The bravery of being out of range

[6./ZG26_5tuka]

Agreed, but as Gielow said above, it will be interesting to see what player controlled flak is going to be like?  

Since we're only getting 20mms I'd not expect planes to get instant killed by player AAA but it will be enought to turn them off. More interesting would be developing proper Flak tactics like barraging or concentrating fire to fight off high valuable attackers (which is ver limited given the effective range of only 1000m max those guns have).

[6./ZG26_Custard]

 

 

Since we're only getting 20mms I'd not expect planes to get instant killed by player AAA but it will be enought to turn them off.

It's rather ironic that according to the video of German flak that I posted in the thread earlier, the 20 mm guns were more or less phased out by 1944 due to their ineffectiveness. Coordinated flak batteries (player controlled) would be interesting though.    

[6./ZG26_5tuka]

It's rather ironic that according to the video of German flak that I posted in the thread earlier, the 20 mm guns were more or less phased out by 1944 due to their ineffectiveness. Coordinated flak batteries (player controlled) would be interesting though.    

Indeed. This is also a good one on deployement and tactics of Flak batteries. It's a bit sad we don't have that complexity ingame (yet?).

https://www.youtube.com/watch?v=FWtQz7qp_v8

[56RAF_Roblex]

 It really depends how you fly.   A few days ago I attacked an airfield all by myself in a Yak,  took out 4 AA emplacements then flew home with empty guns and not a scratch on me.  I just had to dodge a lot which is why I used up all my ammo on jut four AA.   If there had been two of us flying sensibly, one to draw the ack while the other killed it,  then we could have got it all.   On the other hand, last night two of us got impatient and did not co-ordinate our attacks and we both got shot down :-)

Edited December 5, 2017 by 56RAF_Roblex

[unreasonable]

Here is a mission file - called Airfield defense 48 Low - uses Kuban map.

 

In it 48 109s attack an airfield defended by 8 Soviet 61-Ks.  When I say "attack" I mean that I have programmed them to fly in an attack pattern roughly similar to JG3 at Eindhoven, according to my understanding of how it worked.  They do not use the attack ground objects command as this makes them mill about: instead they attack in waves of 6 aircraft in line abreast, which make a low pass at 500 kph then turn and fly a circuit before coming in for a second run which they will do more or less in sequence, after which they head off to a rally area.  The result is that they are doing what amounts to low strafing runs (while not actually firing), while always flying fast and hardly ever flying straight and level.  The first wave is just turning into its second attack when the last wave passes over the airfield on it's first pass, so the airfield is under almost continual attack.  The advantage of this planned attack is that the aircraft all approach out of the rising sun (irrelevant in game but useful in RL) and the planes all travel along a similar path, making collisions far less likely, while still swamping the defense.

 

The 8 guns are on "low" AI, and have limited ammo (which I think just means that they have to pause to reload periodically (?) as I did not see any stop firing while planes were in range. They have an attack area command. They are spaced around the airfield perimeter. Press Ctrl-F5 to see POV from each gun position.

 

There is an Fw190 doing a circuit around the rally area. This should come up as the first aircraft when you press Ctrl-F2. The idea is that after the attackers have all left the airfield, you can count how many there are left using the sole Fw190 as a marker by cycling through Ctrl-F2.

 

I have run it twice and the AI shot down 3 and 1 109s in the two passes which seems the right order of magnitude. Two of the shoot-downs I observed came when planes were almost heading directly away from the guns - did not see any high deflection hits, which again seems right.

 

Be my guest and modify the guns to other AI settings in the Editor if you desire. I will get round to that myself but it might take a while.   Unzip into (My Missions) folder in Data/Missions and I think it should be visible in your game.... 

 

 

Airfield defense 48 low.zip

[Dakpilot]

Great effort unreasonable

 

Cheers Dakpilot

[unreasonable]

Thanks Dakpilot!   Just made a version with AA on "high"   - 35 109s made it to the rally point of which 4 had been obviously hit and of those 2 looked like they would not make it back to base - say 15 downed.   Obviously a huge difference - an order of magnitude between "low" and "high" -  even without being sure of the averages over a number of runs.

