Gonna need explanation

[zelil]

You have to change the AAA mechanic in this game this BS shoot someone down then seconds later get drilled by AAA has to go. I understand the vulching aspect it but were no were near the enemy base, in fact were over our lines when being shot at.

Why would any bomber or plane want to attack the objective when the second you do you get annihilated. And another thing I would like to know why certain players never get hit. I wont name names but 20 passes and AAA is not even firing and the second another enemy gets close that players gets knocked out in seconds once he gets close. Just makes me real suspect.

[curiousGamblerr]

Couldn't possibly be your tactics, right? Noooo waaay

[6./ZG26_Klaus_Mann]

Explanations here. 

 

https://forum.il2sturmovik.com/topic/32578-bravery-being-out-range/

 

Especially the 37mm FlaK Guns are in God Mode once set above Low Skill Setting. In Fact setting Flak at anything but Low creates virtual No Fly Zones with Laser Guided Guns. 

And even in Low Skill they are twice as effective as RL.

Edited January 1, 2018 by 6./ZG26_Klaus-Mann

[Field-Ops]

Naturally you would be easier to hit by ground units as you get closer to the, you guessed it, ground. So maybe staying at 1k+ altitude despite your target being that low might be a better option than staying on their tail when they are lawn mowing the trees? Just some food for thought. And being a ground pounder isnt much different. They arent going to go easy on you just because your job is to be up close and personal. Be unpredictable in your turns and dont get too close, use other teammates in the area as bait or have heavy bombers soften up the target before your entry. If you want to be the lone guy going up against hordes of AAA dont expect to get home.

[NETSCAPE]

I've tested the 3 AI settings for all AA/AAA extensively. 

 

Most of the time "normal" feels too op for anything 20-40mm. But "low" is pathetic... And "high" is insanely laser guided accuracy unless in the 80-90mm weapon range...

 

I tend to do a mix of all 3, depending on what AA and effect I want. 

 

About the AA/AAA not engaging. The AI is already set, tracking a target, or they are reloading, or they are taking time to switch to another target (possibly making it seem like they "aren't shooting that one guy, but are shooting me".

 

I flew a He 111 sortie on wings the other day where the FIRST flak round blew off my wing at 4000 altitude....yeah there's that laser accuracy for ya... 

 

So it's up to the mission designer. In a way you are both right. But I have to say I am getting tired of the "change your tactics, git gud" type of remark. They usually come from fighter pilots  . Us ground pounders that do use the proper tactics will still get ANNIHILATED if all the AA/AAA AI settings are on "high" . 

[[CPT]CptJackSparrow]

SEAD strikes in il2's or 110's prior to strike package reaching IP?

[curiousGamblerr]

SEAD strikes in il2's or 110's prior to strike package reaching IP?

You ever see someone make a post like this that actually takes the time for some pre strike wild weasel? No chance

[[CPT]CptJackSparrow]

Hope springs eternal =)

[Lusekofte]

We tend to do just fine 80% of the time, and in this persenage if you get shot down it is when you fly away from the target, and is stupid enough to climb while doing it. If you stay at ground level and do not fly in straight manner, you are good, the next 20% well you are doomed. All of my squad members have had a very short Friday night bomber flight withing those 20%. And it feels more like 60% the times you get shot down.

We normally fly IL 2 and PE 2, so we are right there ready to get shot. But looking back it is only 20% of the times you get hit in serious servers and FNBF. 

It is very frustrating, and it tend to happened to the same person in the squad, if I am unlucky once , you can bet I am the one going down next time too

[=TBAS=Sshadow14]

Let Devs Lock AAA To Historic Accuracy

Wanna have it unlocked goto special
"Un historic Servers"

[1./KG4_OldJames]

You ever see someone make a post like this that actually takes the time for some pre strike wild weasel? No chance

...erm, I try to.

 

Its the same old story: teamwork, teamwork, teamwork. Fly alone near flak, and you are a gonner. you need one person to fly JUST within range of the aaa and draw the fire which also shows the other team members where the aaa is and allows them to eradicate it. Also, to prevent quick respawning of aaa (on Finnish server at least) destroy depots and factories that supply the areas of aaa.

