Question about vehicle weapon hit probability

[jocke_p]

Hi there fellow commanders!

Let me first tell you all i am a long time fan of the original Combat mission series, i own all three CMx1 titles. Like so many others i have opted not to buy CMSF or its other modern expansions, mainly due to the fact that i am not intrested in modern warfare. I did try the demo on a couple of occasions but it was not my cup of tea.

I have been waiting for CMBN since it was announces, and it has been a blast plying it these last couple of weeks. It has been everything i hoped for and much more.

BUT, there is always a "but" isn´t there

I started to dabble a little with the editor to start my huge campaign project, and i am in the process of testing the AI scripts i made. It was then that i noticed something i did not quite like, it felt a little "wrong".

The situation is this, i advance three crack panthers slowly along a road where all three stop, 600 m or so straight ahead there is a american 57mm AT gun hidden in a treeline. This AT gun starts "pinging" away at the panthers and after expending about 20 shots it has achieved roughly 18 hits. Now to the part i find a little strange, the result of this "pinging" is that all three panthers have had their main gun damaged and is now teethless so to speak. I have replayed this scenario 3 times after this and the end result is the same. All 3 panthers are alive but have their main gun destoyed.

So my question is! Does this feel "right" to you guys? what is the probability of a gun hit? I would have guessed that no more than 10% of the hits scored to the turret front would strike the gun. In my game that number is closer to 30-40%. Also all gun hits scored in my games have resulted in the main gun being out of commission, not a single gun hit has resulted in that the gun is still operational.

Could anyone from Battlefront, or anyone else with knowledge about this, explain to me if what i am seeing is historicly correct or if there in fact is something wrong with the chance to score gun hits?

I accept the fact that AT guns are hard to spot and quite accurate but i do not feel that the results i am getting is correct, am i wrong?

I would like to finish with huge "thank you" to battlefront for making this fantastic game. I am looking forward to years of fun with this game series.

BR

Joakim

[Magpie_Oz]

we did sort of touch on this once before and it may be that things like the sights are being damaged as well, rendering the gun useless.

I have no hard data but the gun, being mounted in the centre at the front and unable to be armoured is pretty vulnerable.

I'd be more concerned from the game perspective that an AT is able to fire that many rounds at the tanks and they do not run away.

[Blackcat]

The first question I would ask is what is the attitude of the Panthers to the ATG? Are they hull down? The TACAi will always aim at the centre of the mass, if all it can see it the turret then that is what it will aim at.

The second, related, question is how many hits landed on the Gun as opposed to the gun mount? The end result is the same (i.e. a disabled gun) but the size of the target from a head on aspect is different.

The Panthers were stationary, each present a big target at 600m. In gun terms that is virtually point blank range so 18 hits out of 20 shots is reasonable, except the gun would have had to switch targets. That does make me wonder. How widely spaced were the Panthers?

Finally, the factthat the Panthers had crack crews is irrelevent in terms of the ability of the ATG to hit. You might expect such teams to have a better chance of spotting the ATG, but they can't affect the chance of their tank being hit, especially if they behave like dumb rookies (which your guys were).

Oh, by the way, someone will be sure to mention it so we might as well get it over with here. There are no hit probabilities in CMx. The game models the flight of each shell and they hit where they hit.

[Blackcat]

"I'd be more concerned from the game perspective that an AT is able to fire that many rounds at the tanks and they do not run away"

Maybe they would have run away if their crews hadn't been "crack". Dunno, but it is worth a thought.

[GreenAsJade]

There are no hit probabilities in CMx. The game models the flight of each shell and they hit where they hit.

This is an interesting one to explore further.

What is it, when it comes down to it and you have a gun and a target each stationary, that makes each shot not hit exactly where the previous one did?

GaJ

[Magpie_Oz]

Maybe they would have run away if their crews hadn't been "crack". Dunno, but it is worth a thought.

