"Un-hittable Gun" Game Bug?

[Sergei]

Originally posted by Schoerner:

This is fundamental ballistics and no matter if CM uses simpler calculations, what counts is, that the effect showed by CM, can be understood and explained in reality.

And "unhittable" units can be explained very well.

Aha. Well, if you're into "explaining", then sure. I suppose you like playing games in which the units try shooting at berms while they could as well hit the gun or the house itself, but they don't because they don't realise they are hitting the berm.

I just am more into having it fixed because it's unrealistic. Your explanation just doesn't apply, not in the game anyway.

[Schoerner]

If you would have read my explanation, you would know, that we should differentiate between the TacAI, that keeps shooting and the thread's title, that "unhittable guns" are a bug per se.

Two different pair of shoes.

[Industrializer]

Originally posted by Schoerner:

This is fundamental ballistics and no matter if CM uses simpler calculations, what counts is, that the effect showed by CM, can be understood and explained in reality.

And "unhittable" units can be explained very well.

Fine, than please explain my unhittable building test, and especially the part that the building can be hit with patch 1.01 but can't be hit with 1.02.

[Industrializer]

All right, it might be easiest if you have a look for yourself, here is a little unhittable gun scenario, run the test scenario with patch 1.01 and the guns gets destroyed almost immediately, run it with patch 1.02 and the guns will kill the attacking StuGs one by one without ever being threatened.

I agree that it should be hard to hit the guns, certainly harder than in 1.01 but 1.02 made it impossible.

[Schoerner]

Industrializer,

my explanation was clearly (i thought so) aiming to destroy the always repeated BS, that "if someone hits me, i must be able to hit him".

Most of the assumptions why it has to be a bug, are therefore wrong.

As the thread was developing, more and more very interesting and good facts were collected (enough to see, that something needs to be adjusted) but if a calculation's result should be 2, and someone claims he knows the right result, because 4-1 was obviously 2, the conclusion is wrong although he maybe is naming the correct result.

In conjunction with the other collected interesting observations the picture is becoming clearer and it seems that there's something wrong, ofcourse.

But not because of the original argumentation with it's way to less data/facts.

Although i think it's a good sign, that BFC still hasn't responded (because of working hard on the engine-rewrite, i guess ), i'm interested, what their point of view is.

[ March 05, 2003, 09:32 AM: Message edited by: Schoerner ]

[Conscript Bagger]

Just a couple of questions that might help zero in on what exactly is happening...

So far it sounds like the effect has appeared when HE is fired at (a) buildings or ( soft units - AT guns, infantry. No one has complained about armor-piercing rounds fired at vehicles (i.e., hull-down for armor still seems to be working correctly), so the problem must lie somewhere in between. I wonder what happens when:

</font>

• AP ammo is fired instead - does it impact in front like HE, or does it strike the building/unit targeted (even if it does little or no damage)?</font>
• HE is fired at vehicles and bunkers in an (evidently) "HE-proof" position - does it impact in front, miss long, or have a chance of striking the target directly?</font>

I'll test this myself when I get a chance, though that's not likely to be today.

edited because you shouldn't use the same word three times in two sentences.

[ March 05, 2003, 10:48 AM: Message edited by: Offwhite ]

[hakstooy]

Hey everyone, I'm going to refer to the thread in which Steve says that this is a bug, and that apparently 1.02 is related to it. go HERE , where nevermind linked on the 2nd freakin post on this thread. The thread this links to has a post by Steve in which he clearly states that this as a BUG.

The topic of this thread relates to a walking in the shots gradually up a berm, which Steve said he was unaware of. This simple plodding away at the same spot he acknowledged was a known problem.

I'd like to think that this should end the debate, but I doubt it will.

Schoerner, you are right about not necessarily being able to shoot at what is shooting at you, but that is NOT the case in these situations. The gun is not entirely below the rise, the barrel is above it (thats how you can see it). Thus, it is very much hittable, and while very difficult to knock out, i'd think the odds of blast damage at least freaking out the crew are pretty good in 700 shots. Not to mention a building where 75% of it is clearly visible.

[ March 05, 2003, 11:14 AM: Message edited by: hakstooy ]

[Eric Alkema]

Originally posted by hakstooy:

The thread this links to has a post by Steve in which he clearly states that this as a BUG.

To quote Steve in the post to which you refer "Er... not correct." What he said was "That is something I don't think we have ever seen before." In my version of reality that is not the same thing as saying that this is a bug.

[hakstooy]

Eric, read his post again.

