AA Stuff

[Symon]

ORIGINAL: oldman45
John, thanks so much for the work and the offer of the spread sheet. Shipboard AAA never felt right and I am looking forward to the "joy" of modding using your numbers.

Thanks!

You are welcome. But you probably won't get a lot more "joy" out of the new numbers. They are close enough to the old numbers to not make signifigant differences in outcomes. What I'm doing is lining up all the data so that it can be shifted up, down, left, or right, and retain its initial character and its "relative" value with respect to all the other weapons.

Perhaps your lack of "joy" is because game raid profiles are nothing like real raid profiles and therefore real gun profiles don't show real (historical) results. Just for example, the B-17s in game have Max_Alt between 35000 and 37500. We all know this is so. And every gamer ever born will note that most AA tops out at 32000 and will just fly above it. They can do that. It is acceptable within the game system.

Is it righteous? Hell no. The standard B-17 mission profile, over Germany, was at 15000, often with an IP to Release altitude of 10000. 1940 era thought put Coast Defense/Anti Ship bombers at 20000. Which was the profile run by the USAAC planes at Midway.

So what do we do? Limit the Max_Alt of airplanes because of the reality of their use? Can you imagine what would happen to a commercial wargame that restricted data to its actual nominal operational values? Dude ... !

Game has very wide "envelopes". Babes is developing "sweet spot" data. Play gamey, this won't help or hurt you much. Play righteous, you will be surprised.

Ciao. JWE

[oldman45]

The joy I was talking about was going line by line and editing. You're right though, the altitudes I find the B-24's used in the pacific was 8 to 10k. Since I tend to fly mission packages that are around the number used in the pacific I get the results I expect.

Another thing that would drive a designer nuts is the feel. When I see the results of raids on my ships, I tend to get disappointed but the effects of my AAA and when my planes go after the IJN the results seem worse than what it "feels" like it should. Are the numbers correct? I don't know, all I can say is it doesn't feel right.

Once I started playing babes, I was hooked. "Feel right" or not, what you guys have done in the mod is magic and it continues to be played by me day after day. Thanks to you and all the rest.

[Symon]

Well, Hope springs eternal from every True heart. Maybe this will perk ya up some. [;)]

Next is Eff for AAA. This won’t be for AAAW (AutoWeaps), since the effect mechanism is different, just for good old AAA (Artillery). The mechanism is fragmentation, so the mechanism is a frag density within a volume. More powerful shells will have a larger volume. So put a 5” shell 10’ from the target and you will do a lot more damage than a 3” shell the same distance (radius) away. A 3” shell will have to get a skooosh closer, or get lucky.

I've never been pleased with the shell wt as a parameter. Simple, easy, quick-and-dirty, but (as you say) somehow unsatisfying.

So, that’s the model plan. Generate the pattern densities and then normalize the values into the range of numbers the program looks for. As always, the model will be based on data that can be found in any ordnance manual, or at places like Navweaps, Nigelef, or some of the nifty Russian sites.

Model is an abstraction (highly simplified) of fragmentation pattern density studies by US Army Armament Research and Development Command.

Ciao. JWE

[witpqs]

When you say (paraphrasing) 'for AAA but not AAAW (AutoWeaps)', I take it that shells like 40mm that were fragmentation (proximity fused) are included as AAA but 20mm, .50cal, etc. are what's classed as AAAW in this sense?

[Sardaukar]

ORIGINAL: witpqs

When you say (paraphrasing) 'for AAA but not AAAW (AutoWeaps)', I take it that shells like 40mm that were fragmentation (proximity fused) are included as AAA but 20mm, .50cal, etc. are what's classed as AAAW in this sense?

I don't think 40mm was proximity fuzed in WW II, AFAIK, small enough fuzes came only in 70s for 40 mm.

[Symon]

ORIGINAL: witpqs
When you say (paraphrasing) 'for AAA but not AAAW (AutoWeaps)', I take it that shells like 40mm that were fragmentation (proximity fused) are included as AAA but 20mm, .50cal, etc. are what's classed as AAAW in this sense?

No, Sir. 40mm was explosive, but on a contact fuse. Fuses, in the day, were contact, or time (time subsumes the delay fuses). The first 'proximity' fuse was the VT (variable time). The smallest it could be squeezed down to was for a 3" shell.

The 40-mike didn't have it, so used the everyday flash contact fuse with a mechanical time delay that 'popped' the round after a timed number of seconds.

Everything was kinetic, but AAA was proximity kinetic, AAAW was direct kinetic, i.e., the round had to physically impact the target. A really interesting juxtaposition of Eff and Acc, wouldn't you say?

[ed] was slow. What Sardakuar said.

Ciao. JWE

[witpqs]

Thanks, guys. I had the mis-impression that 40mm was the smallest proximity fuze in WWII.

