"Un-hittable Gun" Game Bug?

[BadgerDog]

Originally posted by xerxes:

If, on the other hand, you area target your tanks will finish him off much more quickly.

ahhhhh !!!!

I never realized that ... great suggestion... thanks for the tip.

Regards,

Badger

[Soddball]

Originally posted by Industrializer:

OK, I'll try again to illustrate what I think is buggy in CMBB.

I drew a little sketch with the situation I encounter in the PBEM. The targeted building is in a small sinkhole (and no, it's not like 1,50 as one member said, it's really just barely visible!) and the tank is ordered to area fire at the building. What happens is that the tank tries to fire at a point at ground level inside the building but on it's way to that point the shells impact in the small rise infront of the building. The building is clearly visible and if the tank would aim at approx. 1m height it would hit the building easily!

You don't seem to understand. The difference in height cannot be less than 1.25m. That's the smallest possible height difference on a CM:BB map. Go to the scenario editor and try it out.

[YankeeDog]

OK, since everybody seems to be getting all up in arms about this 'bug', and I'm stuck at work on a slow day, but without my PBEM files, I figured I'd step up and run some tests so we actaully have data to work with rather than wild speculation based on individual cases.

Right now, I am testing the original situation brought up in this thread - i.e., the "unhittable gun" situated on a fighting crest. The "unhittable building" issue that is also being discussed in this thread is another, possibly related issue, but it needs to be tested seperately.

Here's the setup I am testing right now:

Gentle slope map.

July 1943, Clear day, cool weather. Allied attack.

A German 50mm ATG (regular exp.) set up on a ridge 3 elevation levels higher than 'base' (3.75m). I have acutally created a 'notch' the ridge by making the the tile that the gun actually sits on one elevation level lower than the rest of the crest. This creates a nearly ideal 'fighting crest' position for the gun as it is sitting in a slight depression. I removed all AP and T ammo from the gun in order to render it virtually impotent against tanks.

In front of the ATG, I have created a long, shallow depression that is a near mirror image of the rise the gun is sitting on, 3 elevation levels lower than 'base' (again, 3.75m). The center of this depression is about 260m from the ATG's position.

Now, I placed a platoon of T-34/76s armed only with HE ammo (100 shells each) in the depression. The T-34/76s are placed so that they just *barely* have LOS to the ATG - A few meters farther forward and they would be out of LOS. In this position, the tanks are hull down relative to the gun, and are sitting on the 'floor' of the depression about 275m from the gun. Theoretically, then the tanks are about 7.5m (6 elevation levels) lower than the ATG. Incidentally, in this position, if you manually target the ATG with the tanks, you get the mysterious "0% exposure" rating. If you place the tanks 20m behind their setup position (and out of the depression), the gun shows 75% exposure.

I need to run more trials before I can say anything definitive, but since the initial results are rather interesting, I though I would post them here.

For simplicity's sake, I am running all tests with no FOW.

Before running the tests, I confirmed that the ATG could hit the tanks by giving the ATG a few of it's AP rounds back and giving the scenario a quick run - the TacAI usually won't have a 50mm ATG with only HE ammo target a T-34 on it's own. The ATG scored a hit within 4 shots.

Next, I ran the scenario again, with no AP ammo for the ATG and just watching to see what the tanks would do. The tanks almost immediately targetted the ATG and began firing away. EVERY shot impacted the rise about 30m in front of the ATG. the impact craters were very closely clustered - there was no 'walking' the fire towards the gun.

Through the first 4 turns, the tanks fired away, firing about 20 shots each. Never once did a shell overfly the ridge.

On turn 5, a curious thing happened. Evidently, the ATG crew got tired of watching the tanks dig holes, and even though they only had HE ammo, the gun crew targetted one of the T-34s. On the third shot, the ATG got the range and hit the T-34 turret front for the first time. Oddly, on the fourth shot, the ATG hit the T-34's upper hull, even though a quick check with the LOS tool after the turn was over showed the T-34 to be 'hull down' relative to the ATG. On the sixth shot, the ATG managed to take out the T-34's gun. As soon as the T-34 went "gun damaged" the ATG stopped targeting it. For the remainder of turn 5, the ATG stopped firing.

I kept the scenario running for 5 more turns, at which point the 2 T-34 with functional guns had fired almost 50 shells each, all impacting about 30m in front of the ATG. The ATG did not fire any more shots at the tanks.

****

I need to run a lot more tests (and put the gun on hide so it doesn't mess up my trial. But a few preliminary observations:

1) Not really related to the bug issue, but interesting: Judging by the gun's odd behaviour on turn 5, the TacAI seems to have a behavior where it will occasionally target units it has very little chance of damaging. Interesting.