Even though i said this in my last post, I will say it again. when Steve says "Er...not correct..." he is referring to when someone said that the walking problem was identified by him as a bug. This is what he is referring to as "something I don't think we have ever seen before." NOT the shooting repeatedly at the same spot, which is what he acknowledges is a bug.

The difference is:

shooting at the same spot always short

vs.

walking each shot incrementally towards the target

To me these seem somewhat different. As they obviously did to Steve since thats what he said. There really is no need to include snide comments, especially when you obviously didn't carefully read the referred posts.

[Neutral Party]

I guess Steve knows what he meant but a theological debate on the interpretation of his sayings ain't gonna help much.

[hakstooy]

Heh, yeah I guess he would Neutral.

Perhaps I am simply wrong in interpreting what he said, although it seems completely straightforward to me. Am I wrong here?

[ March 05, 2003, 12:15 PM: Message edited by: hakstooy ]

[YankeeDog]

Schoerner -

Unfortunately, I do not have the time right now for a detailed response to your post. If I have the time later I will post a more detailed decription of the physics involved.

Briefly, though, parabolic trajectories are completely reversible. IOW, if you have two guns of *exactly* the same Mv and other ballistics characteristics, the trajectory (and therefore angle of elevation, etc.) that allows one gun to hit the other is an *exact* mirror image of the firing solution for the projectile going the opposite way.

Your mistake in your example appears to be your assumption that the trajectories can be "linearized" - by the time the tanks projectile reaches the rise just in front of the gun in your example, it has already reached it's apogee and is on it's way down. Not coincidentally, the downward angle of the tank's projectile in a 'perfect' firing solution in your example is the exact same 2.86 deg of the outgoing gun projectile that hits the tank.

I should note that this does not take into account for wind resistance, but the slowing of the projectile due to wind resistance actually favors the tank, not the gun, if the intervening rise is closer to the gun than the tank. Wind resistance means the tank's projectile will be travelling more slowly as it nears the gun and therefore have a higher angle of attack.

IOW, what the tanks needs to do in order to hit the gun is calculate the trajectory that will *just* clear the rise in front of the gun, basically skimming the ground in front of the gun and impacting the gun itself or the gun shield - NOT hitting the ground. Assuming the gun protrudes above the ground level to any significant degree AT ALL, the tanks can hit the gun.

Mortars etc. in "Tobruk" type gun pits where the gun is completely blow ground level are a totally different matter.

For this situation, in the conditions where the tank can't hit the gun, the gun can't hit the tank, either, as it's projectile will fly over the tank and impact somewhere behind.

Once again, I will try to detail the physics later.

Cheers,

YD

[Schoerner]

Yankee Dog,

i appreciate your effort, but you can save your time, the trajectory is not reversable because of velocity.

You are also right, that the angle of the impacting projectile is higher than the projectile leaving the gun.

You are right, if two identical weapons are on each side, each could theoretically hit the other. But the one that's closer to the ridge in reality is much harder to hit due to the minimal projectile's "window". Only under a certain velocity - over a certain distance - the angle is less than -2.86 degree (from our example).

If the angle is more flat the projectile would hit either the ridge, or if it would fit perfectly, it flys (centimeters) over the gun, while the unit close to the ridge has a (much) wider hit "window" (compared to the minimum/zero exposure of the gun).

Additionally we should take into acount, that the 2.86 degree are referring to the last possible distance for the gun, to be able to hit the tank while larger distances still offer the gun the whole hit "window", while the window of the tanks stays extremely small although the projectile's negative angle increases with distance.

In reality the hit chances are almost zero - and this was well simulated by CM as i thought - before the other facts were presented.

[ March 05, 2003, 03:47 PM: Message edited by: Schoerner ]

[Snorri]

heh, it ain`t a f+#k -bug. It`s a absolet/absolut important (if not even essential) point for an AT-Gun crew to figure out such positions..

what i suggest a "bug-in far meaning", is the point your heavy arti/tank-fire cannot "lower" a hill-top after some time n` "constant" pressure..

what turned out to be a bug, was me, as i saw myself i the position of a really-"stupid"-officer telling his men to carry on with "dumping ammo", without concern about the last-minute`s results...

[YankeeDog]

Schoerner -

I think we are misundertanding each other.

What you are saying is certainly true, it just doesn't apply to the situation I was testing.

I my setup, the tank was in a small valley below the gun. The tank was also as close as it could get to the gun and still be hittable - any closer and the tank would be in the 'shadow' of the hill and the gun would not be able to hit it because the very ridge the gun was sitting on would block it's view of the tank. In this position, the tank was also "Hull Down" to the gun for the very same reason - the hull of the tank was blocked relative to the gun's line of fire by the very ridge the gun was sitting on.