 

The wiki on the Eindhoven raid says that the RAF AA guns shot down 4-5 planes out of total losses of 15 - perhaps the German AAA when the planes were returning was more effective than the RAF  .  

 

Waiting for my Stackpole book on Bodenplatte to arrive: with luck there will be some more detail on all the raids and we can get a rough idea of the loss rates to airfield AAA in each case to compare to the game's outcomes in my Bodenplatte emulator.  

[Gambit21]

Do some reading on Y-29 if you haven't already.

[6./ZG26_Custard]

 

 

Obviously a huge difference - an order of magnitude between "low" and "high" -  even without being sure of the averages over a number of runs.

 

It would be interesting to do a comparison with German triple-A.

 

The disparity between lost aircraft and pilots lost, when looking at the different AAA is very surprising  when you compare stats supplied by TAW

 

 

I can only assume the AAA is set to ace?

[unreasonable]

I could replace the 61-Ks with German - or other Soviet - guns to check for effectiveness, as at least in my emulator, unlike a collection of stats from a MP server, you are getting a controlled experiment where the targets behave (almost) exactly the same each time. 

 

First I want to do a series of 10 runs for low-normal-high AI with the 61Ks to get a firm base for the difference of effectiveness of the AI levels.  Should have time this weekend.

 

To get a good grasp of which setting is more realistic by comparing against RL outcomes I need more detail on the facts of some of the airfield raids, especially the number of guns, which is obviously important, more so than the exact number of targets in a target rich environment.  

 

The only point I can say with certainty at the moment is that people who claim that "the AAA is right" - or for that matter "the AAA is wrong" need to specify what AI level they are talking about, since they clearly cannot both be right. (Although they could both be wrong ).

 

 

 

Do some reading on Y-29 if you haven't already.

 

That was a hell of a battle and a great victory for the Yanks but does it tell us much about AAA?  The thing is the Germans were unlucky to arrive when there were already US fighters in the air and got slaughtered.  Anyway, once   Bodenplatte: The Luftwaffe's Last Hope (Stackpole Military History). by John Manrho, Ron Putz.  arrives from Amazon I will see what information on any of the raids I can get to revise my experiment.

[6./ZG26_5tuka]

Did a test with german AAA formed up in battreries (3 batteries with 6 guns each standing close to each other) and had Laggs circling right above them ( 1000m above the Flak 38s, 1500m above the Flak 37s and 2000m above the 88s). Even on high skill level only once a 37mm managed to shoot one down within 10min while all others registred 'close misses' (not even damaged the aircraft). The 20mms in particular are very bad since it's bullets always crossed a fuselage lenght behind the enemy aircraft without remotely leading shots. The 88s on the other hand did light up the sky very well but seriously struggled to cause damage to fighters above 1.5 km.

Edited December 8, 2017 by 6./ZG26_5tuka

[Gambit21]

I play with AA all the time while building missions - low is the best setting. The end.

Even 'normal' is radar guided accurate.

[6./ZG26_5tuka]

I play with AA all the time while building missions - low is the best setting. The end.

Even 'normal' is radar guided accurate.

That is not the point though. We try to compare the effectiveness of russian AAA to the german one.

 

Generally though I agree even though there are issues with the low skill level (sometimes they dont open fire or way too late). I'd rather have historically sized battery of AAA set to low difficulty than a single gun at high skill.

[Gambit21]

That is not the point though. We try to compare the effectiveness of russian AAA to the german one.

Generally though I agree even though there are issues with the low skill level (sometimes they dont open fire or way too late). I'd rather have historically sized battery of AAA set to low difficulty than a single gun at high skill.

Ahh yes, forgot about the Russian vs German bit.

I have to say after building the Ju-52 missions, and now the A-20 missions (well Peshka standing in for now) 'low' is the only option on both sides 'in general' as you said that's workable, survivable, but adds some pucker factor.