Yes - very interesting, of course!

[Dili]

If you can't get above 55 degrees, you ain't AA.

Some navies employed secondaries against anti torpedo and low level aircraft. Giving a very low ceiling for some current naval gun depending on game engine might be interesting, Yamato guns should be DP's ...

[Mac Linehan]

John -

Just downloaded (with great anticipation) your AAA data spreadsheet.

It is professionally done (I would expect nothing else), and absolutely beautiful - I will plug the numbers into my Babes scenario.

Five Gold Stars, Sir - you have met or exceeded all expectations.

5/54 Mark 42 Mod 9's forever...<grin>.

Edit - I had to change the Mod 10 (3/4 inch grade B armor for the gun house, DD's and above only) to Mod 9 (no armor, weather shield only - much lighter, for Knox Class Frigate, which is what I served on). I forgot, posted incorrectly, till a small voice talked to me at 0300 this morning...

SeaGunBunny Mac

[Symon]

Thank you all, very much. So off to AAA Eff, then. This one is turning out to be pretty easy (except for grinding through the math). There’s some good operational research studies on AAA effectiveness (US, UK, Russia, China, Canada) that are full of juicy math and data. There’s even some bottom line equations that can be rendered down into cube root dependencies of shell weight/area/power, etc.. They tend to consider AAA as one big “glop”, subsuming both effectiveness and accuracy. Maybe because by that time, everyone had decent fire control and VT, or an equivalent.

In the War-2 environment, and in the game system, it makes much sense to bifurcate the model into a strictly Effect portion and an Accuracy portion. The concept goes like this: You got some guns and some shells; If you were to pop a half dozen shells, from different national weapons, into a specific “standard” volume (all other things being equal), what would be an expectation value of damage to a plane within that specific “standard” volume. Obviously, six big shells will generate a higher “hurt” density than six smaller shells.

Now, the real bitch is getting those shells into the specific “standard” volume to engage the plane. That’s the Acc part of the algorithm. But once they get there, what damage will they do? That’s the Eff part of the algorithm.

I could wax enthusiastic about my attempts to look at shell surface area, and explosive charge per unit area, but found they all abstracted (one way or another) into cube root of shell wt. Then the explosive power of an individual shell (the fill wt %) got dialed in, also as a cube root. And so the math model proceeds. I think it looks good.

Anyway, Japan is done, cranking on the Allies.

Ciao. John

[dwg]

ORIGINAL: Symon

ORIGINAL: JuanG
The reason is probably that the majority of MkIV used the MkV single mounting, which was only capable of 55 degree elevation. Only the Battle class had the MkIV mount which was capable of 80 degrees.

Yes. I think the best solution is dink with type for the Mk III/IV (there's 3 entries for these) to make one of them a Nav gun. If you can't get above 55 degrees, you ain't AA.

Except that RN AA doctrine allocated destroyer LA guns to barrage defence of the ships they were escorting. HA AA was recognised as superior, but is only needed in the final phase of an attack, first the attackers have to get through the barrage defence, and even LA guns can get high enough if the range is long enough (you can often see this in the reports of actions in the Med). I don't know how the game internals handle the differences between Naval, DP and AA guns and whether this is a significant point or not, but the 4", 4.5" and 4.7" were all expected to perform at least a partial AA role, even in the LA mountings.

[Symon]

ORIGINAL: dwg
Except that RN AA doctrine allocated destroyer LA guns to barrage defence of the ships they were escorting. HA AA was recognised as superior, but is only needed in the final phase of an attack, first the attackers have to get through the barrage defence, and even LA guns can get high enough if the range is long enough (you can often see this in the reports of actions in the Med). I don't know how the game internals handle the differences between Naval, DP and AA guns and whether this is a significant point or not, but the 4", 4.5" and 4.7" were all expected to perform at least a partial AA role, even in the LA mountings.

Very sorry dwg, but the game only abstractly models vanilla tripple-A. In fact, there's a low ceiling cut-off. For AAA with ceiling above a "fuzzy" 30000, it cuts-off low around a "fuzzy" 6000, i.e., won't (can't) engage planes below its cut-off.

It's an abstract model, not a simulation. What we're trying to do is merely smooth and adjust the data so that it gives more reasonable results over 1600 square mile hexes and across 4 years of development. There's really no way to bend the model to accommodate every nation's doctrinal tricks, special shells, or any of that technical stuff that wargamers love so well. It's a vanilla HA AAA model, and a very abstract one, at that. Again, sorry. [8D]

Ciao. JWE

[dwg]

ORIGINAL: Symon

ORIGINAL: dwg
Except that RN AA doctrine allocated destroyer LA guns to barrage defence of the ships they were escorting. HA AA was recognised as superior, but is only needed in the final phase of an attack, first the attackers have to get through the barrage defence, and even LA guns can get high enough if the range is long enough (you can often see this in the reports of actions in the Med). I don't know how the game internals handle the differences between Naval, DP and AA guns and whether this is a significant point or not, but the 4", 4.5" and 4.7" were all expected to perform at least a partial AA role, even in the LA mountings.