There was only one angle of shot from the gun that could hit the tank: One that just skimmed the rise in front of the gun and hit the tank's turret. Any more angle, and the shot would fly long over the tank. Any less, and the shot would impact the ridge just in front of the gun.

Similarly, there was only one angle of shot from the tank that *should have* been able to hit the gun: One that just skimmed the ridge in front of the gun and impacted either the gun itself or the gun shield and aiming mechanism. Any higher, and the shot flies long over the gun. Any lower, and the shot impacts the ridge.

So both shooters were at a serious disadvantage, and both have to score a direct hit on a pretty small target. In the case of the gun, its only available target is the tank's turret front. In the case of the tank, it's only available target is the gun itself, and the portion of the gun shield that protrudes above the gun barrel (maybe 20-50cm of steel plate, depending on the type of ATG).

This certainly give the ATG an advantage - for most matchups, the turret front of the tank is going to be a somewhat larger target than the top part of an ATG. But only by a factor of two or so, not by enough of a factor to justify a 51% vs. 0% chance "to hit", which is what I was seeing in my test runs.

Just as a WAG, maybe in this situation the tank should have a 3-5% chance of hitting the gun. It just shouldnt be 0%.

[Eric Alkema]

Originally posted by hakstooy:

Eric, read his post again.

he is referring to when someone said that the walking problem was identified by him as a bug. This is what he is referring to as "something I don't think we have ever seen before." NOT the shooting repeatedly at the same spot, which is what he acknowledges is a bug.

I thought he also said that the shooting repeatedly at the same spot bug was fixed by version 1.02. I will admit that that it is not 100% clear to me how Steve's quote relates to the current topic. Maybe I am just dim.

Dim as I may be, it seems rather clear that tanks are not aware that their shots are hitting the dirt short of the target and that this is a Bad Thing.

[UberFunBunny]

I think BTS is not responding here because they are spending their time fixing this bug!

[Broken]

These "unhittable gun" problems confuse many people because there are really three different issues involved:

1) the model CM uses for "hitting" soft targets -it's different from the model for armor or for buildings,

2) the ballistics of whether a gun is hittable or not (Schoerner has the right idea),

3) the aimpoint used by the Tac AI.

SOFT TARGET MODEL:

In the model used for soft targets in CM, there is no "hit" or "miss" like there is for armored targets. The shell simply impacts the ground and the resulting HE blast is factored into whether the gun survives or not. That is why you never see a "hit" message like you do for armor.

So, even if a tank can see an AT gun (soft target) behind a crest, it can't actually hit the gun, only the ground near the gun, due to the soft target model. If the gun is on a reverse slope behind a crest, it can be unhittable because the sloped ground it is sitting on is ballisticly unreachable

Incidently, this soft target model was the source of the "Flak truck" bug in CMBO. Flak trucks were modeled as soft targets, they could not be "hit" like armored targets. Only HE rounds would kill them through blast, but the tank's Tac AI saw the Flak truck as a "Vehicle" and fired AP instead of HE, to no effect.

UNHITTABLE GUN BALLISTICS:

Now, is it possible for a gun to be located behind a crest so that a tank can see it but not hit it? Most definitely.

Take the simple case where the the AT gun mount, the crest, and the tank gun mount are all at the same height. The AT gun can hit the tank, since CM uses the armored target model. No problem. However, the tank HE round must clear the crest with a trajectory slope greater than the ground slope the gun is on, or the round will sail right on past (or through) the gun. Here's a simple formula you can derive from high school physics:

If the Ground Slope is > 9.8 * D/(Vm^2),

the gun is unhittable,

where,

9.8 meters/sec^2 is the downward gravitational acceleration of the shell,

D is the distance between the gun and the tank in meters,

and Vm is the muzzle velocity of the tank gun in meters/sec.

Aerodynamic drag is not included in the formula, but would add about 10% to the maximum hittable slope at 500 meters for a Panther AP shell.

What this formula tells us is that low Vm shells are better at hitting reverse slopes than high Vm shells. In fact, if you halve Vm, you can hit a gun on a slope which is four times steeper. The formula also says that it is easier to hit a reverse slope gun from long range than from shorter ranges. So, if you can't hit it, back up (way up).

The thing is, the formula says it doesn't take much of a slope to render a gun unhittable. A slope of 0.1 (one meter drop in ten meters horizontal), and a gun is pretty much unhittable (plug in some numbers and you will see what I mean).

What if the gun and the tank are at different elevations? A tank shell fired from a higher elevation will start out with an initial downward velocity, so it will be easier to hit the reverse slope. Conversely, a tank shell fired from a lower elevation will have an initial upward velocity and it will be harder to hit the reverse slope.

