat to all this being that I am not sure wtf "20 EXP" or for that matter "70 EXP" equates with in the real world. Some kind of ranking to be sure, but not much more meaningful to me than "bad" vs "good."

In the first extreme, low EXP F4Us vs high EXP A6Ms you pit pilots who presumably know how to find the throttle on their a/c against some kind of veteran. In those instances the situational awareness of the Corsair pilots is going to be pretty much in the real world the instant that bullets start striking their plane. More or less. How this would come to be is open to question. In a head on approach the Corsairs will go into the combat with pretty good situational awareness but, presumably, lose it to a degree in the thick of melee. Now, if these Corsairs make one high speed pass and are unable to stay in the fight they're going to shoot down a certain number of Zekes, lose a few Corsairs and exit stage left.

If they stay in the fight they're being attacked. In that event the attacker (unless the attacker is also being attacked, as in a head-to-head run) will have positional advantage. At this point the F4U pilot can:

1. Die, if the A6M pilot inflicts enough damage on the F4U fast enough that the F4U pilot has absolutely no time to react.
2. React.

If he reacts, the most likely reaction (because he's been told that his aircraft is fast enough to get him out of trouble) is to accelerate. If he does this and his airplane has not sustained much damage his aircraft will sprint, not walk, away from the attacking A6M and he will disengage. His alternative is to try maneuver, but he's at least been told that this is a bad idea.

If the really green F4U pilot lives long enough to react he can control the fight because of his aircraft's superior speed. One benefit that is immediately obtained is that he has the option to run away. Absent good luck, there is nothing that any A6M pilot, regardless of his experience, can do to stop that F4U from escaping, one the F4U pilot has chosen that course of action.

So to get lots of F4Us shot down, you have to presume that the F4U pilots lack any desire to survive and any knowledge of how to run away in a straight-line dive. Is that the level of training that 20 EXP represents?

On the other extreme you have 20 EXP A6Ms against veteran or well trained F4U pilots. The Zeke is flimsy. Most of the time if the Zeke lacks situational awareness with an F4U in an advantaged position on him and shooting, the Zeke pilot WILL die. The A6M is not likely to last long enough to provide any reaction time at all.

Assuming, however that the A6M pilot has the time to react. He can:

1. Attempt to disengage. Effort fails because he can't leave the scene of the combat unless all the F4Us are so busy killing some other Zeke that he slips away unnoticed. The F4U has complete control over the fight whether the A6M pilot wills it or not. The Zeke can't outrun. Given the level of training he probably can't maneuver well.

2. He can attempt to stay in the fight. This is substantively the same as option number 1 in the end because the F4U controls the fight. If he's lucky, a target will cross his gunsights in a way that allows him to maybe get a lucky kill, or simply ram his plane into the enemy.

Either way it's almost certainly a dead A6M pilot.

This by the way is well-represented by the Marianas Turkey shoot, in which upwards of 500 Japanese aircraft were destroyed for less than a score of American aircraft destroyed as a result of damage inflicted by the enemy. Dozens to one is a very believable outcome.

The bottom line is thath the faster aircraft can always control the fight. The rookie F4U pilot has therefore at least a good chance to escape. The A6M2 pilot, in contrast, almost never does.
Show me a fellow who rejects statistical analysis a priori and I'll show you a fellow who has no knowledge of statistics.

Didn't we have this conversation already?
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

Numbers

Post by mogami »

Hi, What were the actual Turkey Shoot numbers. I have seen a low of 240 Japanese and a High of 350 Japanese. Both sources gave the same USN loss 29
The Japanese began with 550 AC. The USN had 900 ac
Image




I'm not retreating, I'm attacking in a different direction!
mdiehl
Posts: 3969
Joined: Sat Oct 21, 2000 8:00 am

Post by mdiehl »

Hi, Mo -

The numbers that I've heard "given" on this are all over the map. 29 lost for the USN is the most consistent one and I suppose that somewhere there is a set of primary sources that support the number, but I've never seen anyone cite the primary source. The 250 to "more than 300" values generally refer to "confirmed kills" -- a way of tracking that I find very suspect. More than 500 Japanese a/c failed to return to base. So the minimum kill ratio for the US is possibly lower than 10:1 and the maximum is "pie-in-the-sky." Where "reality" falls on that spectrum is not entirely clear to me, but 10:1 seems like the minimum given the descriptions of the combat. This was, mind your, primarily F6Fs vs Japanese not F4Us, where the extreme disparities between the planes would have been even greater.

Caveat: Many of these were bombers so the salience is open to question. Also many of the bombers were shot down by flak. But I think if all the Japanese a/c had been fighters, the difference would not be that great.
Show me a fellow who rejects statistical analysis a priori and I'll show you a fellow who has no knowledge of statistics.

Didn't we have this conversation already?
HMSWarspite
Posts: 1405
Joined: Fri Apr 12, 2002 10:38 pm
Location: Bristol, UK

Post by HMSWarspite »

mdiehl wrote:Let me explain where I'm coming from in re the extreme results. Caveat to all this being that I am not sure wtf "20 EXP" or for that matter "70 EXP" equates with in the real world. Some kind of ranking to be sure, but not much more meaningful to me than "bad" vs "good."