 

Sometimes solo is the only way to fly, but you have to swallow the fact that your survival rate is going to be a lot lower.

[=ARTOA=Bombenleger]

So AA is doing whats its designed to do? I dont see the problem.
Except for it shooting through clouds and forests, that is ineed something worthy of improvement.
Unfortunately it would need a huge amount of work and ingame performance would decrease radicaly if that was to be corrected, so its probably not going to happen.

[PatrickAWlson]

I've tested the 3 AI settings for all AA/AAA extensively. 

 

Most of the time "normal" feels too op for anything 20-40mm. But "low" is pathetic... And "high" is insanely laser guided accuracy unless in the 80-90mm weapon range...

 

I tend to do a mix of all 3, depending on what AA and effect I want. 

 

About the AA/AAA not engaging. The AI is already set, tracking a target, or they are reloading, or they are taking time to switch to another target (possibly making it seem like they "aren't shooting that one guy, but are shooting me".

 

I flew a He 111 sortie on wings the other day where the FIRST flak round blew off my wing at 4000 altitude....yeah there's that laser accuracy for ya... 

 

So it's up to the mission designer. In a way you are both right. But I have to say I am getting tired of the "change your tactics, git gud" type of remark. They usually come from fighter pilots  . Us ground pounders that do use the proper tactics will still get ANNIHILATED if all the AA/AAA AI settings are on "high" . 

 

Mostly agree except that LOW is pathetic. IMHO LOW is pretty realistic.  Actually I agree with Klaus Mann: LOW = twice as good as real life.

 

It just wasn't that deadly on a per mission basis. What made it so deadly for real pilots was that they had to do it dozens or hundreds of times.

 

In my SP missions with Flak = low I get hit all the time.  I have been blown out of the air a couple of times in dozens of missions.  If anything that is more deadly than RL.  

 

What I do think is different is SP campaign vs MP environment.  In a campaign you want a high degree of realism over time.  In MP you quite often want to stop unrealistic behaviors like vulching.  The former requires lower accuracy otherwise a campaign is not survivable.  The latter requires higher to make people behave online.  My best suggestion is campaign writers always use LOW.  MP mission developers should use LOW except around critical targets like airfields where MED might be better.  Leave HIGH alone.  The bright side is that the tools are in the box.  Just use the right ones.

 

I do admit to an ulterior motive: as an SP player I do not want LOW flak to get any more accurate just to cater to the wishes of MP players.  If any adjustments are made I want low to be worse, medium to be today's low and high to be today's medium.  It should be possible to meet both needs.

Edited January 1, 2018 by PatrickAWlson

[unreasonable]

Let Devs Lock AAA To Historic Accuracy

 

Wanna have it unlocked goto special

"Un historic Servers"

 

That would be my goal too, but one of the difficulties is knowing what "historic accuracy" actually was.  We can get a rough idea of rounds expended per shootdown, but the historic data is usually aggregated.  So  even if you are just looking at light AA - say up to 40mm - the numbers we get in books typically mix shots fired against low flying ground attack by tactical AA units with shots fired in barrages against level bombers. 

 

My own view is that even on Low, the AA is somewhat too effective - I disagree with Netscape on this. That is after doing my own tests, documented in the thread Klaus-Mann linked.  

 

Without specific details on a reasonably large number of actual RL engagements, however, it is hard to be sure, since there is a large amount of variation in the outcomes when you run the same test several times. Getting the detail we need for specific RL instances is quite hard, for instance how many AA guns were there in some particular engagement, although I now have some information on the Bodenplatte raids that I will post once I have summarized it.

[6./ZG26_Klaus_Mann]

Well, considering that Loss Rates of 5% per Flight were already Catastrophic, while we regularly loose 50% of a Force on the First Pass, no matter the Speed, if 61-K Guns are stationed there. 

And when the Fighters dragging Flak, get shot down at 750kph, well, something aint right. 

