Air Combat Mod with test results

[DFalcon]

ORIGINAL: Sardaukar

You mean you did reduce the "Max load" number and it didn't reduce the hits ?

Yea, I reduced a B-17 from 6800 max load to 680 max load and it had no noticable effect. I should mention that all my tests were conducted at a 4 hex range, well with in normal range for the B-17. So I suspect that Max laod is only used to determine what the load out will be when subtituting weapons at extended range, for air base requirements, and for transport.

I suspected something about Max load when I noticed that there are US torpedo bombers with a Max load below the load of their torpedo and it did not effect them.

[Sardaukar]

Weird !! Well, so much about reducing effects of bombing by that.

[DFalcon]

ORIGINAL: Sardaukar

Weird !! Well, so much about reducing effects of bombing by that.

My reaction exactly! It may be a blessing in disguise though. With Max load tied to air base restrictions I was very reluctant to change it. Also with a reduction for extended range it might have made planes too ineffective. I think it may work out better this way.

[Sardaukar]

You may well be right with that. Airbase limitation is important factor in game.

[el cid again]

* CAP being able to intercept every raid thrown at a target, with no detriment for dealing with multiple raids, or distance to target hex.

This is quite untrue. One of the best strategies in UV or WITP is to throw multiple strikes at the same base. After a while, the CAP gets tired, and it gets less numerous (although the SECOND wave of CAP is usually STRONGER in numbers). The model clearly DOES consider tired pilots, and maintenance issues, and after 3-5 strikes, NO raid will be intercepted at all, as a rule.

[el cid again]

* Bomber and escort aircraft squadrons being able to combine irrespective of launching base location or command structure.

While bombers and escorts CAN combine, they USUALLY do NOT do so. They do this better when in range of an appropriate air HQ. I regard this as realistic and historical. Mostly, bombers attack a base in range order - that is the short range strikes come in first - the long range strikes later. This has always been true, in all mods, going back to UV 1.0.

[el cid again]

* Target bases being treated as a single airfield when they could be many. Using the WitP game scale a single hex would represent SE England in the Battle of Britain with Germany having several for parts of Europe. Although Germany could send large raids totalling many hundreds of aircraft, they could not all be intercepted by every fighter available. Nor could a few day’s raids put every airfield out of action. I believe this point is equally applicable to the shore bombardment effect on airfields.

This was not always so. You can see remnants of the earlier idea in the code - e.g. the "all bases in this hex" button. But you are confused about what they did do: the base situation is abstracted - which is not to say they all are a single base. Rabaul had two airfields BEFORE the war began, and Japan built more "in the same hex." One reason it is harder to take out a higher level airfield is that it represents more than one runway.

This does not mean the concept could not be improved. I would like to see damage reduce the airfield level - rather than take it all down as soon as it goes over 50%. At some point your level 5 airfield becomes a level 4 airfield - etc - until you fix it.

[el cid again]

If I could change the code I would. I do not even know what the code factors in outside of my observations. I can work with the data and if the results are good how I got them does not bother me too much.

Unless you work in isolation, this may be short sighted? Some poor guy comes along after you and does not know what you did, he puts in his plane with correct data - and it is now unbalanced relative to others in the game. Better to keep it simple and obvious how to create a unit - if you can.

[el cid again]

My early test showed that reducing the load value had no effect on the number of hits scored on a base. It also showed it had no effect on what ordinance an aircraft would carry. I suspect it is only used when flying extended range, but I have not tested this. Because I saw no results and it was tied to so many other aspects, like what base could support heavy bombers and extended range load out, I did not change this value. Too mant interactions not enough information.

In my testing I discoverd changing the standard pay load had a small effect. Hopefully enough to counter the problem you addressed of more bombers making it on target.

What you seemed to say was that you reduced the bomb capacity by 25%. That WOULD prevent a plane from carrying a torpedo or OHKA or whatever most of the time. IF instead you changed NORMAL bomb load, that is not a problem at all. Sometimes normal bomb loads were too high in fact.

[el cid again]

Ordnance listed is used only against ships. Payload is used against land targets, be it installations or units, IIRC.

Plain false. See my Ohka test thread. Clearly ordnance listed can be used against land targets. Further, clearly if the plane cannot lift the ordnance, it won't, even if it is listed as an option. IF you reduce payload to below the weight of a torpedo, you cannot lift it, and the AI won't.

[DFalcon]

ORIGINAL: el cid again

If I could change the code I would. I do not even know what the code factors in outside of my observations. I can work with the data and if the results are good how I got them does not bother me too much.

Unless you work in isolation, this may be short sighted? Some poor guy comes along after you and does not know what you did, he puts in his plane with correct data - and it is now unbalanced relative to others in the game. Better to keep it simple and obvious how to create a unit - if you can.

I agree totally! That is why I wanted the changes I made to be simple. I posted my method above, if I am unclear I will be happy to clarify what was done.

[DFalcon]

ORIGINAL: el cid again


What you seemed to say was that you reduced the bomb capacity by 25%. That WOULD prevent a plane from carrying a torpedo or OHKA or whatever most of the time. IF instead you changed NORMAL bomb load, that is not a problem at all. Sometimes normal bomb loads were too high in fact.

