Most of my air groups aren't participating in combat

[kfsgo]

ORIGINAL: LoBaron

- watch out for all the other variables with the potential to influence combat results

Your feeling may be right, the battles are less bloody compared to the early days, but don´t let this draw you to the
wrong conclusions.

In my experience - which is not so much running through campaigns per se but messing around with the relative effectiveness of aircraft - the issue isn't large numbers of aircraft but single large strikes. If you set up a scenario where you have, say, 1000 fighters on sweep attacking 1000 fighters on CAP, divided into five squadrons on each side, what you tend to see happen is that about 80% of each squadron will fly as a single large, coordinated strike (so, ~800 aircraft), while about 10-15% or so of each squadron will fly a strike separately. That's not set in stone - sometimes they all coordinate, sometimes they all fly as squadrons etc - but it seems to be the 'middle ground' as far as odds are concerned.

As far as how this is relevant to the issue - well, when the large strikes meet they do seem to have a much lower number of losses on everyone's side than when smaller aircraft packages are involved. The losses as a relative number for each side are about the same - maybe slightly more in favour of the attacker than they would be, given that the attacker is otherwise split up into fragments while the defender generally isn't - but total numbers are lower. It's not until the fragments start jumping in that losses tend to mount significantly. If I had to guess, I suppose it's probably a function of time - that big strike package has only got so long to hang around looking for something to shoot at, and the sky's a big place...

Now, as to whether this is wrong I'm not really sure I can say. My personal opinion would be that it it's probably a more sensible result than having hundreds of aircraft shot down, simply because that doesn't appear to have been the norm elsewhere - you have people getting worried about loss rates over 5-10% or so per mission etc. The bomber really did get through over Germany, after all; it would be interesting to have people post loss rates as a percentage of aircraft flown in some of these 'problematic' air battles.

[LoBaron]

CT, I just seem to have an easier time than you to admit when I am wrong on a topic, also I
have no problem to apprechiate your effort to identify bugs.

I was not really involved in the land based radar discussion, the strato sweeps are a matter of taste
and/or houserules, what you call "pre cap flak" was investigated partly by using a savegame provided
by Rob and me - we may just be a bit more patient to wait for conclusive analysis by a dev -, and
the data I posted to the coordination guide is correct and confirmed.

Don´t make things up please and lets stay on topic, thanks.

[n01487477]

ORIGINAL: US87891

ORIGINAL: michaelm

I can increase the number of combat rounds when there is a large number of fighters in CAP.

This increases the dogfights without making a lopsided change.

Here is updated EXE.

Michael, this is not a good idea. I thought your function was to fix bugs, but this is not a bug, this is an opinion, and one with no basis in game reality. Chasing after uninformed but loudly expressed opinions by less than 1% of the actual playing community, has already moved the game to the margins of un-playability. This takes it one more step towards instability. Please talk to John and Don before making any more changes to the calculation algorithms.

Thank you, Matt

Can you please explain what you mean by this rather blanket statement.

Just an interested party

[LoBaron]

kfsgo I tried to articulate this difference in post #30, this has not so much to do with
coordination but with the actual combat on approach to target.
The AE is quite variable in simulating different potential outcomes of raids depending
on the success for both sides of an attack.

It might be too variable on rare occasions but I have not seen proof for this yet, for sure
not in this thread.

If you look at different fates of similar sized heavy bomber raids in Europe (which is about
as close to a good comparision value we can get considering the ingame situation) you will
see a broad spectrum of different outcomes (from failed intercepts with nearly no losses
to both sides to heavy losses to the attacking formations) which is decided on a tactical
scale which is not exclusively governed by player decisions in WitP AE (so, chance).

The bottom line, and maybe the most important lesson for new players in this thread is:
Don´t let heavy bombers come into range of mainland Japan. Some raids will fail, some will
get through unscratched, but whatever the individual outcome is, its going downhill very fast
for the Japanese player from then on.

[LoBaron]

ORIGINAL: michaelm

I am not changing the 'formula' of the air battles.
Just the maximum number of times/rounds that are run to determine who/what engages in the dogfight.

As far as the designers are concerned, I was the one who originally set the maximum number of rounds in the code, more as a 'guesstimate', based on the factors I mentioned above.

Thats an interesting bit of information, thanks.