[[DBS]Tx_Tip]

OP is discussing MP flak.

MP flak logic is different than SP flak logic. Flak logic versus AI is different than flak logic versus players in SP and then again different versus players in MP.

The testing that unreasonables'  did, while interesting and served up good information, was versus AI only. Introducing a player into the mix would of resulted in different outcomes as I mentioned in that thread.

What we don't need as Mission Designers is the lack of options to dictate how, when, where and why players and AI get shot down. Having the Devs limiting our options in regards to this matter would lessen mission design, not enhance it.  

The strength and accuracy of flak can be adjusted using the many variables we now have within the ME to the ninth degree.

[7.GShAP/Silas]

OP is discussing MP flak.

MP flak logic is different than SP flak logic. Flak logic versus AI is different than flak logic versus players in SP and then again different versus players in MP.

 

 

 

How so?

[unreasonable]

OP is discussing MP flak.

MP flak logic is different than SP flak logic. Flak logic versus AI is different than flak logic versus players in SP and then again different versus players in MP.

The testing that unreasonables'  did, while interesting and served up good information, was versus AI only. Introducing a player into the mix would of resulted in different outcomes as I mentioned in that thread.

What we don't need as Mission Designers is the lack of options to dictate how, when, where and why players and AI get shot down. Having the Devs limiting our options in regards to this matter would lessen mission design, not enhance it.  

The strength and accuracy of flak can be adjusted using the many variables we now have within the ME to the ninth degree.

 

AFAIK the ME has the same logic for AA whether it is MP or SP. If this is incorrect please point out which boxes in the ME are specific to MP vs SP and how the AI flak logic is different.

 

edit - just to note I am certainly not against options. I would rather more options or a modding ability. At the moment the minimum flak option is probably too effective. 

Edited January 1, 2018 by unreasonable

[coconut]

 

 

AFAIK the ME has the same logic for AA whether it is MP or SP. If this is incorrect please point out which boxes in the ME are specific to MP vs SP and how the AI flak logic is different.

 

Just because the interface is the same does not mean the machinery behind it is the same. I don't know that the logic is different, but I would not be surprised if in MP an AAA refrained from engaging a plane until the server has sent info about the existence and status of that AAA to the client being targeted. It would not be fair if you could get shot by things that were invisible to you. Again, I don't know that anything like that is happening, just speculating to give you an example.

[6./ZG26_Custard]

I genuinely hope that flak AI is something the devs will look into for future updates. I hear the arguments that better tactics need to be used etc, but it seems fairly clear from those individuals that have taken time to test the effectiveness of the AAA (in game) bears little resemblance to real world data.  

[Requiem]

Mostly agree except that LOW is pathetic. IMHO LOW is pretty realistic.  Actually I agree with Klaus Mann: LOW = twice as good as real life.

 

It just wasn't that deadly on a per mission basis. What made it so deadly for real pilots was that they had to do it dozens or hundreds of times.

 

In my SP missions with Flak = low I get hit all the time.  I have been blown out of the air a couple of times in dozens of missions.  If anything that is more deadly than RL.  

 

What I do think is different is SP campaign vs MP environment.  In a campaign you want a high degree of realism over time.  In MP you quite often want to stop unrealistic behaviors like vulching.  The former requires lower accuracy otherwise a campaign is not survivable.  The latter requires higher to make people behave online.  My best suggestion is campaign writers always use LOW.  MP mission developers should use LOW except around critical targets like airfields where MED might be better.  Leave HIGH alone.  The bright side is that the tools are in the box.  Just use the right ones.

 

I do admit to an ulterior motive: as an SP player I do not want LOW flak to get any more accurate just to cater to the wishes of MP players.  If any adjustments are made I want low to be worse, medium to be today's low and high to be today's medium.  It should be possible to meet both needs.