2) The 76.2mm L/42 gun has a lower muzzle velocity that a 50mm PaK, at least when both are firing AP ammo. although the PaK has a somewhat lighter projectile that will be slowed more quickly by wind resistance, I feel fairly confident in stating that the trajectory of the 50mm PaK AP ammo is flatter than the trajectory of the 76.2mm AP, at at least over the relatively short ranges on this test.

3) Given that the 50mm PaK can hit the T-34s' turret with it's flatter trajectory projectile, the T-34 *should* be able to also hit the PaK. The area of the PaK hittable by the T-34 may be very small (basically just the gun barrel and the very top of the shield), but a hit to either of these should be possible.

4) Over somewhat more than 100 shots fired in my initial mini-test, none of the T-34 scored a hit close enough to the PaK to even alert the crew, let alone cause damage or casualties. Furthermore, since not a single shell overshot the gun, the T-34s appear to not even have been aware that their aiming point was off and all their shots were falling short. Logically, at least some of the tank's shots should fly long and over the ridge.

Clearly, I need to increase the amount of data before I can say anything definitive. However, on first examination, it does appear that there is some sort of bug. Sooner or later, a T-34 should get lucky and put a shell basically right down the barrel of the PaK. I am also going to rerun the test with a very low velocity tank gun vs. a very high velocity ATG. A slower shell should equal a higher trajectory and therefore a better chance of hitting the gun.

So anyway, hopefully more data will follow soon. In the meantime, here's some preliminaries to chew on.

Cheers,

YD

[Sergei]

Originally posted by Soddball:

Cretin.

Okay, I admit - you really are useless to this thread, being blind to other people's arguments and only seeing your own, faulty ones. Why faulty? Two reasons. First of all, because it has nothing to do with the LOS. Secondly, because while you can only place elevations with minimum steps of 1,25 metres in the map editor, in the actual preview the contours are smoother.

So, what you have accomplished, is to make false claims and call me a cretin. Gee that must be the highlight of your day?

P.S.

[Porajkl]

No comment from BTS yet?

Just wanted to add that this didn't happen in CMBO. So now in CMBB either it's a bug or a feature ????

Now it's quite annoying (very very mild word).

... silence gives consent ...

[Other Means]

about the height steps, if a tank is travelling along an incline, the incline starts at 0 m & goes to 1.25m after 10m laterally, is the tanks actually at 0m elevation until it is 5m along at which point it steps to 1.25m? i do not believe this to be the case.

even if the terrain was smoothed out from 1.25m steps, there is still a disparity between what the AI thinks it's hitting & what it is actually hitting.

it does seem that guns have an apparent height of 0 when being targetted but can shoot from their actual height.

IMHO Dschugaschwili (bless you) got it spot on.

[Panzer76]

We have all experienced this in one form or another, and when it happens its really annoying.Bug ot not, there is something clearly wrong here, and that should be OBVIOUS to everybody. Nit picking about screenshots etc isnt very helpful.

I would also like to see an offical answer to this, can we expect to see a fix for this? This problem is on the same level as the pre 1.02 AFV that fired right in front of them when in hull down, and that one was fixed.

[Sergei]

Originally posted by Porajkl:

Just wanted to add that this didn't happen in CMBO. So now in CMBB either it's a bug or a feature ????

I was thinking of this... could it have something to do with the changes in how the terrain is drawn (it's somehow "smoother" than in CMBO), or the new 1,25 elevation mode? Just tossing some ideas.

[LGW]

Keep at it industrializer.

I think it's a bug, but I just accept it now, bug or no bug.

I have never taken much stock in morters in CMBB like I did in the CC series. After reading this post I had 2 AT guns knocked out by 82s. Readers luck i guess!

Now I always go in with an 81. Preferably with the 251/2 or 250/7. My first QB before work yeserday and I knoced out 2 pesky 45s within the first 5 min. Just wish they had an MG as I have to be really carefull with them. Thanks Soddball.

[UberFunBunny]

Never once did a shell overfly the ridge.

This is my experience also. It is an important clue.

The "this is a feature" crowd would like you to believe that the gun is well placed, and that tanks fire into the ground in front and also over the head of the gun into the fields beyond. In reality, the game only hits the ground in front.

This is probably the most serious bug I have seen in CM.

[JonS]

Originally posted by Porajkl:

... Just wanted to add that this didn't happen in CMBO. ...

Yes it did. Not to the same (reported) extent but it did happen.

Regards

JonS

[Industrializer]

Originally posted by Soddball:

You don't seem to understand. The difference in height cannot be less than 1.25m. That's the smallest possible height difference on a CM:BB map. Go to the scenario editor and try it out.