 

I'd say on 'low' the AA in general behaves like I'd expect a real world AA 'Ace' to shoot.

Not ideal, but the best option at the moment.

 

If they ever re-work AI I'm going to have a lot of re-testing to do.

[[DBS]Tx_Tip]

Here is a mission file - called Airfield defense 48 Low - uses Kuban map.

 

In it 48 109s attack an airfield defended by 8 Soviet 61-Ks.  When I say "attack" I mean that I have programmed them to fly in an attack pattern roughly similar to JG3 at Eindhoven, according to my understanding of how it worked.  They do not use the attack ground objects command as this makes them mill about: instead they attack in waves of 6 aircraft in line abreast, which make a low pass at 500 kph then turn and fly a circuit before coming in for a second run which they will do more or less in sequence, after which they head off to a rally area.  The result is that they are doing what amounts to low strafing runs (while not actually firing), while always flying fast and hardly ever flying straight and level.  The first wave is just turning into its second attack when the last wave passes over the airfield on it's first pass, so the airfield is under almost continual attack.  The advantage of this planned attack is that the aircraft all approach out of the rising sun (irrelevant in game but useful in RL) and the planes all travel along a similar path, making collisions far less likely, while still swamping the defense.

 

The 8 guns are on "low" AI, and have limited ammo (which I think just means that they have to pause to reload periodically (?) as I did not see any stop firing while planes were in range. They have an attack area command. They are spaced around the airfield perimeter. Press Ctrl-F5 to see POV from each gun position.

 

There is an Fw190 doing a circuit around the rally area. This should come up as the first aircraft when you press Ctrl-F2. The idea is that after the attackers have all left the airfield, you can count how many there are left using the sole Fw190 as a marker by cycling through Ctrl-F2.

 

I have run it twice and the AI shot down 3 and 1 109s in the two passes which seems the right order of magnitude. Two of the shoot-downs I observed came when planes were almost heading directly away from the guns - did not see any high deflection hits, which again seems right.

 

Be my guest and modify the guns to other AI settings in the Editor if you desire. I will get round to that myself but it might take a while.   Unzip into (My Missions) folder in Data/Missions and I think it should be visible in your game.... 

 

 

Airfield defense 48 low.zip

Which flight did you fly with and how many times were you shot down Unreasonable?

 

It would be interesting to do a comparison with German triple-A.

 

The disparity between lost aircraft and pilots lost, when looking at the different AAA is very surprising  when you compare stats supplied by TAW

 

 

I can only assume the AAA is set to ace?

Interesting stats. You could ask them.

 

...

There has always been a noticeable difference in flak testing results when a test is done with different "target" parameters for Single and MP. 

The results using the same flak settings will be different when testing with:

 

Flak vs AI only.

Flak vs AI with 1 Player only.

Flak vs 1 Player only.

Flak vs AI with several Players.

Flak vs Several Players.

 

For instance. Testing a MP mission offline to rate flak effectiveness one can approximate what the flak numbers, types, settings and their disposition should be to achieve light, moderate or heavy losses. Then when the test is run on a server with only yourself you generally have a different outcome. Run the mission again with several players and you will have another outcome. Run the mission with a full contingent of Players, several things going on all at once on different areas of the Map and you will again get a somewhat different outcome depending upon server capacity and strength.

 

The flak piece logic can be manipulated with several wildcards. That it will generally key in on the Player, both SP and MP, or certain Players depending upon the ingress of the Players with regards to the disposition of the pieces is one of them.

While the controlled Flak vs AI plane only scenario is generating interesting results. An addition of a Player or Players in a SP or MP environment would generate their own results as well. 

 

Speaking for myself. The inherent vagueness of how the flak pieces respond to different stimuli makes manipulating flak interesting at times.

Having a set group which gives a well worn and tested out come 80% of the time for what you wish to achieve is the goto option. 

