The Power of Inexperience / GreyJoy(A)-Rader(J)

[JohnDillworth]

Same settings...but with 70% CAP and 0 hex...

ALL the bombers got through....

Just for laughs why not try just 200 fighters? Maybe the math goes bad just based on gross numbers? Might be the software developer in me but I have seen stranger things

[Miller]

Very depressing outcomes from your tests GJ.........[:(]

To be honest I dont think anyone, including the developers has any real idea as to why this is happening, and what can be done to fix it.

[DTurtle]

ORIGINAL: JohnDillworth

Just for laughs why not try just 200 fighters? Maybe the math goes bad just based on gross numbers? Might be the software developer in me but I have seen stranger things

Yeah, I'm with this. If it was that easy to get bombers through, this problem should have been known for a long time - as those are numbers that can be reached very, very early. There has to be something else screwing with this.

Maybe it really is that you have too many fighters defending - definitely worth a test.

[princep01]

Ser Greyjoy, it appears Miller is exactly right. This game is borked at the levels you and Rader are playing. It is apparent that the game was never really tested concerning very large aerial engagements. Frankly, this does not appear to be a problem at much lower intensity or those levels of intensity seen thru mid-43 or so. That is the usual point most games end, so it isn't a huge surprise to see these results are surprising to most of us.

Frankly, as has been noted, naval flak (probably all flak) does not work well and the Japanese E escorts are completely out of whack. I suspect the flak and e-escort problems can be solved fairly easily, but the Cap/Raider problem is soemthing that might require a lot of owrk.

Good luck finding a good compromise with Rader. All of us will benefit if you do. Thanks for trying.

[Dan Nichols]

The ASW and AA problems are corrected to some degree with the DaBigBabes series of scenarios, but they didn't do a data base fix for the official scenarios with the latest official patch. As far as I can tell they are not planning on doing another patch, so there probably will not be any data base fix.

[ADB123]

ORIGINAL: Miller

Very depressing outcomes from your tests GJ.........[:(]

To be honest I dont think anyone, including the developers has any real idea as to why this is happening, and what can be done to fix it.

One might need to carve out the entire AE Air Combat model and replace it with the old WitP Air Combat model. Sure, lots of people said that they "hated" it, but it did give Marianas-type results under Marianas-type conditions.

The current Air Combat model is garbage, and no amount of rationalization can get around that. Every test everyone does shows this more and more clearly.

Designing Air Combat so that some number of Bombers must always get through if there are escorts (and sometimes even if there aren't) was a fundamental fallacy that has hurt the Air Combat part of AE right from the beginning.

House Rules can't fix this problem. Only coding can.

[GreyJoy]

ORIGINAL: DTurtle

ORIGINAL: JohnDillworth

Just for laughs why not try just 200 fighters? Maybe the math goes bad just based on gross numbers? Might be the software developer in me but I have seen stranger things

Yeah, I'm with this. If it was that easy to get bombers through, this problem should have been known for a long time - as those are numbers that can be reached very, very early. There has to be something else screwing with this.

Maybe it really is that you have too many fighters defending - definitely worth a test.

Ok, i'll try to empty hakodate and only use 200 fighters to defend against the same size strike...but i don't have many expectatipns to be honest...

I've been lucky till now cause Rader, untill the last turn, told me he was too scared to send hos betties and frances to attack my Surface fleet at Hakodate and face those 3000 crack fighterrs i have there....if he only had known....[8|]

[veji1]

There is definetly something broken here. I mean how could two players, having played together for 1 or 2 years IRL, being in 44 with their respective CV fleets more or less intact feel confident to keep on playing without risking a single turn savaging 2 years of investment ?

It looks that with this model the only solution for the allies late war is to advance very carefully only in areas where they can achieve complete and total air suppression with LBA alone. limits the avenues...

