Mitsubishi Stuka and Thunderbolt

[Pursuivant]

Originally Posted by FC99

Tail coming off is relict of the past when average player's PC were much weaker than what we have now and it is best viewed as crude visual representation of catastrophic damage.

That might explain wings coming off B-17 or B-24 due to damage, too.

Is there any way to better model that catastrophic damage so that very tough planes don't lose parts in unrealistic ways?

Originally Posted by FC99

You are very wrong about human gunners too, they are way better than AI, obviously you don't fly online much.

Then you've got a decent way to calibrate AI gunner skill. Get a bunch of very skilled human gunners in coop, ask them how many hours they have flying online as gunners, then determine their hit percentage against a variety of targets and a variety of deflections.

Use the hit percentage by your very best and most experienced humans as your "Ace" quality gunner standard and adjust AI skill from there.

Whether or not it's historically realistic, synching AI gunner skill to top human skill has the following benefits: a) It means that nobody can bitch about the AI being "better than human", b) means that offline AI gunners will be good training for people who are practicing before they go online. By definition, if you can beat Ace AI, you can do pretty well against human gunners online.

Likewise, if TD feels like revisiting fighter gunnery accuracy (which went from "lasers o' death" prior to 4.11, to just about right in 4.11, to "nerfed" in 4.12) you could base Ace gunnery standards on hit percentages for the very best human players.

[majorfailure]

Originally Posted by Pursuivant

Then you've got a decent way to calibrate AI gunner skill. Get a bunch of very skilled human gunners in coop, ask them how many hours they have flying online as gunners, then determine their hit percentage against a variety of targets and a variety of deflections.

Use the hit percentage by your very best and most experienced humans as your "Ace" quality gunner standard and adjust AI skill from there.

Whether or not it's historically realistic, synching AI gunner skill to top human skill has the following benefits: a) It means that nobody can bitch about the AI being "better than human", b) means that offline AI gunners will be good training for people who are practicing before they go online. By definition, if you can beat Ace AI, you can do pretty well against human gunners online.

Likewise, if TD feels like revisiting fighter gunnery accuracy (which went from "lasers o' death" prior to 4.11, to just about right in 4.11, to "nerfed" in 4.12) you could base Ace gunnery standards on hit percentages for the very best human players.

While your idea has its merits, I fear it would lead to gunners close to 4.09 standards. I do think gunnery in IL2 is easier than it was in real life, no vibrations, no force needed pulling the gun anywhere, extensive training possible without any risk, no recoil etc.
I still like the gunners the way they are now.

[jameson]

The laser sighted, 40mm, power assisted, zero-g, cannon operator in the back of the IL-2 appears to have strangely missed any "improvement". A de-winged IL-2 tumbling and spiralling towards the deck still has the "terminator" firing his weapon as he hits the water/ground. Any human would have been, lol, over the side somewhat earlier I feel. Perhaps this could be addressed?

[Treetop64]

Originally Posted by jameson

The laser sighted, 40mm, power assisted, zero-g, cannon operator in the back of the IL-2 appears to have strangely missed any "improvement". A de-winged IL-2 tumbling and spiralling towards the deck still has the "terminator" firing his weapon as he hits the water/ground. Any human would have been, lol, over the side somewhat earlier I feel. Perhaps this could be addressed?

Inertia dampers. You forgot to mention the inertia dampers.

[horseback]

Originally Posted by Pursuivant

Originally Posted by FC99:
Tail coming off is relict of the past when average player's PC were much weaker than what we have now and it is best viewed as crude visual representation of catastrophic damage

That might explain wings coming off B-17 or B-24 due to damage, too.

Is there any way to better model that catastrophic damage so that very tough planes don't lose parts in unrealistic ways?

Originally Posted by FC99:
You are very wrong about human gunners too, they are way better than AI, obviously you don't fly online much

Then you've got a decent way to calibrate AI gunner skill. Get a bunch of very skilled human gunners in coop, ask them how many hours they have flying online as gunners, then determine their hit percentage against a variety of targets and a variety of deflections.

Use the hit percentage by your very best and most experienced humans as your "Ace" quality gunner standard and adjust AI skill from there.

Whether or not it's historically realistic, synching AI gunner skill to top human skill has the following benefits: a) It means that nobody can bitch about the AI being "better than human", b) means that offline AI gunners will be good training for people who are practicing before they go online. By definition, if you can beat Ace AI, you can do pretty well against human gunners online.

Likewise, if TD feels like revisiting fighter gunnery accuracy (which went from "lasers o' death" prior to 4.11, to just about right in 4.11, to "nerfed" in 4.12) you could base Ace gunnery standards on hit percentages for the very best human players.

There seems to be an ongoing campaign to conflate my comparison of the off-line ai gunners to the human beings that actually manned the guns of WWII bombers, attack aircraft and heavy fighters with the in-game mouse aimed guns.