I thought the same thing, but planes will carry the default load even if it is above the Max load value. The more I think about it the more I like that I was forced to change the normal pay laod instead of the Max load value.

[DFalcon]

ORIGINAL: el cid again


Plain false. See my Ohka test thread. Clearly ordnance listed can be used against land targets. Further, clearly if the plane cannot lift the ordnance, it won't, even if it is listed as an option. IF you reduce payload to below the weight of a torpedo, you cannot lift it, and the AI won't.

This is not consistent with my tests. Perhaps there is another variable we are unaware of. The only things I know of that will change the load out is airbase size and extended range. In the stock scenario 15 there are US torpedo bombers with a lower Max load than the load value of the torpedo they carry. No problems have been reported so far.

[DFalcon]

Test Result #3

Day Air attack on Port Moresby , at 53,91

Japanese aircraft
A6M2 Zero x 32
Ki-21 Sally x 59

Allied aircraft
Buffalo I x 57

Japanese aircraft losses
A6M2 Zero: 5 destroyed, 4 damaged
Ki-21 Sally: 9 damaged

Allied aircraft losses
Buffalo I: 10 destroyed, 15 damaged
Hudson I: 5 destroyed

Airbase hits 7
Airbase supply hits 1
Runway hits 11

I ran this type of battle 10 times with the following settings;

Defenders
72 Buffalo I 80% Cap
72 Hudson I targets

Attackers
36 A6M 10% Escort
72 Ki-21

Average losses per dog fight (Standard / Mod)

Buffalo 14.6 / 9.7
A6M2 2.4 / 2.4
Ki-21 0.6 / 0.1
Airbase hits 15.1 / 11.7

I want to point out that for all these tests the scenario date is June 1942 so the zero bonus should not be in effect. All planes are set to the same altitude of 10,000 feet.

[DFalcon]

Test Result #4

Day Air attack on Lae , at 54,87

Japanese aircraft
A6M2 Zero x 42

Allied aircraft
Buffalo I x 32
Hudson I x 68

Japanese aircraft losses
A6M2 Zero: 6 destroyed, 18 damaged
Ki-21 Sally: 10 destroyed

Allied aircraft losses
Buffalo I: 8 destroyed, 3 damaged
Hudson I: 3 destroyed, 7 damaged

Japanese ground losses:
7 casualties reported

Airbase hits 2
Airbase supply hits 4
Runway hits 44

I ran this type of battle 10 times with the following settings;

Defenders
72 A6M2 80% Cap
72 Ki-21 targets

Attackers
36 Buffalo I 10% Escort
72 Hudson I

Average losses per dog fight (Standard / Mod)

A6M2 4.0 / 3.9
Buffalo 12.3 / 8.4
Hudson 6.1 / 4.0
Airbase hits 28.8 / 29.2

I am noticing a trend in favour of the attacking planes. Also the Mod had a higher number of hits in this run. It could be standard deviation or more bombers getting through the cap.

Tomorrow I will run the F4U up against the A6M3 another lopsided fight.

[el cid again]

This is not consistent with my tests. Perhaps there is another variable we are unaware of.

Here you must be correct. ONE of the variables is plane type.
OKHA will NOT be lifted by a torpedo bombar, apparently.

It also will not be lifted by a Sally - even if you define the weights/capacities right - but I have no clue why not??? Apparently there are SEVERAL variables that matter.

[DFalcon]

ORIGINAL: el cid again

Here you must be correct. ONE of the variables is plane type.
OKHA will NOT be lifted by a torpedo bombar, apparently.

It also will not be lifted by a Sally - even if you define the weights/capacities right - but I have no clue why not??? Apparently there are SEVERAL variables that matter.

Perhaps the OKHA is hard coded to a plane slot like carrier capability. Have you tried dropping it from a B-17?

[Sardaukar]

Don't you love all that hard-coded stuff... It's always been a trademark of Gary Grigsby...*sigh*.

[DFalcon]

Test Result #5

This test was done in response to a thread on the main board about the Oscar. It recreates an engagement where the Ki-43IIa got the better of some P-38J.

The Oscars were given a huge advantage. They have 20 experience on the P-38, altitude and out number them by 51 to 38 against 18. I think the Mod shows very well in this test and the stock data shows it’s weakness.

Day Air attack on Rabaul , at 61,88

Japanese aircraft
Ki-43-IIa Oscar x 51

Allied aircraft
P-38J Lightning x 18

Japanese aircraft losses
Ki-43-IIa Oscar: 11 destroyed

Allied aircraft losses
P-38J Lightning: 2 damaged

I ran this type of battle 10 times with the following settings;

Defenders
64 Ki-43IIa 80% Cap 10,000ft Experience 85
72 Ki-21 targets

Attackers
20 P-38J 10% Sweep 5,00ft Experience 65

Average losses per dog fight (Standard / Mod)

Ki-43IIa 7.6 / 2.0
P-38J 0.6 / 3.0

Tomorrow I will get back to my F4U vs. A6M3a test.

[el cid again]

Perhaps the OKHA is hard coded to a plane slot like carrier capability. Have you tried dropping it from a B-17?

No - but I don't use stock slots - and more than one type of plane that DIDN"T carry it works.