[EUBanana]

ORIGINAL: veji1

There is definetly something broken here. I mean how could two players, having played together for 1 or 2 years IRL, being in 44 with their respective CV fleets more or less intact feel confident to keep on playing without risking a single turn savaging 2 years of investment ?

It looks that with this model the only solution for the allies late war is to advance very carefully only in areas where they can achieve complete and total air suppression with LBA alone. limits the avenues...

I think the concentration of engineers in Japan is the issue, it means the Japanese airfields are unsuppressable. On top of that, as Japan makes its own supply obviously, and generally sits on a big stockpile of fuel, you can't cut the place off and bring a quick resolution.

So you can't starve them out, you can't bomb them out, and the game engine makes it impossible to achieve local air superiority if they are not "out", as we have seen.

I've not noticed these issues so far though in my own game. I think even somewhere like Luzon could be starved out and bombed out. I think this is a special case regarding Japan, a special case exacerbated by the fact that Greyjoy has attacked on an extremely narrow front and pretty much all of IJ can be marshalled against that front. Even so. It seems to me that any game is going to eventually run into these issues when its invasion of Japan time, assuming Japan is still capable of resistance (by which we mean capable of a 400 a/c strike, which seems a pretty low bar to me).

[EUBanana]

I think a lot of this stuff could be solved with caps probably, maybe soft caps rather than hard ones.

Like the number of of engineers and engineering vehicles that can fix an airfield should probably have some sort of limit. If you have a billion bulldozers it doesn't mean you're going to fix the runway in a nanosecond after all.

The same applies with the size of strike packages, though I do think that the way escorts are handled seems to have fundamental issues with them acting like bullet sponges.

[Miller]

ORIGINAL: ADB123

ORIGINAL: Miller

Very depressing outcomes from your tests GJ.........[:(]

To be honest I dont think anyone, including the developers has any real idea as to why this is happening, and what can be done to fix it.

One might need to carve out the entire AE Air Combat model and replace it with the old WitP Air Combat model. Sure, lots of people said that they "hated" it, but it did give Marianas-type results under Marianas-type conditions.

The current Air Combat model is garbage, and no amount of rationalization can get around that. Every test everyone does shows this more and more clearly.
Designing Air Combat so that some number of Bombers must always get through if there are escorts (and sometimes even if there aren't) was a fundamental fallacy that has hurt the Air Combat part of AE right from the beginning.

House Rules can't fix this problem. Only coding can.

To be fair to the devs that is a rather harsh statement. I think the air model works quite well with reasonable numbers on both sides. My game is at August 43 and things seem to be going fairly well in that department.

Bringing back the old WITP air model would be a huge step backwards IMO. It would just tip the problem way back the other way against the Japs later in the game. 1000 a/c strikes going in against a 1000 a/c CAP would usually mean 1000 attackers shot down without a single attack a/c getting through to drop anyhthing on target. Obviously it has gone totally the other way in this game, but after all this game is very much "Off the map" in terms of how the real war was fought.

Whilst GJs approach has been very entertaining for us readers.....if perhaps he had been more methodical in his advance then Rader would not have been able to mass these numbers, or even produce as much if the oil had been cut off first. I realise this would be hard to do with Allied subs being easy targets for the E class escorts, but that is easier to fix in mods such as DBB.

I am fairly confident that with some further investigation that a solution (of sorts) can be found..........

[GreyJoy]

I'm reporting here the results posted in the main section concerning what happens if i don't overstack Hakodate (a level 9 AF, mind you).
So here i'm simulating an amphib TF composed of 30 ships (3 BBs + 2 CVEs + 25 APAs) loading troops at Hakodate, where is based a single fighter group composed of 200 P-47s with pilots with an avg experience of 80...against a raid composed of 50 Franks and 50 Betties from Tokyo (not a MASSIVE raid as you may guess...)
Results...again...are discouraging...
 
 
 
 
 
 
1st test: 50 Betties + 51 Franks (from Tokyo) against an Amphib TF at Hakodate. At Hakodate is based a fighter group composed of the usual 200 P-47s with 80 experience.