[unreasonable]

Which flight did you fly with and how many times were you shot down Unreasonable?

 

Interesting stats. You could ask them.

 

...

There has always been a noticeable difference in flak testing results when a test is done with different "target" parameters for Single and MP. 

The results using the same flak settings will be different when testing with:

 

Flak vs AI only.

Flak vs AI with 1 Player only.

Flak vs 1 Player only.

Flak vs AI with several Players.

Flak vs Several Players.

 

For instance. Testing a MP mission offline to rate flak effectiveness one can approximate what the flak numbers, types, settings and their disposition should be to achieve light, moderate or heavy losses. Then when the test is run on a server with only yourself you generally have a different outcome. Run the mission again with several players and you will have another outcome. Run the mission with a full contingent of Players, several things going on all at once on different areas of the Map and you will again get a somewhat different outcome depending upon server capacity and strength.

 

The flak piece logic can be manipulated with several wildcards. That it will generally key in on the Player, both SP and MP, or certain Players depending upon the ingress of the Players with regards to the disposition of the pieces is one of them.

While the controlled Flak vs AI plane only scenario is generating interesting results. An addition of a Player or Players in a SP or MP environment would generate their own results as well. 

 

Speaking for myself. The inherent vagueness of how the flak pieces respond to different stimuli makes manipulating flak interesting at times.

Having a set group which gives a well worn and tested out come 80% of the time for what you wish to achieve is the goto option. 

 

I did not fly at all - the mission is AI only.  Easy enough to add a player plane, but I do not see the point at the moment, at least until I have a baseline for what the AI does alone which I can compare to some historic outcomes.

[Jupp]

~S~ Fellas,

 

Hats off to a great thread! Of late, I learn something new about this simulator and history, .etc, every time I skim the forum.

 

Things are a bit off kilter at the moment. We early adopters also know that 777 does come right back to pleasing most of us.

 

Anyway, thanks to all of you contributors for furthering my (re)-education on several topics, including this very one. Salute!

 

https://www.youtube.com/watch?v=rtrn6-cKqkM

Edited December 9, 2017 by Jupp

[Retnek]

To add some landmarks I've searched some of my books (s.b.) to find numbers. How many AAA-rounds Luftwaffe Flak needed to bring one airplane down?

 

Koch p. 71

Ammo average over all (including any kind of engagement, day and night, barrage shooting etc) until 31.12.1942

4915 rounds light AA (<= 20 mm)

3343 rounds heavy AA (>= 88 mm)

 

So this should be a kind of upper limit for the German AAA.

 

 

v. Renz p. 99

Estimations from shooting range exercises before WW2. Gun type isn't given, but most probably for the 88 mm guns:

"47 rounds to bring down a plane"

 

v. Renz p. 143

Rounds to bring down a heavy bomber, standard time fuse as until end 1944

16000 for 88 mm type 36

8500 for 88 mm type 41

6000 for 105 mm type 39

3000 for 128 mm type 40

 

Rounds to bring down a heavy bomber, time and contact double-fuse late 1944 / 1945

5000 for 88 mm type 36

3000 for 88 mm type 41

2000 for 105 mm type 39

 

v. Renz p. 193

Until 31.12.1942 on average for one shoot-down

5000 rounds light AA (<= 20 mm)

3500 rounds heavy AA (>= 88 mm)

(might be the same source as Koch c.a.)

 

Nov. and Dec. 1943 average per shoot-down

 

6500 rounds light AA (<= 20 mm)

4000 rounds heavy AA (>= 88 mm)

 

Westermann p. 292 ff

"First twenty month of war"

5354 rounds light AA (<= 20 mm)

2805 rounds heavy AA (>= 88 mm)

 

Nov. and Dec. 1943 average per shoot-down

6500 rounds light AA (<= 20 mm)

4000 rounds heavy AA (>= 88 mm)

(same as v. Renz)

 

"Over the entire cource of the war" per shoot-down

4950 rounds light AA (<= 20 mm)

3343 rounds heavy AA (>= 88 mm)

 

I've found some reports for combined large heavy AA-batteries shooting at lone (most probably already damaged) US-bombers during daylight at lower alt (2000 - 5000 m) using ca 500 heavy rounds to finally bring the bomber down. At night with radar aiming something like 1000 heavy rounds per shoot-down seemed to be the optimum possible (Banny, Greve, Hupp, Thiel)

 

Sadly I found no reports about the ammo use of light AAA in single engagements under good conditions.