In the first extreme, low EXP F4Us vs high EXP A6Ms you pit pilots who presumably know how to find the throttle on their a/c against some kind of veteran. In those instances the situational awareness of the Corsair pilots is going to be pretty much in the real world the instant that bullets start striking their plane. More or less. How this would come to be is open to question. In a head on approach the Corsairs will go into the combat with pretty good situational awareness but, presumably, lose it to a degree in the thick of melee. Now, if these Corsairs make one high speed pass and are unable to stay in the fight they're going to shoot down a certain number of Zekes, lose a few Corsairs and exit stage left.

If they stay in the fight they're being attacked. In that event the attacker (unless the attacker is also being attacked, as in a head-to-head run) will have positional advantage. At this point the F4U pilot can:

1. Die, if the A6M pilot inflicts enough damage on the F4U fast enough that the F4U pilot has absolutely no time to react.
2. React.

If he reacts, the most likely reaction (because he's been told that his aircraft is fast enough to get him out of trouble) is to accelerate. If he does this and his airplane has not sustained much damage his aircraft will sprint, not walk, away from the attacking A6M and he will disengage. His alternative is to try maneuver, but he's at least been told that this is a bad idea.

If the really green F4U pilot lives long enough to react he can control the fight because of his aircraft's superior speed. One benefit that is immediately obtained is that he has the option to run away. Absent good luck, there is nothing that any A6M pilot, regardless of his experience, can do to stop that F4U from escaping, one the F4U pilot has chosen that course of action.
You presume that a 20 exp pilot has the training to know what to do, the presence of mind not to panic, and fail to attempt it, and the skill to actually achieve what he sets out to do. My view of a 20 exp is that he might know what to do, but is highly unlikely to manage it, either because of panic, or his skills let him down. This is where my question 'what is 20 exp' comes in earlier
So to get lots of F4Us shot down, you have to presume that the F4U pilots lack any desire to survive and any knowledge of how to run away in a straight-line dive. Is that the level of training that 20 EXP represents?

On the other extreme you have 20 EXP A6Ms against veteran or well trained F4U pilots. The Zeke is flimsy. Most of the time if the Zeke lacks situational awareness with an F4U in an advantaged position on him and shooting, the Zeke pilot WILL die. The A6M is not likely to last long enough to provide any reaction time at all.

Assuming, however that the A6M pilot has the time to react. He can:

1. Attempt to disengage. Effort fails because he can't leave the scene of the combat unless all the F4Us are so busy killing some other Zeke that he slips away unnoticed. The F4U has complete control over the fight whether the A6M pilot wills it or not. The Zeke can't outrun. Given the level of training he probably can't maneuver well.

2. He can attempt to stay in the fight. This is substantively the same as option number 1 in the end because the F4U controls the fight. If he's lucky, a target will cross his gunsights in a way that allows him to maybe get a lucky kill, or simply ram his plane into the enemy.

Either way it's almost certainly a dead A6M pilot.

This by the way is well-represented by the Marianas Turkey shoot, in which upwards of 500 Japanese aircraft were destroyed for less than a score of American aircraft destroyed as a result of damage inflicted by the enemy. Dozens to one is a very believable outcome.

The bottom line is thath the faster aircraft can always control the fight. The rookie F4U pilot has therefore at least a good chance to escape. The A6M2 pilot, in contrast, almost never does.
All I was saying on the with the 80 exp F4U vs green Zeros is that you can only shoot down so many a/c in a mission. The best SA in the world doesn't keep you in touch with the position of multiple scattered enemy. The Marianas Turkey shoot was an extreme fight, (or more accurately series of fights) and the Fighter -fighter losses are not certain. Bet they weren't dozens (=24:1 minimum in my book!) I am happy to believe 10:1 order, but again what was the average exp of IJN in that battle? 50? 40? 30? I would guess the game would give 40ish (Mogami, any clue)


The point is we are mostly agreed on the effect of low exp, we are just discussing the price!
:)
I have a cunning plan, My Lord
Mike Scholl
Posts: 6187
Joined: Wed Jan 01, 2003 1:17 am
Location: Kansas City, MO

From MEMORY Alone..

Post by Mike Scholl »

Mogami wrote:Hi, What were the actual Turkey Shoot numbers. I have seen a low of 240 Japanese and a High of 350 Japanese. Both sources gave the same USN loss 29
The Japanese began with 550 AC. The USN had 900 ac
As I recall, the Japanese lost 350-360 Aircraft to the fighters and flak, with
another 100-odd straffed or otherwise destroyed on the ground in the Marianas.
Ozawa was trying to use a "transfer strike" with the A/C from his carriers
supposed to attack the Americans and then land on Saipan, to reverse the
process the following day. This is one reason why he didn't realize the extent
of the disaster for some time. US losses were about 30 A/C not counting
those lost at the end of the fight when Spruance launched his long ranged
attack on Ozawa'f fleeing force and a number of US aircraft had to ditch. Can't
argue with the 29 figure being quoted. The US had every advantage possible
in this fight except range. Their radar coverage allowed them to vector inter-
ceptors out to meet the Japanese long before they could reach the fleet, and
put themselves at the right altitudes to get the jump on them almost every time.
US flak was getting pretty awesome by this time and might be credited with
a fair portion of the kills. But if they got 25%, that still gives a 9:1 ratio in the
air. Now most of the Japanese weren't fighters, and the American's had "the
bounce" in most instances. At best the Japanese were "partially trained" and
might have had a squadron average experiance of 40---At worst the Americans
were fully trained, and had a goodly "leveling" of combat veterans for at least
a 70. Factoring out their other advantages (the bounce, fighting mostly heavily
loaded attack planes, being on the defensive) you still have to give the US AT
LEAST a 4 or 5 to one edge in results just based on A/C stats and experiance.
And the Japanese were NEVER AGAIN going to be even this good, while the
Americans were always going to be, even with rotation. No wonder Kamikazes
were starting to look like a good idea.
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