Two entirely different things. Both criminally bogus of course, but two different things.

The in-game player's mouse gunner model is vastly simpler and less complicated than the operation and aiming of machine guns from a constantly bobbing and rolling gun platform like an actual moving aircraft of that era. You are on a rail smooth, predictable platform and you can easily control your guns; no engine vibration, no jammed or sticky rings or turrets, no gunshake or recoil making that three-to-six round burst scatter across a two or three degree range, and only an occasional (and buttery smooth) change in direction or angle of your platform to potentially spoil your aim.

This differs very little from the all-ai aircraft gunners offline model, except that they enjoy absolutely perfect awareness of their human target's range, speed and direction; they know precisely how fast they are going, they know how fast you are going and to the millimeter how far away you are and where you will be when they fire their guns at ranges well beyond the average player's convergence ranges. They can perfectly compensate for their 'aircraft' turning, banking and diving. And they consistently manage to hit critical components of target (Player) aircraft moving at high speeds from ridiculous angles in microsecond wide firing windows, and they still seem to victimize some aircraft types more consistently than others.

None of that compares remotely with the actual capabilities of the real-life gunners on WWII era aircraft. For the offline fighter campaigner the difference is critical. The 8th Air Force awarded the title of 'ace' to over 300 bomber crew gunners; I would be amazed if any two of them actually destroyed a combined total of five enemy aircraft in flight, and the late war US bomber defenses were the heaviest and most sophisticated of the war. Their gunners were arguably the most extensively trained of the war. If their efforts were so futile, what does that say about the gunners on the lightly armed, less stable types that everyone else fielded?

cheers

horseback

[MiloMorai]

Testing done by the USAAF found that the bullet pattern from ground testing had the following results for 12 rounds to 600yds:

For the B-17:

ball turret > dia. 15' - 8.3mils
upper turret > dia. 21' - 11.7mils
chin turret > dia. 23' - 12.6 mils
waist(closed) dia. 26' - 14.3mils
side nose > dia. 34' - 18.7mils
tail turret > dia 45' - 25mils

For the B-24:

ball turret > dia. 15' - 8.3mils
upper turret > dia. 20' - 11.2mils
nose turret > dia. 23' - 12.9mils (Emerson)
nose turret > dia. 35' - 19.3mils (Motor Prod.)
waist(closed) dia. 23' - 12.9mils
waist(open) dia. 63' - 35.6mils
tail turret > dia 35' - 19.3mils

source: "Gunner" ISBN 1-55046-332-2

This should be easy enough to replicate in game.

[Pursuivant]

Originally Posted by horseback

The in-game player's mouse gunner model is vastly simpler and less complicated than the operation and aiming of machine guns from a constantly bobbing and rolling gun platform like an actual moving aircraft of that era.

At least when AI is at the controls, I find it difficult to aim accurately because the pilot will perform maneuvers without telling me that he's going to do so. Online, it's likely to be easier because the pilot can actually talk to the crew.

But, like you said, there are lots of things that gunnery doesn't currently model in IL2, which makes it easier to be a gunner, relatively speaking, than it was in real life.

So, I don't see your concerns (which are legitimate) and mine as being incompatible. IL2 online gunnery IS too easy for all the reasons we've mentioned before - plane vibration, turbulence, gun vibration, sticky scarf rings or turret rings, G-forces, physical labor and inertia of slewing the guns around (at least by hand) and, of course, slipstream effects.

I think that these effects would all be pretty easy to model just by incorporating a bit more randomness into the bullet dispersal pattern for gunners under various conditions and by building a bit of variable turn speed and randomness into the mouse movement model.

Things that increase bullet dispersal - each shot after the first in a burst, turbulence (synched to weather/wind, although it is possible to build turbulence into the game), G-forces, slipstream/wind buffeting - at least 10 degrees angle off from (plane's vector - 180 degrees), hand-turned guns.

Things that reduce turning speed of turret/Scarff ring/pintle-mounted guns - G-forces, slipstream/wind-buffeting - at least 10 degrees angle off from (plane's vector - 180 degrees), inertia (modeled as a bit of initial slowness in getting the guns to track if they're not already in motion in the directions you want to track, greater inertia for larger or multiple guns due to mass).

Plus, you automatically build a tiny bit of randomness into mouse tracking movement to represent stickiness and "Murphy's Law."

If TD were kind enough to include all those problems into the human-controlled gunnery model, after the shrieks of outrage fade to whimpers of grudging acceptance, THEN you calibrate maximum human skill to get maximum AI skill for gunners.

Of course, as with any option of this sort, there should be a button to turn it all off, so people who can't cope with the aiming problems that real gunners faced can still have their simplified gunnery model.