 

Most problematic for our purposes in IL2-BoX are the light AA-guns. German 20 mm guns are most deadly killing machines when used with standard settings. Here's a table from several sources:

 

                  20mm-type 30    20mm-type 38    37mm-type 36

Rpm                 120                     220                    120

max range(a)   2200                  2200                  3500

max range(b)   2000                  2000                  4000

 

Max. range was given by a self destructing fuse.

 

Rules of engagement to open up for the German 20 mm guns:

incoming 1000 m, passing by 800 m and and departing 600 m

 

Russian sources state an average of ca. 1000 rounds (25 and 37 mm) per shoot-down. No further information on this number, but it's somehow in range with the numbers from German sources.

 

A rough estimation finally: If we divide the German numbers by two because of (night) barrage shooting we're somewhere around 2000 - 3000 rounds. Further we should take into account the shooting training of German light AAA crew had serious problems in contrast to the Soviet crews. Resulting in bad results for green crews. Both v. Renz and Koch complain about it.

 

Later in war the planes were faster, more agile and better protected while the German light guns more or less remained the same. So the numbers of rounds per shoot-down went up because of that, too.

On the other hand the German AA-gun-crews in the rear had comparatively low losses. So after some years of war even with serious training errors a lot of them became very professional AA-gunners.

 

All in all I'd suggest for an realistic simulation of German 20 mm AA an per shoot-down average of

1500 rounds for an ace crew

2500 rounds for a green crew

 

Given a battery of four 20mm-type 38 guns and a plane flying at 300 m height directly to them with 100m/s needs

1000 m incoming = 10 s = 4 x 35 rounds

600 m departing = 6 s = 4 x 22 rounds

 

Ca. 230 rounds for one pass. Ace crews will hit that planes not more than once or twice. Given a few 20mm hits were needed for a shoot-down of a light air-plane the result is a serious damage for one (stupid) direct attack. Green crews might hit once, with luck. Given the 20mm-type 30 with half the fire rate, even a battery of ace-AA-crews won't shoot down that fighter on a second pass.

 

Anyone flying ground sorties in IL2-Box has been shot down by the overdone light AAA VERY MUCH more often than the numbers presented here would allow. Some of the Russian light AAA is even more deadly than the German 20mm are. The gunners of the new MTB are divine even in rough seas ...

 

 

The efficiency of light AAA in IL2-BoX should be scaled down seriously.

 

 

Sources:

 

Banny, Leopold (1994): Dröhnender Himmel - brennendes Land. Der Einsatz der Luftwaffenhelfer in Österrreich 1943-1945. 2. Ausgabe. Unter Mitarbeit von Othmar Tuider: Eigenverlag, Lackenbach.

 

Greve, Friedrich August (2000): Die Luftverteidigung im Abschnitt Wilhelmshaven 1939 - 1945. 2. Marineflakbrigade. 2. Aufl. Jever: Lüers.

 

Hupp, Klaus (1998): Bei der Marineflak zur Verteidigung der Stadt und Festung Kiel im 2. Weltkrieg ein Beitrag zur Kieler Stadtgeschichte. Husum: Husum.

 

Koch, Horst-Adalbert (1954): Flak. Die Geschichte der deutschen Flakartillerie ; 1935 - 1945. Bad Nauheim: Podzun.

 

Renz, Otto Wilhelm von (1960): Deutsche Flug-Abwehr im 20. Jahrhundert Flak-Entwicklung in Vergangenheit und Zukunft. Frankfurt/M.: Mittler.