Very sorry dwg, but the game only abstractly models vanilla tripple-A. In fact, there's a low ceiling cut-off. For AAA with ceiling above a "fuzzy" 30000, it cuts-off low around a "fuzzy" 6000, i.e., won't (can't) engage planes below its cut-off.

It's an abstract model, not a simulation. What we're trying to do is merely smooth and adjust the data so that it gives more reasonable results over 1600 square mile hexes and across 4 years of development. There's really no way to bend the model to accommodate every nation's doctrinal tricks, special shells, or any of that technical stuff that wargamers love so well. It's a vanilla HA AAA model, and a very abstract one, at that. Again, sorry. [8D]

Which I understand, however my point is that even LA guns have an AA ceiling high enough they really need to be considered in a general AA model. If we look at the 4.5" Mk IV, on the 50 degree elevation single mount (iow the gun you were discussing), then the generally reputable navweaps, while noting limitations in the AA role, gives it an AA ceiling of 29,910 ft. See http://www.navweaps.com/Weapons/WNBR_45-45_mk1.htm

Picture the engagement envelope as a hemisphere with a cone cut out of it at the elevation limit, which is intersected by a plane at the altitude of the target aircraft. Both HA and LA guns may have an AA ceiling high enough to allow the target to be engaged, but what the low elevation limit does is restrict the time in the engagement envelope in comparison with a higher elevation mounting, which has a narrower cone into which it is unable to fire. You can model this by incorporating the AA ceilings of all weapons, whether LA, DP or HA, but modifying their effectiveness with regard to the elevation limits, as that will govern the time within the engagement envelope/number of shots able to be taken. No need to change the AA algorithm, just work out what it's trying to model IRL and modify the data to reflect reality.

[oldman45]

ORIGINAL: dwg

ORIGINAL: Symon

ORIGINAL: dwg
Except that RN AA doctrine allocated destroyer LA guns to barrage defence of the ships they were escorting. HA AA was recognised as superior, but is only needed in the final phase of an attack, first the attackers have to get through the barrage defence, and even LA guns can get high enough if the range is long enough (you can often see this in the reports of actions in the Med). I don't know how the game internals handle the differences between Naval, DP and AA guns and whether this is a significant point or not, but the 4", 4.5" and 4.7" were all expected to perform at least a partial AA role, even in the LA mountings.

Very sorry dwg, but the game only abstractly models vanilla tripple-A. In fact, there's a low ceiling cut-off. For AAA with ceiling above a "fuzzy" 30000, it cuts-off low around a "fuzzy" 6000, i.e., won't (can't) engage planes below its cut-off.

It's an abstract model, not a simulation. What we're trying to do is merely smooth and adjust the data so that it gives more reasonable results over 1600 square mile hexes and across 4 years of development. There's really no way to bend the model to accommodate every nation's doctrinal tricks, special shells, or any of that technical stuff that wargamers love so well. It's a vanilla HA AAA model, and a very abstract one, at that. Again, sorry. [8D]

Which I understand, however my point is that even LA guns have an AA ceiling high enough they really need to be considered in a general AA model. If we look at the 4.5" Mk IV, on the 50 degree elevation single mount (iow the gun you were discussing), then the generally reputable navweaps, while noting limitations in the AA role, gives it an AA ceiling of 29,910 ft. See http://www.navweaps.com/Weapons/WNBR_45-45_mk1.htm

Picture the engagement envelope as a hemisphere with a cone cut out of it at the elevation limit, which is intersected by a plane at the altitude of the target aircraft. Both HA and LA guns may have an AA ceiling high enough to allow the target to be engaged, but what the low elevation limit does is restrict the time in the engagement envelope in comparison with a higher elevation mounting, which has a narrower cone into which it is unable to fire. You can model this by incorporating the AA ceilings of all weapons, whether LA, DP or HA, but modifying their effectiveness with regard to the elevation limits, as that will govern the time within the engagement envelope/number of shots able to be taken. No need to change the AA algorithm, just work out what it's trying to model IRL and modify the data to reflect reality.

Could the engine really handle that?

[Symon]

ORIGINAL: dwg
Which I understand, ....

You have an excellent point. Can’t fault your logic. Just a couple points:

Babes tries to keep things working as close as possible within the game code algorithms. Yes, we use lots of code tricks, but only those that were developed in cooperation with Michaelm, and that expressly default to the stock condition if the magic data isn’t in the magic place. So we try to keep things as kosher as we can, but make them less filling and better tasting.