Yeh true enough, I would have thought that even the crackest of the cracked would "retire" if they took 6 hits from a hidden ATG. Even just the noise would be deafening.

Check out this:

http://www.lonesentry.com/tigerflorence/index.html

For the results when Tigers are attacked by various things you would expect to hurt them.

I love #3 BTW where the driver gets back in the tank and does the bolt and leaves his mates to leg it on their own.

[Magpie_Oz]

This is an interesting one to explore further.

What is it, when it comes down to it and you have a gun and a target each stationary, that makes each shot not hit exactly where the previous one did?

Good point, does the game model all the tank/ATG moving because of recoil, atmospheric changes, differences in propellant, barrel temperature, gunner scratching his bum on this shot, continental drift etc that cause each shot IRL to be different?

[jocke_p]

The first question I would ask is what is the attitude of the Panthers to the ATG? Are they hull down? The TACAi will always aim at the centre of the mass, if all it can see it the turret then that is what it will aim at.

there where some hits to the upper hull on some of the panthers, in each test i did. The AT gun is on aprox 10 meter higher ground so i am quite sure they where not hull down.

That being said, even if they where "hull down", and the AT where aiming for the centre mass, i would have thought there to be a little more variation to the shot hitting a panther turret. That is why i mentioned 10% gun hits being the number i expected, not 30-40%.

The second, related, question is how many hits landed on the Gun as opposed to the gun mount? The end result is the same (i.e. a disabled gun) but the size of the target from a head on aspect is different.

Actually, the hits NOT scoring against the gun but the weapon mount where quite a few as well. when i think about it that makes it even more odd, that there where so many hits localized to the weapon mount and the weapon itself.

The Panthers were stationary, each present a big target at 600m. In gun terms that is virtually point blank range so 18 hits out of 20 shots is reasonable, except the gun would have had to switch targets. That does make me wonder. How widely spaced were the Panthers?

Yes, i am expecting high accuracy at that range, the panthers where about 15 m apart, so it was not a big ajustement for the AT gun. I am NOT questioning the 18 hits of the 20 fired, but the fact that so many of them was gun hits that ALL three panthers ended with the gun damaged.

Finally, the factthat the Panthers had crack crews is irrelevent in terms of the ability of the ATG to hit. You might expect such teams to have a better chance of spotting the ATG, but they can't affect the chance of their tank being hit, especially if they behave like dumb rookies (which your guys were).

Yepp, i understand the concept, in the last test one panther managed to finally spot the AT gun before his main gun was disabled, but 2 seconds later the AT gun struck the weapon of that panther as well, so it never managed to get a shot of. By the way, the weather is dry, clear.

Oh, by the way, someone will be sure to mention it so we might as well get it over with here. There are no hit probabilities in CMx. The game models the flight of each shell and they hit where they hit.

Then i would like to rephrase my question. What is the probability that the shell hits the exact aim point, being the centered gun in the turret. I would think that the spacing of the hits would be a little more spread out along the turret surface. It would be interesting to learn how many hits against the turret where actually hitting the gun, in real life WW2. I have a feeling it where not this common. Then again this is just a feeling and i have no problem with being proven wrong.

[Blackcat]

Good point, does the game model all the tank/ATG moving because of recoil, atmospheric changes, differences in propellant, barrel temperature, gunner scratching his bum on this shot, continental drift etc that cause each shot IRL to be different?

Dunno, and I don't suppose we will ever find out.

On a related note do you gents recall a post from Steve a while back stating explicitly that [tank] gunners have a memory? I am fairly sure he not only mentioned that fact that gunners retained, for a limited time, where a target was if it suddenly went out of view, but, and here is the cruncher, actually acted on the fall of the last shot.

So, unless my dementia is getting much worse, what we should see is, for example, first round over, second round short, third round - split the difference - "Target!". Well I haven't seen that too often.