 

Thiel, Reinhold (1995): Die bremische Flugabwehr im Zweiten Weltkrieg. Bremen: Hauschild.

 

Westermann, Edward B. (2001): Flak German anti-aircraft defenses, 1914 -1945. Lawrence, Kan.: Univ. Press of Kansas (Modern war studies).

Edited December 9, 2017 by 216th_Retnek

[Stig]

OTOH, it's entirely credible for a player controlled ground attacker to destroy a convoy of trucks, or column of tanks all on it's own; or a single bomber to sink a couple of ships, right? 

[unreasonable]

OTOH, it's entirely credible for a player controlled ground attacker to destroy a convoy of trucks, or column of tanks all on it's own; or a single bomber to sink a couple of ships, right? 

 

When I attack ground targets I am lucky to destroy one!

 

That some can get the results you refer to is mostly a function of the fact that some players have very high skills, honed after hundreds of virtual deaths, and can fly without fear of real death, a luxury not available in RL.    If it is due to the damage effects of the player's weapons being incorrect that can be corrected: but hard to see how the game can take away skills some human players actually have.

 

Anyway, hard to see why this should dissuade us from   trying to understand the RL effectiveness of AAA. 

[Eicio]

When I attack ground targets I am lucky to destroy one!

 

That some can get the results you refer to is mostly a function of the fact that some players have very high skills, honed after hundreds of virtual deaths, and can fly without fear of real death, a luxury not available in RL.    If it is due to the damage effects of the player's weapons being incorrect that can be corrected: but hard to see how the game can take away skills some human players actually have.

 

Anyway, hard to see why this should dissuade us from   trying to understand the RL effectiveness of AAA. 

 

We're kind of cylons, we spawn back... 

[Stig]

I'm not trying to dissuade anyone, that would be pointless; there is after all no doubt that the AAA in game is way more accurate than in RL. However, as long as player controlled aircraft individually are causing as much damage as 10, 20 or more aircraft  would in  RL; reducing AAA accuracy markedly, will only make the players even more effective in destroying enemy assets. So what do mission makers do? Put more objects in the missions, defended by more AAA to insure that the mission isn't done and dusted in no time. We know what more objects and AI do for server performance, don't we.  

[unreasonable]

All I want is the option for SP to contain AAA that does roughly RL levels of damage. You know, like a simulation.   As it happens "Low" AI may be close to that already, but I suspend judgement for the moment until I finish my analysis. (Still waiting for source book ).

 

Saying that individual players in MP are causing as much damage as 10 or 20 or more aircraft seems to me to be a gross exaggeration.  If, however, that is really true you would be better off trying to work out what is missing from MP that allows that to happen and working out how to fix it - and it is not AAA.

 

As long as there are a range of settings for AI mission makers can balance simulation against game-play concerns depending on the context.  

 

All I am trying to do is get a handle on what I think the RL simulation levels should be: if there is a "death ray" setting for use in MP I care not, just as you should not care if there is a simulated RL level for use in SP. 

If MP designers use the RL level in MP and you do not like it, that is not my problem and you cannot expect  SP players who want a simulation level to forgo it.

[Stig]

Hey, no problem Unreasonable! I'm not against what you're trying to do and if it improves  the AAA, great!

I agree also that the differences between MP and SP do make for some annoying problems, particurlarly effecting SP.

 

I do still find it a little amusing when players complain about the accuracy of the AI, while the players accuracy is every bit as unreal.

 

 

 

[Gambit21]

Not sure I agree with your assertions or logic.

[hames123]

Unreasonable, how about you make a test like this. Make a convoy of 50 German trucks driving around towards the East. Then, get 48 Il-2s to attack them, until every truck is dead. Have 12 AA trucks, and set them from low to high. Then, count how many Il-2s get back to their air base, and how many are shot down. Repeat this with the Luftwaffe, use Stukas or ME 110s. I think you will find that the Luftwaffe planes are a little more vulnerable to AA, the Il-2 is very tough.