Trying to model LA AAA is a monster PITA in itself. Add to that the issue of making many more guns DP to support a limited AA capability (even with reduced ceiling values) and you immediately skew the default stock values into strange behavior. Not goodnik, my friend.

We are actually walking a very fine line here, skating on the thin ice of the code and the thin ice of the data as it relates to the code. Doing what you suggest is something that is red meat for a modder. I would certainly be willing to help. But it just doesn’t fit within the present concept space, if you know what I mean.

JWE

[el cid again]

ORIGINAL: DaveConn

Is is possible/feasible to represent FC upgrades by defining different "weapons" (maybe with changed accuracy) when the FC upgrades, and then requiring ships to upgrade to get the "new weapons" and benefit from the FC improvements?

--Dave

Sure it is. Specify the upgrade weapon slot and the upgrade date it takes effect.

[el cid again]

ORIGINAL: oldman45

ORIGINAL: dwg

ORIGINAL: Symon



Very sorry dwg, but the game only abstractly models vanilla tripple-A. In fact, there's a low ceiling cut-off. For AAA with ceiling above a "fuzzy" 30000, it cuts-off low around a "fuzzy" 6000, i.e., won't (can't) engage planes below its cut-off.

It's an abstract model, not a simulation. What we're trying to do is merely smooth and adjust the data so that it gives more reasonable results over 1600 square mile hexes and across 4 years of development. There's really no way to bend the model to accommodate every nation's doctrinal tricks, special shells, or any of that technical stuff that wargamers love so well. It's a vanilla HA AAA model, and a very abstract one, at that. Again, sorry. [8D]

Which I understand, however my point is that even LA guns have an AA ceiling high enough they really need to be considered in a general AA model. If we look at the 4.5" Mk IV, on the 50 degree elevation single mount (iow the gun you were discussing), then the generally reputable navweaps, while noting limitations in the AA role, gives it an AA ceiling of 29,910 ft. See http://www.navweaps.com/Weapons/WNBR_45-45_mk1.htm

Picture the engagement envelope as a hemisphere with a cone cut out of it at the elevation limit, which is intersected by a plane at the altitude of the target aircraft. Both HA and LA guns may have an AA ceiling high enough to allow the target to be engaged, but what the low elevation limit does is restrict the time in the engagement envelope in comparison with a higher elevation mounting, which has a narrower cone into which it is unable to fire. You can model this by incorporating the AA ceilings of all weapons, whether LA, DP or HA, but modifying their effectiveness with regard to the elevation limits, as that will govern the time within the engagement envelope/number of shots able to be taken. No need to change the AA algorithm, just work out what it's trying to model IRL and modify the data to reflect reality.

Could the engine really handle that?

Yes it can. The engine has a number of things one can use. The obvious one is ceiling - which isn't too high for LA. Another one is accuracy.

More important, perhaps, than LA guns used in a DP role is DP guns per se. And a special case is the "short guns" - which are "short" in terms of range and ceiling most of all - never mind the term refers to bbl length.

In general, effective range and ceiling is what matters. And these are not in identical units - range is in yards while ceiling is in feet. There is a tendency to use max range and ceiling - but this is not a good thing. IF AA reform is COMBINED with aircraft operational ceiling reform, things work very well indeed. Planes do not fly at service ceiling (ROC = 100 ft/min) nor at absolute ceiling (ROC = 0) - but a fraction of that. The fraction is greater for jets and rockets than for piston engines. And greater for turbo supercharged engines than for not.

[Symon]

ORIGINAL: dwg
Picture the engagement envelope ....

Hi brother dwg, Going to post the Eff numbers. Thought of you and ran some values for the LA 4.5s. You are basically looking at a “ceiling” of 5900-6300. And then you are looking at an AA-Eff that’s much lower than the Nav-Eff and an Acc that’s way lower. So make those guns DP and put them into stock, and you will nerf Brit DDs Nav combat performance unmercifully. There’s no way, in the present program, to cite a type attribute and have it default to stock.

In fact, I think any (every) naval gun works within the same parameters. An LA shotgun. Hootz gazottiez! Special shells, special doctrine, and none of it worth the time. One would have to have to figure out every single possible “ceiling” for every possible gun, and every possible Eff for every possible shell; and then figure out the Acc for each and every one of these.

Have done it for your 4.5 Mk IV, and will send the specs by pm. You will see what I mean.
Ciao. JWE

[Symon]

So, preview of Japan's AA Eff. Don't forget, these are the KABLAMMO (tm) values for individual shells.

Darn, the Babes "pull it out your butt" alborithm seems to work pretty good.

Ciao JWE

[Symon]

And then the Allied stuff.