I zoomed in on a scrap yesterday and the sequence went:

1. Under

2. Over

3. Under - exactly where round 1 landed

4. Way over - much further than round 2

5. Under - back to exactly the round 1 spot

6. Hit

There were no trees in the way either, this was pure crew action.

I have got used to the fact that my Sherman drivers can't manage a clutch or steer in a straight line, but to have gunners who have worse short-term memory than I do, is bit galling. Perhaps the reason I have never seen the ultra-aggresive bailed out crew syndrome is that all my crews forced to evacuate are still looking to sort out their zimmer frames when they are mowed down (the occasional survivor I put down to the German infantry laughing so much they can't shoot straight).

[Blackcat]

Jocke P,

Well on the information you have provided I am as lost as you are as to why you got the results you did. I expect there will be a grown-up along soon to explain it.

[jocke_p]

Jocke P,

Well on the information you have provided I am as lost as you are as to why you got the results you did. I expect there will be a grown-up along soon to explain it.

Hi again.

I did a simple test with a 700 meter plain map with 3 panthers against 1 AT Gun in a treeline.

I am NOT having the same results. 90% of the hits are against the upper hull of the panthers.

So my conclusion is that in my original battles, my panthers must have been almost hull down, seeing as i got at least a couple of upper hull hits. But most of the hits where against the turret.

My original/rephrased question remains though. How many hits should be expected to hit the gun if the center mass aim point is the turret. My feeling is that this is a little high.

Well i am off to run a similar test with all "hull down" panthers.

BR

Joakim

[Blackcat]

I am impressed by your spirit of enquiry and the willingness to run lots of tests. As you might well find out when the grown-ups appear, you will need hard data not just feelings to convince them of the possibility that there might, just might, be a problem. Be prepared for the energy vampires.

P.S. When recording your results you might want to remember that the Gun and "Weapon (Gun?) Mount" are regarded in the game as two seperate spots

[jocke_p]

Hi again!

Ok i did a quick restest, 700 m open field, AT gun in tree line. All three panthers where in hull down and facing the gun.

The AT gun fired 22 shots:

3 misses

1 hit the "front turret"

9 hit the "weapon mount"

9 hit the weapon.

Result : 3 panthers with their guns disabled. That is roughly 41% gun hits of all shots fired at the hull down panthers. Does not that seem a bit high?

Edit! Also, let me explain how i got the test to run that long without the AT gun being knocked out. I manually "refaced" the panthers every couple of seconds, and i also buttoned them to lessen the chance of them spotting the AT gun. The Panthers did spot the gun after a couple of minutes but lost their contact sporadicly. I was also suprised it took 5-6 hits per panther before they started popping smoke.

Does anybody here know of good historic information sources where one might find information about hit locations on tanks, i suspect it will be hard to find. I would really like to know if my hunch is correct or if weapon hits actually where as common as this.

If NOT, then Battlefront might be persuaded to adjust their center mass aiming formula and increase the hit-width from center point a bit, so that the resulting hits are a little more historical. I am not saying it is unhistorical NOW, but IF facts would support it.

BR

Joakim

PS! I will rerun this a couple of times to get better coverage.

[Redwolf]

Technically this should be called a "weapons kill". Anything that disables the main gun.

This question is more complex than people think. The first debate is whether hitting the gun mantled damages the gun. Or, since it's very thick, getting a hit on the mantled is actually a good thing.

Then there is the main gun sights. A round at the optics is bad. But the front lens is very small. And many tanks even in WW2 would allow rudimentary aiming with other sights, namely from the commander's position. CM might not allow rudimentary aiming and count the gun as disabled.

In modern tanks a hit that disabled turret traverse is counted as a weapons kill, but quite obviously if you already are in a battle you'd be more than happy to use hull traverse to somewhat aim (that isn't the same thing as an assault gun, though).

That's really the core of it: a broken thing that would make you not send a tank into battle is not the same thing as the same broken thing if you already are in the battle. Fighting on with a degraded vehicle was common, there is plenty evidence of that.