Japanese

Post by mogami »

Hi, You know in WITP I can't ever see the Japanese doing such things.
(I won't) In UV I withdraw airgroups back to Truk and cease all offensive operations when the ratios goes slightly south. In WITP I will not send airgroups below 60 out of Home Islands. When combat resuces a group to this level it will
return to Japan to train. You can always have the newest Japanese fighter or bomber in every group but you can make sure the groups are somewhat trained.
Image




I'm not retreating, I'm attacking in a different direction!
Alden
Posts: 4
Joined: Tue Oct 07, 2003 8:56 pm

Post by Alden »

I apologize for intruding, but...

If it results in a series of actions over say, one month, that greatly favor the Japanese in the early war, it's gonna be a broken model...

Results of combat are determined by so many factors other than the basic coefficients (Av. Exp, Mvr, Spd, what-have-you) of the units involved. What if, through superior deployments, making better tactical choices, acheiving better force concentration, etc, the Japanese player can create these favorable (to him) results *despite* any qualitative disadvantage he may have?

If the Japanese player can do this consistantly, he should be able to kill those shiny, hyper-trained American pilots as fast as the schools can produce them. A truly broken model would be one where Allied superiority ratchets up, inexorably, at some fixed rate, regardless of events in the game. Because that's not a model.

I work in modeling, of a more humdrum nature - I write computer models for gravity-powered hydraulic systems (i.e. sewers ;)) and I see a similar impulse in my users all the time - they know what the flow rates are at a particular manhole because they've taken measurements. Then they want to force the model to exhibit these flows at that location. But they are just feeding results in as input - which will give you "accurate" graphs and charts but teaches you nothing about your system, the real one in the ground, that you didn't already know. The nature of the system upstream of the calibration point is still mysterious. This is useless because it doesn't allow them to then try out "What Ifs".

This is just like in the game. You can't say "by the end of 1944, the USN should be shooting Japanese planes down at such and such a rate" - because this condition is dependant on upstream events. The conditions at the end of 1944 are the result of events a year, two years before. This is a point Mogami has made a few times in the threads on these topics. Its been generally ignored, though. Maybe because its self-evident (in which case I needn't have bothered writing this post) but I think its a really important one on a theoretical level.

The outcome of 100 Zeros v 100 Wildcats both at 0 Fatigue and 80 Av Exp, both at 10,000 feet on CAP missions might be interesting academically, and might be a useful baseline (if people can agree on what the outcome "should" be), this "series of events over one month" is meaningless because it can occur only in a broader context. Surely the esteemed mdeihl isn't arguing that no matter how well he plays or how foolishly the US player plays, from 1/1/42 to 1/1/43 the Japanese player should always suffer 1.0-1.2 (or whatever) Air to Air losses? And from 1/1/43 to some other date the ratio should be something else and so on. Again, that's not a model - its results from the one time the "model" ran during the actual interval of 1942-1945.

I realize that my point risks lapsing into the realm of the obvious. Like, if the US player were to sail his carriers, one at a time, deep into enemy territory and get them all sunk, obviously he'd have a hard time winning the game no matter how good his pilots and aircraft are. But I'm not talking about overtly stupid - just a US player who is, unlike the historical American Admirals, consistantly outplayed.

Anyhow, keep up the good work. These threads are fun to read. :)
User avatar
TheElf
Posts: 2800
Joined: Wed May 14, 2003 1:46 am
Location: Pax River, MD

Post by TheElf »

Alden wrote:I apologize for intruding, but...

If it results in a series of actions over say, one month, that greatly favor the Japanese in the early war, it's gonna be a broken model...

Results of combat are determined by so many factors other than the basic coefficients (Av. Exp, Mvr, Spd, what-have-you) of the units involved. What if, through superior deployments, making better tactical choices, acheiving better force concentration, etc, the Japanese player can create these favorable (to him) results *despite* any qualitative disadvantage he may have?

If the Japanese player can do this consistantly, he should be able to kill those shiny, hyper-trained American pilots as fast as the schools can produce them. A truly broken model would be one where Allied superiority ratchets up, inexorably, at some fixed rate, regardless of events in the game. Because that's not a model.

I work in modeling, of a more humdrum nature - I write computer models for gravity-powered hydraulic systems (i.e. sewers ;)) and I see a similar impulse in my users all the time - they know what the flow rates are at a particular manhole because they've taken measurements. Then they want to force the model to exhibit these flows at that location. But they are just feeding results in as input - which will give you "accurate" graphs and charts but teaches you nothing about your system, the real one in the ground, that you didn't already know. The nature of the system upstream of the calibration point is still mysterious. This is useless because it doesn't allow them to then try out "What Ifs".

This is just like in the game. You can't say "by the end of 1944, the USN should be shooting Japanese planes down at such and such a rate" - because this condition is dependant on upstream events. The conditions at the end of 1944 are the result of events a year, two years before. This is a point Mogami has made a few times in the threads on these topics. Its been generally ignored, though. Maybe because its self-evident (in which case I needn't have bothered writing this post) but I think its a really important one on a theoretical level.