Settings: incoming strike at 6,000
P-47s set to Escort, 50% CAP, 50% Rest, 15k feet, range 0
--------------------------------------------------------------------------------
Morning Air attack on TF, near Hakodate at 119,53

Weather in hex: Moderate rain

Raid detected at 54 NM, estimated altitude 8,000 feet.
Estimated time to target is 18 minutes

Japanese aircraft
G4M3a Betty x 50
Ki-84a Frank x 51



Allied aircraft
P-47D25 Thunderbolt x 50 [&:]...50%CAP + 50% rest should mean fighters available for CAP duties...[&:]

Japanese aircraft losses
G4M3a Betty: 13 destroyed, 31 damaged
G4M3a Betty: 1 destroyed by flak
Ki-84a Frank: 19 destroyed

Allied aircraft losses
P-47D25 Thunderbolt: 1 destroyed

Allied Ships
CVE Chenango, Torpedo hits 4, and is sunk
BB Arkansas, Torpedo hits 1
APA William P. Biddle
CVE Sangamon, Torpedo hits 3, and is sunk
APA Heywood, Torpedo hits 1, on fire, heavy damage
CA Indianapolis



Aircraft Attacking:
43 x G4M3a Betty launching torpedoes at 200 feet nearly all the bombers got through...
Naval Attack: 1 x 18in Type 91 Torpedo
2 x Ki-84a Frank sweeping at 10000 feet *

CAP engaged:
52nd Fighter Group with P-47D25 Thunderbolt (16 airborne, 34 on standby, 0 scrambling [&:]
16 plane(s) intercepting now.
Group patrol altitude is 15000 , scrambling fighters between 1000 and 42000.
Time for all group planes to reach interception is 26 minutes

Ammo storage explosion on CVE Chenango
Ammo storage explosion on CVE Sangamon
Ammo storage explosion on CVE Sangamon
Banzai! - Hidaka C. in a G4M3a Betty is willing to die for the Emperor









2nd Test.

Same as usual with the P-47s on escort + 50% CAP (0% rest), alt 15k, range 0


--------------------------------------------------------------------------------
Morning Air attack on TF, near Hakodate at 119,53

Weather in hex: Thunderstorms

Raid detected at 80 NM, estimated altitude 8,000 feet.
Estimated time to target is 28 minutes

Japanese aircraft
G4M3a Betty x 50
Ki-84a Frank x 51



Allied aircraft
P-47D25 Thunderbolt x 100

Japanese aircraft losses
G4M3a Betty: 7 destroyed, 17 damaged
G4M3a Betty: 2 destroyed by flak
Ki-84a Frank: 25 destroyed

No Allied losses

Allied Ships
BB North Carolina, Torpedo hits 1
BB Arkansas, Torpedo hits 1, heavy damage
CA Indianapolis
APA Zeilin, Torpedo hits 1



Aircraft Attacking:
37 x G4M3a Betty launching torpedoes at 200 feet
Naval Attack: 1 x 18in Type 91 Torpedo

CAP engaged:
52nd Fighter Group with P-47D25 Thunderbolt (0 airborne, 67 on standby, 0 scrambling)
33 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 1000 and 42000.
Time for all group planes to reach interception is 26 minutes





3rd test

P-47s on 100% CAP, 15k, 0 range

--------------------------------------------------------------------------------
Morning Air attack on TF, near Hakodate at 119,53

Weather in hex: Light cloud

Raid detected at 76 NM, estimated altitude 7,000 feet.
Estimated time to target is 26 minutes

Japanese aircraft
G4M3a Betty x 50
Ki-84a Frank x 51



Allied aircraft
P-47D25 Thunderbolt x 200


Japanese aircraft losses
G4M3a Betty: 8 destroyed, 16 damaged
G4M3a Betty: 2 destroyed by flak
Ki-84a Frank: 24 destroyed