[Redwolf]

Oh, by the way, someone will be sure to mention it so we might as well get it over with here. There are no hit probabilities in CMx. The game models the flight of each shell and they hit where they hit.

This is a misnormer. It just shifts where you pick a probability model.

You have to decide which way the gun pointed before going into flight path physics. So you have probabilities for range misestimation errors, gun sight disadjustment, traverse errors, crew screwups (drivers rocks the tank while firing), lateral misaim etc etc etc. Then even after you decided which way the gun points you need a probability to have a worse or better performing round, a better or worse gun quality and gun wear. The latter will be an issue for the Russians in particular. Ever heard of thermal sleeves for tank guns? They were put on for a reason.

And that doesn't begin to touch the issue of trying to hit a moving target which has all kinds of gunner performance and estimation probabilities involved.

[jocke_p]

Technically this should be called a "weapons kill". Anything that disables the main gun.

This question is more complex than people think. The first debate is whether hitting the gun mantled damages the gun. Or, since it's very thick, getting a hit on the mantled is actually a good thing.

Then there is the main gun sights. A round at the optics is bad. But the front lens is very small. And many tanks even in WW2 would allow rudimentary aiming with other sights, namely from the commander's position. CM might not allow rudimentary aiming and count the gun as disabled.

In modern tanks a hit that disabled turret traverse is counted as a weapons kill, but quite obviously if you already are in a battle you'd be more than happy to use hull traverse to somewhat aim (that isn't the same thing as an assault gun, though).

That's really the core of it: a broken thing that would make you not send a tank into battle is not the same thing as the same broken thing if you already are in the battle. Fighting on with a degraded vehicle was common, there is plenty evidence of that.

Well, i would just like to point out that CMx2 tracks damage to sights, main gun and weapons control, all of these systems took damage when taking a gun hit (i am not sure about weapons control though), during my battles/tests. I did notice that the main gun was almost always disabled when a gun hit was scored, sometimes other systems as well. subsequent gun hits just worsened the damage to all the subsystems. The first thing to go was almost always the main gun itself, sometimes leaving optics and such as yellow (lightly damaged).So i am guessing if a tank took heavy damage to the sights but the main gun was intact it would still be able to fire in CMBN.

My point though is that i would like to know if there is some historic/realism similarities with the results i am seeing. 40% off all shots fired hitting the weapon on a hull down tank. If there is not i would hope the aim/hit formula

will be modified accordingly. If these are historical/realistic results then i will back down and bow to the gods of reality

BR

Joakim

[Blackcat]

This is a misnormer. It just shifts where you pick a probability model.

You have to decide which way the gun pointed before going into flight path physics. So you have probabilities for range misestimation errors, gun sight disadjustment, traverse errors, crew screwups (drivers rocks the tank while firing), lateral misaim etc etc etc. Then even after you decided which way the gun points you need a probability to have a worse or better performing round, a better or worse gun quality and gun wear. The latter will be an issue for the Russians in particular. Ever heard of thermal sleeves for tank guns? They were put on for a reason.

And that doesn't begin to touch the issue of trying to hit a moving target which has all kinds of gunner performance and estimation probabilities involved.

Ok, and CMx2 models all that does it? Really? Including the persistant variables?

[TheVulture]

Ok, and CMx2 models all that does it? Really? Including the persistant variables?

Nope (well, as far as I know), and I don't think that redwolf was claiming that it does. The point is that CMx2 may well do away with a "hit probability" in the sense that it does some complex calcualtions, comes up with a probability, and then generates a random number to see if it hit or miss, but what it does do is functionally the same. The code for the gunner might decide on an aim point, assign a random error in how well the gunner can acquire that exact aim point (either taking such factors as redwolf listed into account explicitly or by fudging it with a 'design for effect' plausible spread of errors, and then maybe add in some random spread to the shots from the actual (rather than intended) barrel position to account for other randomness (amount of propellant in the round, wind, wear and heating in the barrel, whatever) which assigns a second 'design for effect' spread of plausible errors. Or the randomness in aim error and the randomness in shot spread are rolled in to one uncertainty.