Oh well, than for god's sake it's 1.25m, but that's not important. What's important is how much of the building is visible to the TC and presents a valid target. here is another screenshot, taken at elevation 1 at the position of the PzIV G with zoom x8.

It's obvious that most parts of the building are visible to the TC but he keeps hitting the road infront of the building (in this picture covered by the roadblock).

My question plain and simple is: Is it possible for a PzIV G equipped with good optics to hit a non moving target that is as big as this building at a distance of <300m, that is as visible as shown in the above screenshot, yes or no. If you answer is yes than it's a bug in the game, if your answer is no than I wonder how tanks could have hit anything in WW2.

[Sgian Dubh]

Well I just ran a few short tests with the following result.

Map: 800x800 clean. All tiles at height of 7.

I put 4 houses spaced evenly apart 60 meters from the North edge. I placed a 1.5M berm three tiles wide just in front of one of the houses and a 1 tile berm 1.5 M high in front of another house. The berms are in the tile immediately next to the tiles containg the houses.

I placed 1 45mm Soviet AT gun as close to each house (south side) as possible, located as close to the 'middle' as possible.

Here is a cheapo illustration:

H H H H

G G G G

BBB B

P P P P

(my illustration doesn't work because whatever is parsing the text on thie forum isn't preserving the spaces that used to space things out - sorry about that.)

I placed 4 Panthers on the South edge of the map, so about 700 meters distant. I forced each Panther to target the AT gun directly in front of it to begin.

I let 'er rip.

Turn 1, all AT guns but the one behind the 3 tile wide berm knocked out. The AT gun behind the 1 tile wide berm lasted a bit longer. It was nailed by the Panther 2nd from the right and not the one directly in front of the gun.

Turn 2, all Panthers fired at remaining AT gun.

All shells from the Panthers strike the berm.

Turn 3, I manualy targeted area fire onto the house from 3 of the 4 panthers (leaving the one directly in front of the gun targeting the gun on its own).

All of the shells from the 3 panthers with area fire on the house still struck the berm.

Turn 4 I moved the 2 outer panthers to area fire aournd the house where I thought they would not strike the berm.

The shells struck the ground around the house on both sides.

The AT crew abandonded the gun on this turn.

I ran this same test in this was 5 times with the same results. The crew always abandoned the gun on turn 4.

I ran a new test where I kept all of the panthers firing at the gun on Turns 2 & 3. I five out of five tests the crew abandons the gun on turn 3.

So then I thought to myself, what if I didn't have any MGs firing on the gun crew?

So I replaced the Panthers with early StugIIIs.

In five test runs I got the following results:

2 AT guns with no protection taken out in turn 1,

AT gun behind 1 tile berm lasted until turn 2 in every test.

The AT gun behind the 3 tile berm out-lasted the HE ammunition in all of the Stugs in every test.

AND the AT gun took out or immobilized at least 2 of the Stuggs in every test.

So my suggestion is that if you don't have indirected fire weapons and can't maneuver aroudn the AT guns' protection, then pour MG fire into it and you might be able to get the crew to abandon the gun. And never send in an early Stug to try and dig out an AT gun in a position like this.

[PeterX]

So my suggestion is that if you don't have indirected fire weapons and can't maneuver aroudn the AT guns' protection, then pour MG fire into it and you might be able to get the crew to abandon the gun.

Yes, that works, especially with a platoon's worth of AFVs. Problem is, the TacAI will override the MGs and revert to HE. Also, FWIW, this bug tends to invalidate a lot of Single Player scenarios; there's not horde of AI tanks big enough to prevent me from killing every single one with a 'hull down' AT gun.

And never send in an early Stug to try and dig out an AT gun in a position like this.

Amen to that! Unless it tossing canister.

[mike the wino2]

Quick question and I hope I am not repeating an all ready answered question.

Are you targeting the building or the base of the building? In CMBO I had great success targetting an upper floor of a partially exposed building. I haven't played too many battles in town where I get to test this.

If a certain AWOL person, yes lenakonradI am talking about you, ever shows up again I will test in game.

[RoughRoad]

Originally Posted by YankeeDog:

OK, since everybody seems to be getting all up in arms about this 'bug', and I'm stuck at work on a slow day, but without my PBEM files, I figured I'd step up and run some tests so we actaully have data to work with rather than wild speculation based on individual cases.