 

Then add in a player controlled flight of 109s or Yaks so we can enjoy the slaughter.

Edited December 10, 2017 by hames123

[unreasonable]

Unreasonable, how about you make a test like this....

 

Why not do it yourself?  

 

What I do intend to do once I have finalized my airfield attack test is to run it using the German 3.7cm Flak 36 which is similar to the 61-K, but setting them to Russian and still using the 109s to attack. This will test whether the AAA is noticeably different for each side.  If you run German AAA vs Soviet fighters instead of 109s the issue of DM differences then comes into play in comparing with the Soviet guns vs German fighters case.

[6./ZG26_Gielow]

Don't be shy Devs. We need the Wirbelwind

 

https://en.m.wikipedia.org/wiki/Wirbelwind

 

I hope we can get 3.7 auto guns later on the first tanks expansion when 20mm proves to be inadequate to deal with aircraft.

Edited December 10, 2017 by 6./ZG26_Gielow

[hames123]

Why not do it yourself?

 

What I do intend to do once I have finalized my airfield attack test is to run it using the German 3.7cm Flak 36 which is similar to the 61-K, but setting them to Russian and still using the 109s to attack. This will test whether the AAA is noticeably different for each side. If you run German AAA vs Soviet fighters instead of 109s the issue of DM differences then comes into play in comparing with the Soviet guns vs German fighters case.

I am afraid that I have tried. And failed epically. You seem to be able to somehow work the mission builder, so I was proposing that as a fair test, since you have to take into account that Il-2s and Stukas attack targets differently, with different exposure to AA. And what hatter way to show this than to have them attack targets en masse?

[Dakpilot]

Wirbelwind/Ostwind are very cool machines, but with a total of 87-105 (wirbelwind) produced, all after mid 44, they would not be that common, Ostwind was first combat tested (single unit) at Ardennes Dec 44 with production totalling about 45.

The earlier Mobelwagen was produced from April 44 total 245 built but was not considered very good due to extreme crew vulnerability (compared to wirbelwind design) as far as I know it seems most of panzer IV based AA were used on the Western front

 

Cheers Dakpilot

[6./ZG26_Gielow]

Who cares?? It is a game.

 

Or are you suggesting that only 35000 players are allowed to buy the 109 on its different versions??

 

Are you trying to broke 1C company??

 

It is just flak.

 

This time line and production numbers madness already made enough damage on this new IL2 series.

Edited December 11, 2017 by 6./ZG26_Gielow

[unreasonable]

I am afraid that I have tried. And failed epically. You seem to be able to somehow work the mission builder, so I was proposing that as a fair test, since you have to take into account that Il-2s and Stukas attack targets differently, with different exposure to AA. And what hatter way to show this than to have them attack targets en masse?

 

I can only use the ME if I keep things relatively simple.     My test bed has many aircraft and waypoints, but all they are doing is flying around in circles, albeit fast and low circles, to simulate a mass airfield strafing attack.

 

If you want to test something specific, you have to be able to alter just one variable at a time in order to see what difference it makes.  As you say, Stukas and Il-2s usually attack in different ways, have very different construction and hence DMs, and would be facing different AAA units.

So it is hard to see what specifically is being tested in your proposal. It is pretty obvious that - other things being equal - Stukas are more vulnerable to AA fire as they are much less armoured.

 

Since Amazon have now told me that I will not be getting my book - which I hope has some good information in it - for at least another week, I am going to run the comparison tests of the AI levels this week.  The main difficulty seems to be counting the shots fired - I think I need to find one of those ticket counters and do it manually.....

[unreasonable]

Preliminary results from Test 1: which is to remind you 48 109 G-4s attacking an airfield in waves, defended by 8 61-Ks.  This is with AAA AI on Low.