No Allied losses

Allied Ships
BB North Carolina, Torpedo hits 4, Kamikaze hits 1, on fire, heavy damage
APA John Penn
APA American Legion, Torpedo hits 2, heavy fires, heavy damage
APA Henry T. Allen
APA Harry Lee, Torpedo hits 1, heavy damage
CA Indianapolis



Aircraft Attacking:
47 x G4M3a Betty launching torpedoes at 200 feet ...again, most of the bombers got to the target unmolested.-..
Naval Attack: 1 x 18in Type 91 Torpedo

CAP engaged:
52nd Fighter Group with P-47D25 Thunderbolt (0 airborne, 134 on standby, 0 scrambling)
66 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 1000 and 42000.
Time for all group planes to reach interception is 33 minutes
48 planes vectored on to bombers

Banzai! - Agawa T. in a G4M3a Betty is willing to die for the Emperor
Inoue N. gives his life for the Emperor by ramming BB North Carolina
Heavy smoke from fires obscuring BB North Carolina
Banzai! - Ishimori H. in a G4M3a Betty is willing to die for the Emperor
Heavy smoke from fires obscuring APA American Legion
Banzai! - Miyajima B. in a G4M3a Betty is willing to die for the Emperor




....so even with small numbers (50+50 attacking planes against a max number of 200 defending fighters) the result is always the same: put a few escort and the bombers will hit their target...and hit hard!

I don't know if i'm doing something wrong in placing my CAP...i'd like to be supported by some one else doing some tests to see if it's me or the code...

In my tests however CAP doesn't work as i thought it should...no matter the numbers...1000 defending fighters are ineffective exactly like 50....

[TheLoneGunman_MatrixForum]

GJ, try having 4 groups of 50 aircraft each. Perhaps having 200 aircraft in one group is breaking the code.

[GreyJoy]

ORIGINAL: TheLoneGunman

GJ, try having 4 groups of 50 aircraft each. Perhaps having 200 aircraft in one group is breaking the code.

Ok, will try!

[castor troy]

ORIGINAL: Miller

ORIGINAL: ADB123

ORIGINAL: Miller

Very depressing outcomes from your tests GJ.........[:(]

To be honest I dont think anyone, including the developers has any real idea as to why this is happening, and what can be done to fix it.

One might need to carve out the entire AE Air Combat model and replace it with the old WitP Air Combat model. Sure, lots of people said that they "hated" it, but it did give Marianas-type results under Marianas-type conditions.

The current Air Combat model is garbage, and no amount of rationalization can get around that. Every test everyone does shows this more and more clearly.
Designing Air Combat so that some number of Bombers must always get through if there are escorts (and sometimes even if there aren't) was a fundamental fallacy that has hurt the Air Combat part of AE right from the beginning.

House Rules can't fix this problem. Only coding can.

To be fair to the devs that is a rather harsh statement. I think the air model works quite well with reasonable numbers on both sides. My game is at August 43 and things seem to be going fairly well in that department.

Bringing back the old WITP air model would be a huge step backwards IMO. It would just tip the problem way back the other way against the Japs later in the game. 1000 a/c strikes going in against a 1000 a/c CAP would usually mean 1000 attackers shot down without a single attack a/c getting through to drop anyhthing on target. Obviously it has gone totally the other way in this game, but after all this game is very much "Off the map" in terms of how the real war was fought.

Whilst GJs approach has been very entertaining for us readers.....if perhaps he had been more methodical in his advance then Rader would not have been able to mass these numbers, or even produce as much if the oil had been cut off first. I realise this would be hard to do with Allied subs being easy targets for the E class escorts, but that is easier to fix in mods such as DBB.