I'll be anxious to hear the final results of YankeeDog's tests. But in the meantime, let me ask the question in reverse:

Has anyone running v1.02 ever had a situation where:

1) A gun in defilade was engaging a large number of AFV's

2) For say, 7 or 8 turns, the AFV's fired away with HE but had no effect rather than overshots or digging the proverbial trench a few meters in front. (This may sound like a long time, but it's nothing compared to the 23 turns w/o a hit from the situation that originated this post. Besides, if the gun gets hit before this, it might be an indication that the level of defilade was not comparable to the ones discussed here.)

3) On say, turn 9 or so, without changing position, the AFV's suddenly score a direct hit with HE and knock out the gun?

If the answer is yes (and I don't expect this), it would seem to me that this would make a pretty strong case that this is not a bug-- that instead defilade just creates an incredibly low probability shot.

But if people aren't seeing hits like this occuring, it would seem to me to be yet another indication that were dealing with a legitimate bug.

[hakstooy]

it seems pretty obvious from all the entries here that whats happening is the tank targeting the base of AT guns and buildings. how can anyone be arguing that this is a bug? you can ignore trajectory and crew experience if what you have is a AFV being programmed to shoot only at the portion of a gun that is obscured by cover. it seems to me that steve acknowledged that this is indeed a bug, but that the "walking" illustrated in the initial pictures was new to him. what i don't get is that if the guns are programmed to have a height for their own fire, why can't this height be used to create a target which is larger than a point?

you can argue all you want about "using cover" which is indeed a good tactic, but not entirely whats happening here. this is an example of a flaw in programming that COMPLETELY eliminates the possibility for a kill, instead of merely reducing it.

[hakstooy]

sorry, i was referring to the post in the thread referred to by nevermind on page 1. the thread that steve responded in.

[Industrializer]

Originally posted by mike the wino:

Quick question and I hope I am not repeating an all ready answered question.

Are you targeting the building or the base of the building? In CMBO I had great success targetting an upper floor of a partially exposed building. I haven't played too many battles in town where I get to test this.

It's a single store light building, you can _only_ target the base.

[Visom]

Just ran a quick test.

5000 pts meeting. 11 platoons of Crack StuGIII(early) (33 tanks) against a veteran 76.2mm L/51 M36 gun.

Placed gun in 'good' spot out of C&C.

Turn 1-5. 33 tanks tries to take out the gun with HE. Gun crew is pinned.

Turn 6. Appr 50% of the StuGs now run out of HE.

Turn 7. All StuGs have run out of HE. I didn't monitor every single shot, but it seems like not a single one overshot the gun. In the crest in front of the gun there are more than 700 hits. Gun crew is now cautious.

Turn 8-15. StuGs start using AP, roughly 5 shots per turn in total per turn. All of them, hits the ground 20 meters short of the gun except one, which hits the ground 10 meters in front of the gun. Gun is still unharmed, crew is back to rested.

[ March 04, 2003, 06:39 PM: Message edited by: Visom ]

[YankeeDog]

OK more tests.

This time, just to make things a little bit more interesting, I switch things around and pitted a platoon of PzIVFs (with the L24 gun) against a Russian 45mm gun.

Otherwise, the conditions were exactly as before - the ATG in a near perfect 'fighting crest' position, the tanks about 275m in front of the gun, in a slight depression.

I also didn't even bother to put the gun in a foxhole this time - it was just out in the open with no protection other than the fighting crest position.

I made this switch because the 75mm/L24 on the PzIVF has a significantly lower Mv (420m/s with HE) than the Russian 45mm ATG (820 m/s with AP). This means that there is NO doubt that the trajectory of the ATG is 'flatter' than the tank. Accordingly, if the ATG can hit the tank (or more specifically, it's gun), then the Tank should DEFINITELY be able to at least score a 'lucky' hit on the actual gun (or gun shield), even if it can't land a shell on the ground in the immediate vicinity of the gun.

So the results? Once again, ALL HE fired by the tanks falls at least 30m short. I did one full run with the full PzIVF platoon, meaning that a total of 400 HE shells were fired at the ATG (5 tanks x 80 shells each).

I was going to run the test a couple more times until I had a nice, round sample size of 1000 shots, but I got bored, and anyway no shot so far has come even close, so I think the results are pretty clear.

Previous to running the test, I again gave the ATG some AP ammo in a "pre-run" to make sure it could hit the tanks. It was not only able to score turret hit, but actually also managed to score a couple of upper hull hits, again something I was not aware was possible when a tank is supposedly 'hull down' to a given target.

Given that the ATG could hit the upper hull of the tank, well below the muzzle of the tank's gun AND the ATG has a Mv nearly double that of the tank's gun, it is physically impossible that the ATG would be able hit the tank without any chance of return fire from the tank hitting the gun or the gun shield.