 

 

 

 

Hard to say anything about the loss rate in absolute terms without a historical case for comparison given the specific circumstances, but:

 

- The shoot downs per ammo expended ratio is a on the HIGH (edit - I mean fewer shells per shoot down)  side compared to the German data in - previous post, although this is to be expected since a 37mm shell has many times more destructive power than a 20mm, and I have counted all aircraft that were smoking or leaking at the rally point as shot down, since a safe RTB seems unlikely.  Most (75%) aircraft hit crashed before the rally.  Assuming, optimistically, that the smokers do in fact RTB, the shot down ratio is one per approx 660 rounds expended which does not seem outlandish and in the same order of magnitude as per the earlier post by Retnek.

 

- Watching the guns in action you do get a sense of the lag while they pick a target and track it: by the time they have lined it up the target is often over them and flying away.  

Edited December 12, 2017 by unreasonable

[6./ZG26_Custard]

 

 

This is with AAA AI on Low.

 

As mentioned by yourself and others, the AI settings seem to be the significant factor. Just from a personal viewpoint, When flying in MP any setting above normal for AI AA seems just far to accurate.

 

Many thanks for the continued testing that you are doing     

[unreasonable]

Thanks Custard - inspired by your thread and the prospect of actually having a Bodenplatte map where we can model this type of attack in detail!

 

And here is the Normal test - exactly the same except for the AI change.

 

 

As you can see, in terms of kills per ammo used, the Normal is about three times as effective as Low in this test. While Low was ballpark plausible this is not so much.

 

Watching the guns, what is clear is how much more quickly they can change targets: hard to measure the accuracy as such, but the ability to get on target quickly allows the guns to swing onto a plane that has just flown right over them and hit it while it is still close.  The gunners obviously have eyes in the back of their heads..... 

 

The ammo used is pretty robust - I counted shells used for 4 additional guns in different locations for this test, and only one was more than +/- 10 from 240. 

 

Not sure if there is any point in doing the High now....  but open to suggestions.

[6./ZG26_5tuka]

I'd certainly be interested. Given my mission had a way more simple layout (because I'm bad with the ME) I'd be curious to see what your results would be like (especially with german AAA).

 

From my test the 20mm espeically proved to be useless beyong 700m. No lead at all even on High level.

[=EXPEND=Tripwire]

Same, Id like to see Ace as well just to show how overdone it really is.

 

Mission designers need to see that the use of AI settings above low is making the impact much more severe than losses encountered in real life.

Edited December 12, 2017 by =TBAS=Tripwire

[unreasonable]

Alright chaps, if there is some interest I will get it done using High settings tomorrow. Full set of 10 runs (to get a robust average) takes me a couple of hours to do... but for completeness I agree that High has to be done. 

 

Changing the gun types is also doable - we can have German 20mm shooting at 109s!      But I do not fancy changing the planes much - all 48 of them.......

[Retnek]

Thanks for the energy and time, test is best! However the ca. 1000 rounds per shoot-down for the Soviet light AAA were determined, your numbers show there's some need to adapt. Or at least choose the lowest AAA-abilities to stay somehow near realism. Did you limit the engagement range of the AA-guns?

 

  Next to the deadly precision the light AAA usually fires at ranges far beyond the limits gun crews were ordered to use in WW2. If the AAA needs time to detect and adjust it limits the range against incoming planes automatically. Most annoying to me those goblins enjoy to shoot after departing planes up to the absurdly long ranges.

  Being killed over a distance of 2000 m from a 20 mm gun with a series of hits ... one already had adjusted the seat, radio operator tuned in Radio Belgrad, a thermos-flask, the smell of coffee - blam-blam-BLAM - down goes my Pe-2-tank! That's simply mean.

[6./ZG26_5tuka]

Being killed over a distance of 2000 m from a 20 mm gun with a series of hits ... one already had adjusted the seat, radio operator tuned in Radio Belgrad, a thermos-flask, the smell of coffee - blam-blam-BLAM - down goes my Pe-2-tank! That's simply mean.

The Flak 38  does have an effective range of 1000m and won't even fire beyond that regardless of ai skill level.