I am fairly confident that with some further investigation that a solution (of sorts) can be found..........

absolutely with Miller here, saying the air routines should be replaced with WITP's ones is too harsh as AE's routines are far better than WITP's, until we reach a certain stage. From that stage on, I see WITP's UBER Cap working better than AE's routines because swapping out Turkey shoots with Allied fleet slaughters is further away from a realistic outcome than WITP's results.

Besides that, in WITP I've seen enough people (and have done it myselve) that were able to overcome the so called UBER Cap and get bombers through to hit enemy fleets (not totally whiping them out), so in those late war scenarios it seems WITP really does a better job. Now show me someone that gets AE's thousands of fighters act as realistic Cap.

[GreyJoy]

ORIGINAL: TheLoneGunman

GJ, try having 4 groups of 50 aircraft each. Perhaps having 200 aircraft in one group is breaking the code.

Here u are... 4 groups of 50 planes each. Escort + 50% CAP, 0 range, 15k alt....


AFTER ACTION REPORTS FOR Sep 01, 45
--------------------------------------------------------------------------------
Morning Air attack on TF, near Hakodate at 119,53

Weather in hex: Partial cloud

Raid detected at 77 NM, estimated altitude 11,000 feet.
Estimated time to target is 22 minutes

Japanese aircraft
P1Y2 Frances x 50
Ki-84r Frank x 50



Allied aircraft
P-47D25 Thunderbolt x 100


Japanese aircraft losses
P1Y2 Frances: 11 destroyed, 31 damaged
P1Y2 Frances: 5 destroyed by flak
Ki-84r Frank: 10 destroyed

No Allied losses

Allied Ships
CVE Chenango, Torpedo hits 2, on fire, heavy damage
BB West Virginia, Torpedo hits 1
BB Valiant, Torpedo hits 1
CVE Sangamon, Torpedo hits 3, and is sunk
BB Colorado



Aircraft Attacking:
33 x P1Y2 Frances launching torpedoes at 200 feet
Naval Attack: 1 x 18in Type 91 Torpedo

CAP engaged:
347th Fighter Group with P-47D25 Thunderbolt (0 airborne, 17 on standby, 0 scrambling)
8 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 1000 and 15000.
Time for all group planes to reach interception is 26 minutes
347th Fighter Group with P-47D25 Thunderbolt (0 airborne, 17 on standby, 0 scrambling)
8 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 9000 and 42000.
Time for all group planes to reach interception is 36 minutes
347th Fighter Group with P-47D25 Thunderbolt (0 airborne, 17 on standby, 0 scrambling)
8 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 2000 and 15000.
Time for all group planes to reach interception is 30 minutes
347th Fighter Group with P-47D25 Thunderbolt (0 airborne, 17 on standby, 0 scrambling)
8 plane(s) not yet engaged, 0 being recalled, 0 out of immediate contact.
Group patrol altitude is 15000 , scrambling fighters between 5000 and 15000.
Time for all group planes to reach interception is 27 minutes

[TheLoneGunman_MatrixForum]

33 got through that time.

As opposed to 47, 37, and 43 when you had one huge super-group.

You changed the type of bombers involved though, which could account for the losses.

[Miller]

Actually that last attack result looks fairly realistic, if anything the only thing that looks a bit off is the accuracy of the bombers...7 hits from 33 is a bit high but they are launching against slow targets.

[JohnDillworth]

33 got through that time.

As opposed to 47, 37, and 43 when you had one huge super-group.

You changed the type of bombers involved though, which could account for the losses.

The results should be more consistent with the Marianas Turkey Shoot. Very Roughly same number of bombers in the strike package vs hellcats. A HANDFULL got through to release, but probably not 33. GJ thanks for trying. Looks like 1/10 the fighters got about the same result as 2000. This is strange. I've had big raids earlier in the war that that did not have this problem. I wonder what is influencing these results?

[TheLoneGunman_MatrixForum]

Hop on over to GJ's other thread in the general forum.

He posted another result with the same CAP at 6k feet and they wiped the bombers out.

So altitude is critical.