The outcome of 100 Zeros v 100 Wildcats both at 0 Fatigue and 80 Av Exp, both at 10,000 feet on CAP missions might be interesting academically, and might be a useful baseline (if people can agree on what the outcome "should" be), this "series of events over one month" is meaningless because it can occur only in a broader context. Surely the esteemed mdeihl isn't arguing that no matter how well he plays or how foolishly the US player plays, from 1/1/42 to 1/1/43 the Japanese player should always suffer 1.0-1.2 (or whatever) Air to Air losses? And from 1/1/43 to some other date the ratio should be something else and so on. Again, that's not a model - its results from the one time the "model" ran during the actual interval of 1942-1945.

I realize that my point risks lapsing into the realm of the obvious. Like, if the US player were to sail his carriers, one at a time, deep into enemy territory and get them all sunk, obviously he'd have a hard time winning the game no matter how good his pilots and aircraft are. But I'm not talking about overtly stupid - just a US player who is, unlike the historical American Admirals, consistantly outplayed.

Anyhow, keep up the good work. These threads are fun to read. :)
Amen Brother...I want everyone to read this post 10 times, except Mogami of course ;) . I haven't seen it said better.
IN PERPETUUM SINGULARIS SEDES

Image
User avatar
TheElf
Posts: 2800
Joined: Wed May 14, 2003 1:46 am
Location: Pax River, MD

Post by TheElf »

Mogami,
I ran some tests a couple months ago much like the F4F v A6m ones you've been posting around here lately. Instead of the Zero though I tested the George and the Frank. Two much improved aircraft over the Zero, and in some circles, the George in particular was said to rival, even be better than the F6F... (standing by for Allied response that is sure to be overwhelming)
But here's the evidence to back this belief up..."According to VF-17 Skipper, LCDR Marshall Beebe, an accomplished Test pilot(he flew the George post war) and ace the Shiden-kai(George) '...was equal to or better than the F6F'"
So what am I getting at? I think an important test of the air Combat model's "sense of fairness" let's say, would be an aircraft that should put the Japanese on at least even terms with the 2nd gen Allied aircraft like the F6F or F4U. Coupled with varying EXP levels this test might show the OTHER extreme of the Model.
When I ran these tests, though I lost the results months ago, the outcomes in a vaccum, were disheartening to say the least. I'm talking right up there with the Zero v F6F/F4U. Even with allied pilots set to 10 EXP! and Japs at 99! I would be interested to see how the WiTP Alpha treats such engagements. Perhaps also, it would shed some light on other aspects/shortcomings of the model

I also realize that when we tested the George a higher octane fuel was probably used, but I'm really not interested in starting THAT debate again.

v/r
Elf
IN PERPETUUM SINGULARIS SEDES

Image
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

Octane

Post by mogami »

Hi, I'm no mechanic. What is the performance impact of using 100 octane instead of 92 octane?

Japanese Navy fighters used 92. Japanese Army fighters used 87.

I've no idea how much if any 100 ocatane the Japanese captured. I do know
prior to the embargo Japan tried to corner the market on US 100 octane fuel.
(So I'm guessing they had both some use for it and some supply of it.)

They designed a recon aircraft (Diniah) for 100 octane. This is the only Japanese aircraft I have found 100 octane specificly mentioned with.

In UV's long campaigns the Allies have several (well prehaps 4 ) fighter Sqdns flying P-39/P400. The pilots are not bad. (in fact I often disband 2 of the groups early just to have 2 full sqdns stocked with surplus pilots all of whom are over or close to 70. I make these my 'pocket' sqdns (from a chess term used by hustlers who keep a surplus piece in their pocket and put it on the board when their opponent is distracted in speed games)

I never use these groups except in dire straits. They are kept for when their ac change to P-38. (then I have 2 full well trained/rested/morale P-38 groups)

In WITP it is impossible for the Japanese player to predict how many airgroups will be needed at any location 2-3 years down the road. However when I am planning for Japans turn 1, I pocket four fighter daitai 2 Army and 2 Navy. The 2 Army Daitai train up to max level the 2 Navy Daitai are already fully trained. (pilots between mid 80's to mid 60's) These 4 groups will never fly while there exist other groups. Their purpose is to one day be equipped with the best aircraft available and be introduced in number to a battle where their number and quaility can produce a reversal of fortune. (as much as 72 IJA and 54 IJN aircraft can impact a single battle) 126 fighters are not much but depending on the circumstance it might be a deciding factor. (Adding 126 of my best to Turkey Shoot type battle might allow a few bombers to get through)

I also don't intend to ever use a group in combat that contains more then 1/3 pilots below trained levels (60 for IJN 55 for IJA) I will disband groups to maintain quality before I dilute them. (2 shot up Daitai, rather rebuild both with untrained pilots combine the trained into one group and rebuild the other from scratch)


Well the point of this post is I think the Japanese player should always keep one eye on the future and insure to maintain a reserve of aircraft and pilots (the pilots really) To have for use at a time and place the enemy is unprepared to encounter them. Having a Daitai or 2 of George fighters will be nice. Having the Daitai's pilots be skilled will enhance their performance. (I hope)
Image




I'm not retreating, I'm attacking in a different direction!
User avatar
TheElf
Posts: 2800
Joined: Wed May 14, 2003 1:46 am
Location: Pax River, MD

Post by TheElf »

One last thing before I go for the night. I think it is important that we should start talking about the things that we feel are important to our Air Model and making a list and checking it twice. Perhaps, Mogami, you could jot this stuff down or print it out and compile it for the designers who care. It may not result in any change, but at least through Mogami, who's open mind and generous posting keep my hopes alive, at least we'll be heard.