TAC AI AIMPOINT:

Where, exactly, is the TAC AI aiming the tank gun? From my experience and the tests mentioned on this thread, it appears the tank is aiming at the spot of ground the gun is sitting on, REGARDLESS of any intervening hillcrest or wall. This seems to be the cause of all the impacts on the front face of a hillcrest. This, if not a bug, is a poor choice of aimpoints. A better aimpoint, as pointed out in this thread, would be the hillcrest peak itself.

The same problem makes buildings behind a crest unhittable as well. The tank is aiming at the base of the building.

This aimpoint "bug" should be fixed, or at least improved on. However, the fix should not take away the advantage a reverse slope gun has in real life.

[PeterX]

Broken posted:

This aimpoint "bug" should be fixed, or at least improved on. However, the fix should not take away the advantage a reverse slope gun has in real life.

Do you feel the reverse slope effect was accurate in 1.01?

[UberFunBunny]

This "reverse slope effect" that everyone is talking about - can the current CM engine actually take this into account? I have my doubts. C'mon BTS, give us the goods on this....

[wadepm]

I set up my own little test - AT gun behind low rise facing 10 elite Panthers. In version 1.01 the gun is taken out after about 6 shots. In version 1.02 the gun is invulnerable, all shots (100s of shots) hit the slope in front of the gun. There were sounds of MG rounds hitting the shield which eventually caused the crew to abandon the gun after a couple of turns. I wonder what Charles changed in 1.02?

[Battlefront.com]

First of all guys... I am simply dumbfounded how anybody could think that I confirmed that this was a bug in that previous thread. I have read it and read all the commentary about it and still have no clue how anybody could think that.

Ironically, my comments in that thread were aimed at dispelling a similar false claim that we had confirmed a bug that we had absolutely not confirmed.

Folks, when we find a bug and confirm it we leave absolutely NO DOUBT in anybody's mind about it. We also explicitely state if the bug will be addressed or not. And if we do not think it is a bug, or are not sure, we say so straight out. If you find yourself splitting hairs and reading tealeaves to determine if what we said confirms a bug, then the answer should be obvious... someone is misreading our comments.

Now that that's out of the way, let's look at this new bug that I have not even heard about until now

There appears to be three hit issues that have been raised on this BBS. Here they are and where we stand on each:

1. Shots hitting ground when shooter is Hull Down - this was understood to be a bug going all the way back into CMBO, but was not apparent until CMBB 1.0 (and wasn't even seriously seen until 1.01 for some reason). This was absolutely a bug and was fixed for 1.02.

2. Shots walking to target one right after the other in a straight line - looks like something is wrong, but Charles has not looked into it yet and therefore it is not confirmed as a bug. If it turns out to be a bug Charles will try and fix it.

3. Shots impacting at base of a small hill and not hitting some types of targets (i.e. guns) when using more or less flat trajectory fire - my hunch is that this is a bug, but since I only passed the info onto Charles 5 minutes ago I don't have confirmation yet at this point. If it turns out to be a bug Charles will try and fix it.

Is this clear enough are people going to find a 4th situation and say that I confirmed it is a bug before I ever heard of it?

Steve

[Battlefront.com]

UberFunBunny,

This "reverse slope effect" that everyone is talking about - can the current CM engine actually take this into account? I have my doubts. C'mon BTS, give us the goods on this....

Er... how could we have Hull Down positions if this wasn't taken into account? It is not calculated to the mm, but it is simulated. If something is partially exposed enough to shoot, it can in theory be directly hit. But the size of the target means a whole heck of a lot. An ISU-152 in hull down is a lot easier to hit than a Pak36.

In theory flat trajectory rounds should have problems hitting such targets when on the same or higher plane. Rounds will tend to impact in front of or sail over the target. The degree of difficulty hitting the target dead on will, as mentioned above, depend heavily on the target's size.

Steve

[UberFunBunny]

Steve,

how could we have Hull Down positions if this wasn't taken into account?

Very good point!

But I was really talking about AT guns and similar as targets. It appears your post answers this as well, so thanks ...

... but to clear up any confusion (for me!), are ALL targets treated in this way (a form of hull down) internally?

[Battlefront.com]

UberFunBunny,

.. but to clear up any confusion (for me!), are ALL targets treated in this way (a form of hull down) internally?

Sure, if they are significantly behind the terrain in question in relation to the target. A unit behind a wall is not "hull down" to an enemy airplane, but at the same time could be "hull down" to an enemy ground unit. Depends on the units in question, their relative positions, and the terrain inbetween.

Steve