I preferred using Low AI as well but with a caveat. One of the things I liked doing when making missions in RoF was to only use low AI for the AA, but compensate by adding the 10x script to it so it fired 10 times as fast. I thought it reflected overall accuracy and the "pucker factor" pretty well too. Nothing like seeing a mass of flak bursts start appearing around you and your squadron as they approached the target

[56RAF_Roblex]

Why would any bomber or plane want to attack the objective when the second you do you get annihilated.

 

There you hit the nail on the head.  We *do* attack the objective because we *don't* get annihilated unless we are careless.   Sure I rarely attack an airfield single handed but I have and I have got home.   In real life it was acknowledged that it was suicide to attack a heavily defended target at all even with buddies and when they had to attack anyway they made one fast attack then ran away jinking like hell.

 

There is nothing wrong with the AA.  Learn to deal with it.

Edited January 1, 2018 by 56RAF_Roblex

[Oubaas]

Having never been shot down by AAA in IL-2 BoX, it's difficult for me to comment. However, I suspect that my experiences may have something to do with the fact that I try to spot the AAA as early as possible and then take it out. After that, I carry on with other targets. It seems to work, they haven't got me yet.

[6./ZG26_Klaus_Mann]

Well, try attacking a Tank Column on TAW with a Stuka, 110E or 129, while Fighters are providing Anti-Fighter Support. In a Formation of 4 129s they were able to Kill 2 of us before we fired the first Shot and only 1 got out alive, limping back home on 1 Engine. And that's a regular Occurence. And we are a good at what we do. 

That's on a day with Overcast at 2k. 

Edited January 1, 2018 by 6./ZG26_Klaus-Mann

[7.GShAP/Silas]

It's like being on high seas, getting tossed around.  That part is actually pretty fun.  The space lasers are not.  Gamey tactics win the day under those conditions.

[Gambit21]

Mostly agree except that LOW is pathetic. IMHO LOW is pretty realistic.  Actually I agree with Klaus Mann: LOW = twice as good as real life.

 

 

My own testing/building supports this.

[CUJO_1970]

There is nothing wrong with the AA.  Learn to deal with it.

 

[Oubaas]

In real life the majority of gunners couldn't hit a bull in the butt with a bass fiddle, such are their marksmanship skills. But every now and then you find one who can hit what he's shooting at with a fair degree of consistency. The AI gunners should reflect this by missing a lot and scaring the crap out of you now and then.

[216th_Jordan]

In real life the majority of gunners couldn't hit a bull in the butt with a bass fiddle, such are their marksmanship skills. But every now and then you find one who can hit what he's shooting at with a fair degree of consistency. The AI gunners should reflect this by missing a lot and scaring the crap out of you now and then.

 

Yep, setting "Normal" as it is now could maybe reflect an ACE gunner. But even then it would have to be used seldomly.

You really have to game the game here to succeed against the AA and thats pretty lame IMO.

 

I do not know what should be different about MP or SP. The only thing I could see here is that the MP gunners shoot at a position the netcode tells them, due to the nature of communications and delays or lags the plane can appear still or warping which influences accuracy.

On some occasions the server was lagging as I was attacking some targets, I dove in fast wiggled around dropped the bombs etc. I saw that the flak was shooting at a defined point in the sky that was not moving, all the guns were shooting there, it was well behind me. All of a sudden my complete plane exploded, desintegrated and everybody died. For the server I was at that still point all the time and a sitting duck while I was doing an actual 600 kph dive. This certainly also needs some love.

Edited January 1, 2018 by 216th_Jordan

[LLv24_SukkaVR]

Best strategy is to have a friend who baits the AAA to shoot at him, while you can engage the AAA and get rid of it. Attacking them alone is just suicidal.

[CUJO_1970]

As I've said before -

 

The AAA in this sim behaves exactly like what it is - computer controlled aim-bots.

 

The challenge is to program it not to behave like what it actually is.

 

Program it to behave like what it is supposed to represent - WW2 era anti-aircraft guns controlled by hand cranks for traverse and elevation - looking through iron sights and guided by a guy with binoculars.