I'm talking about things that might be added to the current model to "flesh it out" if you will, give it more depth, more variables etc.

Consider this.

For those of you who don't know I fly Hornets for the US Navy. VFA-131 to be exact. And these are the things that we as fighter pilots see as important, in no particular order(for now)...

1)Experience(EXP)--You're overall time spent in the air. Flight hours, traps, air/air engagements(training or real), Kills (a big one), etc. this provides a sense of comfort in the air to a pilot, even in a combat situation for the first time. A sense of the air, his aircraft, how he can expect his men to react under fire etc. by no means should it be absolute, (ie a 90 EXP pilot should not ALWAYS best 65 EXP pilot.) There are INFINITE variables that affect Air Combat. This one attribute encompasses/affects/modifies ALL the following attributes. A catch all of sorts.

2)Airborne leadership(LDR)--This develops over time. Training helps, but combat is the real test, and will really develop this attribute. Some people are born with the innate ability to lead, and that extends to the air.

3)Training(TRA)--a starting level or initial standard of training, based on the current demand on the pilot pool. Low losses at the front, ie early success=fewer KIA/MIA, and more time to train better RPs. This attribute could be a composite with later experience, or just turn into EXP after they reach the front. Don't forget the pilots that are rotating back to the theater, or may be transfering in from another theater.

4)Proficiency(PROF)--a direct result of the quality and frequency of your training OR your combat. A pilot who flys every day is said to be "proficient". A Combat Pilot who flys Combat every day would be a proficient combat pilot. This attribute might flux depending on time in the AOR, intensity of operations, or steadily decay (much like ship SYS DMG increases over time) while resting. ie when the yanks pull into Sydney and are pre-occupied with the local Auzzie beauties ;)

5)Situational Awareness(SA)--this actually has a definition form TOP GUN, but I'll paraphrase..."The mental picture that a pilot paints of the events and circumstances around him" It is extremely important, but isn't constant. It varies with EXP, the evironment, luck, fatigue, the sun in your eyes, a well executed enemy tactic, a disadvantageous position etc. The analogy we use is that of the bucket. Is it full, overflowing, or do you have room for more?

6)Tactics(TAC)--A knowledge of, experience with, and the ability to...EMPLOY your Airplane. But not JUST your airplane, but those in your section, those in your division, and those in your squadron. For example the section Thach weave, the division Thach, slash attacks, the squadron bounce out of the sun. The list goes on and on.

You could also include the same knowledge/exp with ENEMY tactics. Are these the same guys you've been dueling with over the Owen Stanleys for the past month, or are they newcomers to the fight, complete with a new type of improved aircraft, and using new tactics.

7)Raw Talent(TAL)--let's face it. Not everyone is meant to be a fighter pilot, dive bomber pilot, level bomber pilot etc. Some guys can really bend their bird around in ways others can't. Yes its true this can be learned, but to some it is innate.

8)Luck--This randomizer shows up EVERYWHERE. It is what makes Aces into asses, and asses into Ace's(yes I just made that up)

How would this apply to the model? A traditional Navy squadron deploys 12 aircraft, or three DIVISIONS of four aircraft. Even in WWII as we know, the division was a basic unit (2 sxns of 2 a/c) I think the best way to resolve comabt in WiTP is to form divisions via some algorithm that uses a given attribute(Like LDR) to select division leaders(each with their own EXP, LDR, SA, TAC, etc). The Squadron leader, usually the skipper, or another high ranking member of the squadron, leads the whole unit in the intial stages, but the divisions split up, or gang up on their adversaries after the initial advantage is resolved. Those divisions enter combat together and either succeed or they don't relative to the attributes, of their enemy, and the initial advantage at the start of the fight.

Once combat is started, the attributes of the LDRs and the rank and fiel pilot, the current morale level, SA determine how well they maintain section and division intergrity. When a new round starts this would affect the following combat, perhaps the first combat seperated the division, and the 2 individual sections are now working independently. One section finds an enemy plane at a disadvantage, and dipatches it. The othe is beset upon by a larger sub unit of the enemy force and they are further seperated into singles or even dispatched themselves. Each step of combat though is affected, modified, rolled based on these attributes, special situations could be created basedf on tactics, luck, environmentals, leadership, expeience, and of course the relative performance othe involved aircraft.

Feel free to add to or expand this list, and keep track of these things.

Mogami if you think this "List" warrants its own thread, make it happen. I'll be at sea for the next 2 weeks with limited internet access at best.

v/r
Elf
IN PERPETUUM SINGULARIS SEDES

Image
User avatar
TheElf
Posts: 2800
Joined: Wed May 14, 2003 1:46 am
Location: Pax River, MD

Post by TheElf »

I'd just like to apologize for my poor grammar, and spelling my previous post. Been a long day...
IN PERPETUUM SINGULARIS SEDES

Image
User avatar
Daniel Oskar
Posts: 112
Joined: Fri Dec 15, 2000 10:00 am

Post by Daniel Oskar »

TheElf wrote: 8)Luck--This randomizer shows up EVERYWHERE. It is what makes Aces into asses, and asses into Ace's(yes I just made that up) Elf
"I'd rather be lucky than good" - anonymous naval aviator
User avatar
TheElf
Posts: 2800
Joined: Wed May 14, 2003 1:46 am
Location: Pax River, MD

Post by TheElf »

Daniel Oskar wrote:"I'd rather be lucky than good" - anonymous naval aviator
Yup..