In such a situation, it rightly should be VERY difficult for the tank to hit the gun, as the exposed area of the gun is very small. But it should not be impossible. You can call it a 'bug', you can call it a 'model limitation', or whatever else strikes your fancy. It is unrealistic and physically impossible for the ATG to be invulnerable under the laws of Newtonian Mechanics. If anybody feels like explaining how Quantum Mechanics actually comes into play in this type of situation, rendering the ATG invulnerable, I'm all ears.

If it can be fixed, I certainly hope BTS will take the time and effort to do so. However, some comments on this thread and elswhere about how targeting of non-vehicle units works makes me fear that this may well be a model limitation that will not be easily patched.

The good news is that this problem should actually have a pretty limited effect on gameplay as long as you know what to look for.

First of all, if you are engaging an ATG set up in a good 'fighting crest' position by positioning tanks in front of it and slugging it out, you deserve to get your @ss kicked. Hitting an ATG in this position should be extremely difficult. It just shouldn't be impossible. Just like you can take out a KV-1 by targeting it with an entire company of Pz38(t)s and hoping for lucky gun and track hits, you should be able to target a fighting crest-positioned ATG with a company of tanks and be pretty confident that you'll score that lucky hit before you lose your entire company of tanks. Hardly an elegant or efficent strategy, but you should be able to do this is you really want to.

It trick is to look for that "0% exposure" flag when you check your LOS. If you see that, you probably can't actually hit the target with HE. Maneuver and reposition rather than engaging from your current position. If you really need to supress the gun from your position, use your MGs, or even better use smoke shells, which will have much the same effect, at least for as long as the smoke lasts.

As well, it is difficult for the player controlling the ATG to exploit this bug. The offending behavior only occurs in a fairly narrow band (about 20-30m), and since only the opposing player can use the LOS tool to see when the "0% exposure" flag comes up, the ATG player has to guess whether or not his ATG is 'invulnerable' against a tank in a given location.

So, at least as far as Tanks vs. ATG, this bug is not a game wrecker. It's annoying, but as long as you check your targetting lines, at worst you'll waste a turn's worth of HE before you can order your tank to reposition.

The buildings issue is another matter, but I suspect it is basically the same problem. The really annoying thing here is that there is no way of knowing for sure if you are at "0% exposure" when issing an "Area Fire" order. I guess you just have to watch for clumps of craters on a small rise in front of the building, and reposition if you see your tank coming up short consistently. Unfortunately, this means risking loss of substantial HE and valuable time.

Here's hoping for a fix. . .

YD

[UberFunBunny]

YankeeDog:

The good news is that this problem should actually have a pretty limited effect on gameplay as long as you know what to look for.

First of all, if you are engaging an ATG set up in a good 'fighting crest' position by positioning tanks in front of it and slugging it out, you deserve to get your @ss kicked. Hitting an ATG in this position should be extremely difficult. It just shouldn't be impossible.

This bug affects *all* opposing units as far as I can tell. In other words, tanks will fail to hit anything behind a crest be it an AT gun or a squad of infantry.

As well, it is difficult for the player controlling the ATG to exploit this bug.

This is not correct. I can now plant my units in "HE free" positions quite easily.

the opposing player can use the LOS tool to see when the "0% exposure" flag comes up, the ATG player has to guess whether or not his ATG is 'invulnerable' against a tank in a given location.

Has this been confirmed? I'm not so sure if the un-hittable units always have a 0% exposure rating.

[PeterX]

Hitting an ATG in this position should be extremely difficult. It just shouldn't be impossible.

If you mean by hitting the gun barrel or a curious crew member on the fly, yes, I think it's impossible given the constraints of the current game engine. I don't see how BTS can do it. Stretching the blast zone forward toward the rear? Making the tanks lob their HE shells like a mortar? Defaulting to the MG?

There are work arounds, at least in MP. But I don't look forward to taking on a player armed with this knowledge in, say, a free wheeling steppe battle. I see a lot of burning tanks. It's a bug.

[Sgian Dubh]

One critical thing that I forgto to mention about my test is that the target tool reported the exact same % of exposure for all 4 guns.

This means that as far as the tool is concerend, the berm was providing no aditional coverage for the AT gun. I am talking about only the Stugs or Panther directly in front of each gun.

[Visom]

Originally posted by UberFunBunny:

Has this been confirmed? I'm not so sure if the un-hittable units always have a 0% exposure rating.

I had a 75% exposure in my tests. Doesn't that depend only on the terrain? Mine was not dugin and in open terrain. I think the exposure is less in other terrain types and in foxholes... I guess we should be careful so we don't mix exposure with hit probablilty.