[Gambit21]

In real life the majority of gunners couldn't hit a bull in the butt with a bass fiddle, such are their marksmanship skills. But every now and then you find one who can hit what he's shooting at with a fair degree of consistency. The AI gunners should reflect this by missing a lot and scaring the crap out of you now and then.

"Low" setting basically does this - probably close to a real life ace gunner.

[71st_AH_Hooves]

You ever see someone make a post like this that actually takes the time for some pre strike wild weasel? No chance

Both the 71st and TWB do this nearly everytime we go after a groind target. Maybe join a squadron?

[curiousGamblerr]

Both the 71st and TWB do this nearly everytime we go after a groind target. Maybe join a squadron?

Same with 19// when we have the numbers and are feeling serious. It's the best way to survive.

 

Maybe I reacted too strongly to this post, I don't mean to imply the AAA is perfect in this game. But the title and tone of the initial post rubbed me the wrong way and sounds like a lot of whining more than the reasonable discussion that followed it here.

[NETSCAPE]

Mostly agree except that LOW is pathetic. IMHO LOW is pretty realistic.  Actually I agree with Klaus Mann: LOW = twice as good as real life.

 

It just wasn't that deadly on a per mission basis. What made it so deadly for real pilots was that they had to do it dozens or hundreds of times.

 

In my SP missions with Flak = low I get hit all the time.  I have been blown out of the air a couple of times in dozens of missions.  If anything that is more deadly than RL.  

 

What I do think is different is SP campaign vs MP environment.  In a campaign you want a high degree of realism over time.  In MP you quite often want to stop unrealistic behaviors like vulching.  The former requires lower accuracy otherwise a campaign is not survivable.  The latter requires higher to make people behave online.  My best suggestion is campaign writers always use LOW.  MP mission developers should use LOW except around critical targets like airfields where MED might be better.  Leave HIGH alone.  The bright side is that the tools are in the box.  Just use the right ones.

 

I do admit to an ulterior motive: as an SP player I do not want LOW flak to get any more accurate just to cater to the wishes of MP players.  If any adjustments are made I want low to be worse, medium to be today's low and high to be today's medium.  It should be possible to meet both needs.

 

My default for single player is this:

 

the 80mm+ stuff = NORMAL by default, with a few set to high and low for desired "spread" effect

everything else = LOW by default, with a few set to normal

 

Granted my 100s of hours in the ME are mostly dealing with the player flying the Ju 87 and Bf 110. 

[ShamrockOneFive]

I'm still reminded of a gripping story I read about a Typhoon or Tempest attack on an airfield in 1945. Flak defenses were already ready as the group attacked almost line abreast across the airfield. At the start of the attack 8 aircraft went in and on the way out only two remained. It was a brutal demonstration of how deadly flak and low altitude flying could be.

 

Mind you, this base was defended by several times more guns than we typically see (for CPU/frame rate reasons). I'm ok with reasonably accurate flak although sometimes they seem to get shots that seem unlikely.

 

I wouldn't mind there being more nuance between the skill levels for example.

[Lusekofte]

Forgive me for bringing in another brand here. But it is difficult creating historical flak. Have anyone here read books and articles about strafing hard defended bases. Loss rate was pretty bad every time. Mostly because the AA and AAA was well hid and protected. Japanese gunners did not use tracers on bases in Guinea and was impossible to see at all. If 3 or 4 PE 2 attack a artillerybase with heavy AAA and AA , it is no everytime any of these get hit, In Cliffs over Dover I have flown in circle over heavy defended cities and airstrips at a few 100 meters and not gotten hit once, at the same map I have flown at 5 K with a Heinkel and got my wings blown up.

It is not historical not being afraid of the Flak either. It is all about simulating a chance to getting shot. It is not giving real soldiers Vodka so they get less accurate. What we talk about here is something all brands struggle with, and my guess the developers will try to improve it. But they will not promise anything , because there is no answer to give if their fix will not do it.

I get a bit amused when people claim historical inaccuracy while strafing enemy bases alone or just two, in real life those would be dead meat on second pass.