Hey Dan,
I was a Doer Bird in 98'. I vaguely remember your name, but can't place the face. I don't think we ever flew together.

Elf
IN PERPETUUM SINGULARIS SEDES

Image
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

Forum input

Post by mogami »

Hi, 2by3 reads these threads (at least when they pertain to the game)

Also Kid comes through and adds any new idea to the "list without end"

But I'll also keep track of them.
Image




I'm not retreating, I'm attacking in a different direction!
User avatar
Tristanjohn
Posts: 3027
Joined: Wed May 01, 2002 4:50 am
Location: Daly City CA USA
Contact:

Post by Tristanjohn »

Alden wrote:I apologize for intruding, but...

If it results in a series of actions over say, one month, that greatly favor the Japanese in the early war, it's gonna be a broken model...

Results of combat are determined by so many factors other than the basic coefficients (Av. Exp, Mvr, Spd, what-have-you) of the units involved. What if, through superior deployments, making better tactical choices, acheiving better force concentration, etc, the Japanese player can create these favorable (to him) results *despite* any qualitative disadvantage he may have?

If the Japanese player can do this consistantly, he should be able to kill those shiny, hyper-trained American pilots as fast as the schools can produce them. A truly broken model would be one where Allied superiority ratchets up, inexorably, at some fixed rate, regardless of events in the game. Because that's not a model.
It could be a model and an accurate one at that. Just depends.
I work in modeling, of a more humdrum nature - I write computer models for gravity-powered hydraulic systems (i.e. sewers ;)) and I see a similar impulse in my users all the time - they know what the flow rates are at a particular manhole because they've taken measurements. Then they want to force the model to exhibit these flows at that location. But they are just feeding results in as input - which will give you "accurate" graphs and charts but teaches you nothing about your system, the real one in the ground, that you didn't already know. The nature of the system upstream of the calibration point is still mysterious. This is useless because it doesn't allow them to then try out "What Ifs".
This made me think of The Bay Model over in Sausalito. I talked to a worker there--he either was or is still is with the Army Corp of Engineers--who says they're going to be looking to completely refurbish that model for future studies of the bay. That might beat your current Ed Norton gig. :)

Anyway, your point's a good one. We don't want WitP to be built as a self-fulfilling prophecy of some kind.
This is just like in the game. You can't say "by the end of 1944, the USN should be shooting Japanese planes down at such and such a rate" - because this condition is dependant on upstream events. The conditions at the end of 1944 are the result of events a year, two years before. This is a point Mogami has made a few times in the threads on these topics. Its been generally ignored, though. Maybe because its self-evident (in which case I needn't have bothered writing this post) but I think its a really important one on a theoretical level.
Well, Mogami's point does have merit but unfortunately it was made amidst the general claim and clamor that the current UV is not broken in its own right (when of course it most certainly is) and that even if it were brken before early results from tests of its successor's alpha version indicate that "all would be well" even if the older model was off by "a little bit" with regard to early-game Japanese success because the functioning nature of the WitP Japanese replacement-pilot pool combined with ingame losses to air combat: only 10 pilots per month into this pool with ratings of 60 should in theory keep the Sons of Nippon down after any initial spike of success versus the Allies, always dependent on some critical-mass loss of Japanese pilots along the way.

The thing is that's not an accuarate model, either. Necessarily.

The thing is UV has the Japanese flying superior aircraft with superior pilots under superior leadership; even goes so far as to "model" Allied USN/USMC pilots/planes as starting a given scenario (#14 leaps readily to mind in this regard) fatigued and with seriously-low morale, so you see it isn't as if one were just trying to maintain some healthy balance in play with the game but rather it is very much more the case that one is attempting to affect a reasonably historic model balance for the first time.
The outcome of 100 Zeros v 100 Wildcats both at 0 Fatigue and 80 Av Exp, both at 10,000 feet on CAP missions might be interesting academically, and might be a useful baseline (if people can agree on what the outcome "should" be), this "series of events over one month" is meaningless because it can occur only in a broader context. Surely the esteemed mdeihl isn't arguing that no matter how well he plays or how foolishly the US player plays, from 1/1/42 to 1/1/43 the Japanese player should always suffer 1.0-1.2 (or whatever) Air to Air losses? And from 1/1/43 to some other date the ratio should be something else and so on. Again, that's not a model - its results from the one time the "model" ran during the actual interval of 1942-1945.
I agree with that.

But frankly, the game was screwed up fundamentally years ago with its early-war pro-Japanese slant. From what I can see UV is no more than a glorified patch of sorts to Gary's old notes to a mistaken database going back many years! :)

To get it "right" one would need to revisit and re-examine the project from the ground up, and that's just not going to happen at this late date. You can see for yourself what the forum's calling for at the moment: more multipliers yet, which of course can only offer more complication and thus greater chance still for unexplained confusion in the results.

Not good.