Does changing the max number of rounds also impact large strikes resolved in smaller packages or
just those where the whole formation is engaged in one large combat animation?

I guess it´s the latter?

[LoBaron]

ORIGINAL: GreyJoy
I think that a raid composed of 200 escorts and 400 bombers against 400 fighters CAP "should" suffer more losses (at least in terms on escorting fighters) than what we're seeing right now.

It might be, but only if you previousely are able to exclude other causes from influencing what you see.
And there are lots of potential causes stretching from pilot morale to CAP diverted to other
intercepts or tasks.

Sadly the combat reports are not as telling as one might wish, so there often is only "hidden" or
"cryptic" information available which takes some effort to uncover as well as misleading information.
Getting the picture is often only possible if you combine the right data from the combat report,
the combat animations and the settigns of your savegame plus a good picture on the situation
of your opponent.

E.g. misleading information is the plane count in the combat report. It does not tell you
anything about planes actually involved in the specific battle.

See viberpol´s combat report posted in this thread for a good example (e.g. post #44):

If you focus only on this:

Japanese aircraft
A6M5b Zero x 298

Allied aircraft
PBY-5A Catalina x 14

No Japanese losses

Allied aircraft losses
PBY-5A Catalina: 7 damaged
PBY-5A Catalina: 1 destroyed by flak

Japanese Ships
CVL Nisshin
CV Kaga, Torpedo hits 1
BB Nagato

then it might escape you that only a tiny fraction of fighters was even capable of
attempting an intercept and did not make it.

There is a lot of such information available, it takes quite a while to begin interprete
the data available in a correct way, some people never learn how to.

Personally I am not too happy with how the combat reports are structured in WitP, as a
more direct approach to the information would save a lot of discussions and confusion,
but things are as they are.

[kfsgo]

ORIGINAL: LoBaron

kfsgo I tried to articulate this difference in post #30, this has not so much to do with
coordination but with the actual combat on approach to target.
The AE is quite variable in simulating different potential outcomes of raids depending
on the success for both sides of an attack.

Yes. I am aware of that. It has everything to do with coordination, though, in that a greater number of dispersed strike packages seems to increase lethality - for everyone involved. So - you can say that for each particular strike there is X chance that not much combat will occur and Y chance that a lot will occur, or whatever, and that's all fine, but those numbers will tend to be buffered one way or the other by the number of separate strikes independently of how the separate combats go, I guess because the end result will more or less inevitably be that more of the dispersed aircraft actually make it into combat than they would as part of the coordinated package, given that we have a maximum number of combat rounds per fight to begin with.

[LoBaron]

Could be, for this reson I am interested as well if the modified exe changes this
behaviour to make large strike package losses slightly more realistic compared to what see now.
(Edit: without implying that they currently are not)

I am not in the position however to test the exe myself, as in my PBEM I am up to now perfectly content with
the results and prefer to leave it as it is, and I am unable to perform a conclusive series of test runs
required to determine the differences due to RL time constraints.

The exe is there, so I´d suggest that those who are not content with the current situation give it a try and
share their impressions.

[viberpol]

Not trying to brag but I used to be a Matrix playtester and pointed out several bugs in the early stage of AE... And such comments as...

ORIGINAL: US87891
Chasing after uninformed but loudly expressed opinions by less than 1% of the actual playing community, has already moved the game to the margins of un-playability. This takes it one more step towards instability. Please talk to John and Don before making any more changes to the calculation algorithms.

makes me scratch my head with despair... [8|]

It's exactly because of such loudly expressed opinions the game is being changed and made better. At first there are opinions of dedicated players who have thousands of turns played, chose not to sit quiet, ask and comment...
If the comments are taken seriously by the devs and the changes/fixes are made accordingly, it's just a different matter.

Don't be afraid. I trust Michael can sort out the sheep out of goats because he has inner understanding of the code. [;)]
We, the humble players, are only to give comments, nothing more. And that's the right place for it -- the tech forum.

ORIGINAL: LoBaron

See viberpol´s combat report posted in this thread for a good example (e.g. post #44):

If you focus only on this:

Japanese aircraft
A6M5b Zero x 298

Allied aircraft
PBY-5A Catalina x 14

No Japanese losses

Allied aircraft losses
PBY-5A Catalina: 7 damaged
PBY-5A Catalina: 1 destroyed by flak

Japanese Ships
CVL Nisshin
CV Kaga, Torpedo hits 1
BB Nagato

then it might escape you that only a tiny fraction of fighters was even capable of
attempting an intercept and did not make it.