[unreasonable]

There you hit the nail on the head.  We *do* attack the objective because we *don't* get annihilated unless we are careless.   Sure I rarely attack an airfield single handed but I have and I have got home.   In real life it was acknowledged that it was suicide to attack a heavily defended target at all even with buddies and when they had to attack anyway they made one fast attack then ran away jinking like hell.

 

There is nothing wrong with the AA.  Learn to deal with it.

 

People keep repeating this "it was suicide" business - what does it even mean? What loss rate is "suicidal"? 100%? More than 50%?  Or just 2.5% - which gives a chance of surviving 30 sorties of under 50%?

 

What it was was dangerous, no doubt about that, RL pilots only had one life. The question is how dangerous. The empirical evidence is that the AI setting makes a great deal of difference and at anything above "Low" gets results that are much better than historical research suggests is likely on a one gun = one gun basis.

 

 

Just because the interface is the same does not mean the machinery behind it is the same. I don't know that the logic is different, but I would not be surprised if in MP an AAA refrained from engaging a plane until the server has sent info about the existence and status of that AAA to the client being targeted. It would not be fair if you could get shot by things that were invisible to you. Again, I don't know that anything like that is happening, just speculating to give you an example.

 

True in principle - but I can only test what is specifically proposed.  For instance I could - and will, since the topic is still generating interest - test whether the AA seems to target the player plane in an SP mission more than that of the AI wingmen.  I know we have had trouble with that in other contexts.

Obviously I cannot test MP, but I would be very surprised if the core inputs that determine whether or not a gun hits - eg target prediction, rof, group size etc are any different. Why would they be?

 

 

<snip>

I get a bit amused when people claim historical inaccuracy while strafing enemy bases alone or just two, in real life those would be dead meat on second pass.

 

I agree with your post for the most part - only want to point out that in BoX (and sims generally) the attackers would often be dead on the first pass, since one of the things that makes game AA more effective than it's RL equivalent is that it is never surprised.  Actually I could live with "Low" as a plausible representation if there was a mechanic that could model surprise to a degree, although I am not sure how this could be done in the game since in RL it is very much dependent on the specific context. 

[56RAF_Roblex]

The empirical evidence is that the AI setting makes a great deal of difference and at anything above "Low" gets results that are much better than historical research suggests is likely on a one gun = one gun basis.

 

 

I don't think you can use that argument as I have seen more than one mission maker say that they could not put in the correct number of guns as it affects frame-rate badly so they user fewer guns with more accuracy per gun.  That actually helps the attacker because if 6 guns have the firepower of 12 then each gun you destroy takes away the firepower of two.

 

 

 

Actually I could live with "Low" as a plausible representation if there was a mechanic that could model surprise to a degree, although I am not sure how this could be done in the game since in RL it is very much dependent on the specific context. 

 

 

I have often wished this myself.  Give the gunners a 'Sleep' state they go into when they have not seen an enemy for 15 minutes and have them take 1 minute to 'wake up' when they see an enemy.  This might allow people to make a surprise attack then get the hell out after one pass.   If one minute to wake up seems excessive, bear in mind that the AA see people a long way out and even behind trees and hills.   It is a shame the AA could not be stopped from seeing through obstacles.  I wonder if they could be coded to say they cannot see aircraft below 50m unless they are within 500m of the gun.  Maybe a simple limitation on how low they can depress the guns would help?

 

The other side of the coin though is that the gunners do sleep when the sun is down even when being attacked :-)    I get that in the middle of the night the heavy ack should not be hitting level bombers at 3000m but surely just before dawn they should be at least trying to shoot back at low level attackers.

Edited January 2, 2018 by 56RAF_Roblex

[=RvE=Windmills]

 

 

I don't think you can use that argument as I have seen more than one mission maker say that they could not put in the correct number of guns as it affects frame-rate badly so they user fewer guns with more accuracy per gun. That actually helps the attacker because if 6 guns have the firepower of 12 then each gun you destroy takes away the firepower of two.

 

iirc the less guns with more accuracy to save fps was also the case in 1946, this solution might be carried over into this game.