And as you've noted (or clearly implied) no agreement has been reached even on what a reasonable baseline might be, much less a definition of what "good results" should look like, so there you go.
I realize that my point risks lapsing into the realm of the obvious. Like, if the US player were to sail his carriers, one at a time, deep into enemy territory and get them all sunk, obviously he'd have a hard time winning the game no matter how good his pilots and aircraft are. But I'm not talking about overtly stupid - just a US player who is, unlike the historical American Admirals, consistantly outplayed.
A somewhat false perception there, as USN admirals were consistently "outplayed" early in the war; these men and the forces they commanded still ultimately prevailed, however. To tie this observation practically to the model: the suggestion would seem to be that America's war potentiality was not just merely something all comprehensive and vast, but the integral combat systems which went to make up this war matrix must have been natively imbued with no minor quality of resilience to have so surely and rapidly overcome this shaky start.
Anyhow, keep up the good work. These threads are fun to read. :)
I just knew we were having fun! :)
Regarding Frank Jack Fletcher: They should have named an oiler after him instead. -- Irrelevant
User avatar
mogami
Posts: 11053
Joined: Wed Aug 23, 2000 8:00 am
Location: You can't get here from there

Quit the BS

Post by mogami »

"The thing is UV has the Japanese flying superior aircraft with superior pilots under superior leadership; even goes so far as to "model" Allied USN/USMC pilots/planes as starting a given scenario (#14 leaps readily to mind in this regard) fatigued and with seriously-low morale, so you see it isn't as if one were just trying to maintain some healthy balance in play with the game but rather it is very much more the case that one is attempting to affect a reasonably historic model balance for the first time."


It gets a little old after a while. This taking an item you don't like and then distorting it. In Scenario 14 for those of you who don't know the 3 USN CV have morale from 60 to the mid 70's on turn 1. By turn 4 the first of the 12 Airgroups hits 99 morale. There is no combat before the 9th or 10th.
TJ overlooks the fact that many of the groups are short on aircraft that arrive while the groups rest. Also they all have surplus pilots. (So the groups do not get rookies) It appears to me the groups are being reorganized. They certainly do not suffer in any combat as a result.
TJ implies these USN pilots are being exposed to the Japanese while suffering under low settings. (There are 24 Operating A6M2 on the map on turn 1. and 10 A6M3. These groups are 10-15 points higher in morale but there is no way for any combat to take place. TJ knows no combat will take place before the Allied player decides. These settings have no impact on the historic model balance.
Using them only shows TJ doesn't know what he is talking about or he will stoop to anything to prove his point. (What OOB settings have to do with the model is also beyond me. Since these OOB setting change every turn and the USN airgroups will be 99 morale 0 fatigue when they enter enemy airspace. )

He has posted his own errors on more then one occasion as examples of UV being flawed. Where he gets the idea that there is an added "Japanese bonus" early in the games is silly. (Go load scenario 14 yourselves and look at the bias in action)

The A6M2 had longer range. The pilots were good and as of the period where most long scenarios begin undefeated. The USN pilots are just as good. Look for yourself for once TJ and then shut up. (load scenario 14 H2H look at the Japanese pilots then look at your USN pilots, then shut up)

His real beef with the makeup removed is the Japanese player is not forced into making futile air attacks and being wiped out.
He is carefull to omit the fact that the real bias in the game works against the Japanese. On Aug 7th the Japanese player can not tell Tanaka to go defeat the larger USN surface force on the night of the 8th. There is nothing in the game that defeats the USN that night. It might happen butit might also be a total defeat of the Japanese TF.

I'll tell you why units have the settings and levels they do in any scenario.
All units are May 1 1942 status. When you make a new scenario you select a new starting date and then add units. The units keep the same numbers. As a result at the start of any scenario there is a period in which all the units "adjust"
Airgroups short of aircraft in May 1942 get replacements. Airgroups with low morale recover. Fatigue works off. These things have no impact on any of the scenarios because no combat occurs. They are not "implanted bias" they are just a side effect of the way scenarios are built. Anyone who wants to can edit specific scenarios to raise levels at start. (but why bother the game will restore the units by the time they are used in combat)
Image




I'm not retreating, I'm attacking in a different direction!
mdiehl
Posts: 3969
Joined: Sat Oct 21, 2000 8:00 am

Post by mdiehl »

In re the George, F6F and F4U. One man's opinion about the George, even a pilot's, is worth almost nothing. If you want to look to the real merits of the a/c the information that you need is in the technical specs. Some guys for example hated the P-40. Others loved it (like Erik Shilling). When you go to anecdotal stuff it's pretty much "pick from amon the range of opinions the one you want to hear." Instead, I always go to the tech specs.

The Georgew was comparable to the F6F but not as durable. It was greatly inferior to the F4U. Consider that the difference between the F4U and the George is greater than the difference between the A6M2 and the F2A3, and note that Buffaloes fared most poorly against Zekes, and you get a better idea of the problem that George pilots faced fighting the best American planes. The F6F was not the best American plane. It was simply a decent middle-quality design that worked well on CVs. The F4U was far superior but resistence to its use was initially very strong owing to it's limited forward view -- a real problem when landing on a CV deck. After the kamikazes started showing up, however, F4Us were in high demand because of their overwhelming superiority as interceptors. Nothing in the Japanese aresenal came close to them.
Results of combat are determined by so many factors other than the basic coefficients (Av. Exp, Mvr, Spd, what-have-you) of the units involved. What if, through superior deployments, making better tactical choices, acheiving better force concentration, etc, the Japanese player can create these favorable (to him) results *despite* any qualitative disadvantage he may have?
Non sequitur. I was talking about comparable situations. Frankly, however, there are no altered tactics or options in the Japanese (or for that matter, American) player's hands that could affect CV vs CV engagements. If the model makes it likely that US CVs will lose in a face to face encounter with a comparable number of Japanese CVs at any time in 1942, the model is flawed.
If the Japanese player can do this consistantly, he should be able to kill those shiny, hyper-trained American pilots as fast as the schools can produce them. A truly broken model would be one where Allied superiority ratchets up, inexorably, at some fixed rate, regardless of events in the game. Because that's not a model.
Actually, the latter model would be better than one that pretends that the Japanese have much capacity to alter tactics, supply or other circumstances to favor the Japanese. Pretty much any sustained series of air combats against anything should wipe out Japanese pilots faster than they can be trained, owing to operational losses alone. But in air combat. Assume that for some reason the Allied player is limited to the use of F2A3s and pilots who have just completed flight school. It'll take longer, but the end result, depletion of the Japanese experienced pilot pool faster than they can be trained or learn, should result. If it does not, the model is broken.