LoBaron, in the same combat report you cite, there's 20 fighters actually intercepting the raid -- you call 7% of CAP a tiny fraction?
My example is of not the same "heavy" nature as the GreyJoy/Rader's ones, but it can be a symptom of the same problem... and that's exactly why I wanted to share it.

[LoBaron]

ORIGINAL: viberpol
LoBaron, in the same combat report you cite, there's 20 fighters actually intercepting the raid -- you call 7% of CAP a tiny fraction?

Yes, miniscule. More so if the OP refers to "Almost 300 fighters on CAP", your post #31, which simply is not the case.
But this has been discussed already. [;)]

My example is of not the same "heavy" nature as the GreyJoy/Rader's ones, but it can be a symptom of the same problem... and that's exactly why I wanted to share it.

No, it isn´t even the same league. You have an issue to get fighters in the air and those that are, are flying circles
at 29k. Simple as that.
This is something completely different to what Michael´s modification tries to adress.

For this very reason you did not notice any difference between the original and the modified exe.

[USSAmerica]

I can see where this thread is heading.  

[LoBaron]

To the airport? [:D]

[GreyJoy]

At a certain point i think that we all must simply stop and wait for the devs to answer the question if these results shown are WAD or not.
And i think this moment has come[;)]

[hades1001]

anyone has the chance to test the EXE given by Michael?

Please share some results here.

[GreyJoy]

I'd be glad to test it but i'd need a scenario modified by someone who knows how to use the editor

[LoBaron]

GreyJoy, would you and rader give the exe mod a try? Don´t feel compelled to do it, its just that you two are in the perfect situation
to evaluate the impact on large scale raids.

[LoBaron]

Sorry you´re right, better not impact a running game without preliminary test runs.

[rader]

ORIGINAL: LoBaron

It might be, but only if you previousely are able to exclude other causes from influencing what you see.
And there are lots of potential causes stretching from pilot morale to CAP diverted to other
intercepts or tasks.

Sadly the combat reports are not as telling as one might wish, so there often is only "hidden" or
"cryptic" information available which takes some effort to uncover as well as misleading information.
Getting the picture is often only possible if you combine the right data from the combat report,
the combat animations and the settigns of your savegame plus a good picture on the situation
of your opponent.

E.g. misleading information is the plane count in the combat report. It does not tell you
anything about planes actually involved in the specific battle.

This is exactly what the root of the question is... why are they not participating? I am curious about finding an answer to that ~ I am not saying that it is definitely a bug. Maybe it is working perfectly well in the developers' eyes; I am just curious what is causing 80+% of the defending fighters to sit back and do nothing (and if this is intentional). As you say, you can't tell from the combat animations or reports, so I am asking here to determine if there is anything that can be done to get them more involved.

Additionally, I am making sure this is not a new and accidental change in one of the betas, because we never noticed this behaviour before (possibly because we never had 1000+ a/c air battles before).

[PaxMondo]

ORIGINAL: michaelm

I am not changing the 'formula' of the air battles.
Just the maximum number of times/rounds that are run to determine who/what engages in the dogfight.

As far as the designers are concerned, I was the one who originally set the maximum number of rounds in the code, more as a 'guesstimate', based on the factors I mentioned above.

Michael,

I think your proposal has merit and would like to see it in the beta. What is the purpose of a public beta anyway? <rhetorical question>

I beleive we have sufficient evidence that large numbers of ac in combat produce unintended results. As you state, there is a hard limit on the number of rounds in the model (for good reason, a combat DOES need to end) ... that clearly is a variable to be tested ... and we now have quite a few games now where these large air battles are being fought. Potentially ripe testing grounds I would say ....

[michaelm75au]

The number of planes shown is the total number of planes available to CAP. This is not the number that will engage in combat. Due to the way the animations and the engine works, the number of planes is needed upfront before the actual combat takes place. Planes can be out of contact, out of position, landing to rearm/fuel, etc.

I can look to see if there is a way to limit the numbers to those actually available a particular air battle (or somehow reporting the actual number versus the number available), but this could just as easily start a 'None of my CAP is flying" threads...[:D]

This would not have any impact on the actual mechanics of the combat.