If TD wanted to be extra nice to us, they could model the effects of injury to gunners' limbs. A hit to the arm means that you have lots of trouble turning and shooting hand-turned, hand-triggered guns. A hit to the leg means that you can't turn foot-operated turrets in a particular direction. And, of course, bleeding means that gunners will eventually bleed out, getting weaker and less accurate until they fall unconscious or die.

Originally Posted by horseback

The 8th Air Force awarded the title of 'ace' to over 300 bomber crew gunners; I would be amazed if any two of them actually destroyed a combined total of five enemy aircraft in flight, and the late war US bomber defenses were the heaviest and most sophisticated of the war.

This might be a bit too extreme. Some planes flying solo really did shoot down multiple enemy aircraft - or at least damage them seriously enough that they were "probables" and out of the action. So, aerial gunners weren't completely useless - especially the tail gunners who accounted for a majority of the 8th Air Force gunner aces. It's also worth mentioning that the USAAF kept tailgunners long after they ditched every other gunner position (last tailgunner kill was over Vietnam).

But, as you said, the USAAF (and every other Air Force) had problems with overclaiming kills. Often, when some hapless Bf-109 diving through a formation B-17 or B-24 coughed up smoke because pilot mishandled the fuel mixture, every gunner in the formation would claim it as a kill because they saw the 109 coughing up black smoke were sure that their gun was the one that "hit." With claims like that, even the most skeptical debriefing intelligence officer was likely to believe that the fighter was a "probable" even if the Luftwaffe plane wasn't scratched.

[Pursuivant]

Originally Posted by MiloMorai

Testing done by the USAAF found that the bullet pattern from ground testing had the following results for 12 rounds to 600yds:

This is very good information, the sort of quantitative data that's hard to come by. Thank you for posting it. Since the numbers for the gun positions on the two different bombers were reasonably close, the data could possibly be used for similar weapons mounted on other U.S. bombers, like the B-25 or the A-20.

[horseback]

Originally Posted by Pursuivant

This is very good information, the sort of quantitative data that's hard to come by. Thank you for posting it. Since the numbers for the gun positions on the two different bombers were reasonably close, the data could possibly be used for similar weapons mounted on other U.S. bombers, like the B-25 or the A-20.

The data for the open waist positions is the most relevant to the majority of man-served guns; it was a pintle-type mount rather than one in a reinforced ball socket mount, similar to the majority of the rear gunners in the sim. Scharf ring and pintle mounts are much more subject to vibration and flex.

Note also that the data is for ground testing, which means that there is no relative motion to joggle the gunner's elbow or guesstimations about where the target was going to be when the bullets got there. Chances are good that the guns were sighted in and then clamped down and fired by a fixed remote to get those figures in order to eliminate human error.

cheers

horseback

[horseback]

This might be a bit too extreme. Some planes flying solo really did shoot down multiple enemy aircraft - or at least damage them seriously enough that they were "probables" and out of the action. So, aerial gunners weren't completely useless - especially the tail gunners who accounted for a majority of the 8th Air Force gunner aces. It's also worth mentioning that the USAAF kept tailgunners long after they ditched every other gunner position (last tailgunner kill was over Vietnam).

Most of the stories we hear about a solo aircraft's gunners managing to destroy or damage attacking single engine fighters usually turns out to be apocryphal if we try to investigate; about 20 years ago, I was commissioned to build a 1/72nd scale B-24J with the markings of a Foggia based aircraft that was shot down in the sort of circumstances you describe for one of the surviving aircrewmen. I was invited to the presentation, and the honoree confessed to all present that the claims of four or five enemy fighters destroyed in their heroic last fight (over Turin, I think) were all bulls**t (his word, used as he pointed right at the Groups' former Public Information Officer), but he wasn't giving his Air Medal back.

As for the B-52's stinger, the 'gunner' operated a radar aimed gun remotely, with the help of a slightly more advanced stabilization system than that used on late-WWII era battleships's guns. It took shameless advantage of the limited range and acquisition cone of early Warsaw Pact heat guided missiles like the Atoll.

But, as you said, the USAAF (and every other Air Force) had problems with overclaiming kills. Often, when some hapless Bf-109 diving through a formation B-17 or B-24 coughed up smoke because pilot mishandled the fuel mixture, every gunner in the formation would claim it as a kill because they saw the 109 coughing up black smoke were sure that their gun was the one that "hit." With claims like that, even the most skeptical debriefing intelligence officer was likely to believe that the fighter was a "probable" even if the Luftwaffe plane wasn't scratched.

Almost any Allied fighter pilot in the theater for more than two days (i.e., long enough to visit the Officers' Club bar) could have told them that German fighters tended to belch black smoke whenever the throttle was shoved forward or back too quickly, and the Allies' commanders were well aware that the Luftwaffe wasn't taking even a tiny fraction of the losses the the gunners claimed to be inflicting. "Morale" was the only justification for awarding the overwhelming majority of gunners' claims for destroyed enemy aircraft, and they beat it to death.

cheers

horseback