I suppose that you can take it to reducto ad absurdem. "Yeah. What if American player commits his CVs 1 at a time and equips each of them with the Wright flyer?" My response is that such is not expectable or even reasonable play, so why waste my time with that kind of argument.
I work in modeling.... This is useless because it doesn't allow them to then try out "What Ifs".
I honestly don't care. If the "what ifs" allowed are extreme it does not represent anything that could have happened in WW2. It's like saying that the model is no good if it does not allow the John Stennis to appear off of Marcus Island on 10 December 1941.
You can't say "by the end of 1944, the USN should be shooting Japanese planes down at such and such a rate" - because this condition is dependant on upstream events.
No, it actually isn't that dependent on "upstream events" unless you change the entire history of American fuel development in the 1930s, engine development from 1939-1944, and undo a bunch of congressional acts in 1940. By 1941, American naval aviators were trained in superior tactics that allowed them to take a marginally inferior machine and fight, to a draw, much more experienced pilots in a marginally better machine. When you throw in the F4U (specifications first laid down in 1940 IIRC) and the F6F (1939 IIRC) you see that aircraft far superior to most of the planes available to the Japanese at any time, flown by pilots trained in superior tactics and available in huge numbers, were "in the pipe" before the first Japanese plane left the PH Strike Force CV decks on 7 December.

The presumption of the use of "reasonable" effort by the Allies pretty much means the Japanese lose. I can't see much that the Japanese player can change. There are no settings for "tactics" that will allow the Japanese to switch from 3-plane to 4-plane sections, or to mutual support. There should be no particular capacity to greatly enhance Japanese logistics (at least, not without starving other theaters of maintenance techs, fuel, and spare parts). So if the Allied player plays "reasonably well" and keeps pressure on in multiple fronts, the Japanese should lose pilots far faster than they can replace them.

Sorry if that does not appeal to your veltanschuang but that is simply and inexorably the way it was. To the extent that the Japanese are given some ahistorical flexibility to concoct a "miracle fix" the model is trash. It'd be like pretending that B17s were really the ship killers that everybody imagined them to be in 1939.
The conditions at the end of 1944 are the result of events a year, two years before. This is a point Mogami has made a few times in the threads on these topics. Its been generally ignored, though. Maybe because its self-evident (in which case I needn't have bothered writing this post) but I think its a really important one on a theoretical level.
It is generall ingnored because it is factually incorrect. Very little about the American strategic plan for dealing with Japane was affected by the events of 1941-1942. The need for more pilots was recognized in 1939. The need for more aircraft in 1940. The design specs for superior aircraft that outperformed everything cranked out by the Japanese -- largely laid down before the first shot was fired.
The outcome of 100 Zeros v 100 Wildcats both at 0 Fatigue and 80 Av Exp, both at 10,000 feet on CAP missions might be interesting academically, and might be a useful baseline (if people can agree on what the outcome "should" be), this "series of events over one month" is meaningless because it can occur only in a broader context.
Depends on what you mean by broader context. If you mean that in the early war the Japanese will have an advantage in the PI and SE Asia owing to interior lines of supply, more airfields, more aircraft, and the general advantage held by an attacker with well laid set-piece plans against a numerically inferior and logistically isolated force, sure. OK. When you speak of engagements like the Solomons campaign there should be very little that the Japanese player can do to alter the basic fact that he has insufficient pilots, machines, and logistical capacity to fight a sustained campaign against a comparably equipped foe.
Surely the esteemed mdeihl isn't arguing that no matter how well he plays or how foolishly the US player plays, from 1/1/42 to 1/1/43 the Japanese player should always suffer 1.0-1.2 (or whatever) Air to Air losses? And from 1/1/43 to some other date the ratio should be something else and so on. Again, that's not a model - its results from the one time the "model" ran during the actual interval of 1942-1945.
I can think of few examples that would substantially alter Japan's situation in the short term. In the long run, 5:1 loss ratios in 1944 and thereafter were inevitable, pretty much without regard to where the combats occurred.
I realize that my point risks lapsing into the realm of the obvious. Like, if the US player were to sail his carriers, one at a time, deep into enemy territory and get them all sunk, obviously he'd have a hard time winning the game no matter how good his pilots and aircraft are. But I'm not talking about overtly stupid - just a US player who is, unlike the historical American Admirals, consistantly outplayed.
All that should do is buy the Japanese player a little extra time by making each Japanese defeat a little more costly to the Allies. But as I said, it's hard to see what control players have (or SHOULD have) beyond sending certain ships and troops to certain destinations and letting the pieces have at each other.
Show me a fellow who rejects statistical analysis a priori and I'll show you a fellow who has no knowledge of statistics.

Didn't we have this conversation already?
Post Reply

Return to “War In The Pacific - Struggle Against Japan 1941 - 1945”