At any rate it doesn't matter much; the net result is that you pick a random number (or two, which is mathematically equivalent) and that ultimately determines whether you get a hit or not. So while there isn't an explicit "hit probability", there is still an implicit one, albeit in a form that you can't easily say which ranges of random numbers will result in a hit, or what the probability is.

[Blackcat]

"My point though is that i would like to know if there is some historic/realism similarities with the results i am seeing. 40% off all shots fired hitting the weapon on a hull down tank"

I suppose the question could be rephrased into what % of the target area is made up by the gun in a head on aspect. Is it 40%? Seems unlikely, on fag-packet calculation I would think it no more than 5%. However, that 5% is smack in the middle of the target area, which does seem to suggest one of two things; the gunners are very good or Redwolf's complex probability model to decide on the precise point of aim is perhaps not as complex or as probabalistic as one might think (and neither, perhaps, is the in-flight balistics model).

The only one of those options a user can test is the crew quality. Do you get the same sort of results with Green, Regular, Veteran etc. crews?

As for confirmation of your results by a historical facts, you might well be out of luck. However, if there ever was a WWII study on such detailed accuracy of AT guns one of the grown ups on this site will know of it.

Good luck.

[Blackcat]

Thanks for that, Mr. Vulture. I am relieved to know that all this talk of complex modelling actually comes down to picking some random numbers within a defined spread. It certainly makes the results I see in game a lot more understandable, for all Steve's talk about gunner memory and the rest of it. My guys just keep rolling ones when they need sixes.

[WineCape]

I zoomed in on a scrap yesterday and the sequence went:

1. Under

2. Over

3. Under - exactly where round 1 landed

4. Way over - much further than round 2

5. Under - back to exactly the round 1 spot

6. Hit

I have got used to the fact that my Sherman drivers can't manage a clutch or steer in a straight line, but to have gunners who have worse short-term memory than I do, is bit galling. Perhaps the reason I have never seen the ultra-aggresive bailed out crew syndrome is that all my crews forced to evacuate are still looking to sort out their zimmer frames when they are mowed down (the occasional survivor I put down to the German infantry laughing so much they can't shoot straight).

Well, given ideal firing situations I have these test results on record

Setup:

-- All regular tankers;

-- M4A3 Shermans versus PzIV(H)'s;

-- All AFV’s are immobilized, each in separate lane, non hull down;

-- Distance 800m, all tanks unbuttoned, facing each other head-on;

-- Only counted hits on AFV’s, not on bailed crews;

-- "Billiard table" ground conditions between opposing AFV's, no trees;

-- 8 x Bosche tankers squaring off against 8 x USA noddies in separate lanes in each Test below:

RESULTS: REGULAR TC's @800m

1st Test: 19/37 misses fired on both sets of AFV’s (9 German/10 USA)

2nd Test: 12/28 misses fired on both sets of AFV’s (5 German/7 USA)

3rd Test: 27/40 misses fired on both sets of AFV’s (10 German/17 USA)

4th Test: 18/32 misses fired on both sets of AFV’s (10 German/8 USA)

5th Test: 28/46 misses fired on both sets of AFV’s (13 German/15 USA)

Total AFV shots = 183

Total misses @800m are 104/183 = 56,8%

NOTE: 5 Allied gunners + 1 German gunner went 4 x Misses in a row

Stressing the point again: Above results are in ideal firing conditions; that is good weather, motionless, immobilized, etc.

Add, increase or change the above parameter(s) and you will find even greater % misses.

[jocke_p]

"My point though is that i would like to know if there is some historic/realism similarities with the results i am seeing. 40% off all shots fired hitting the weapon on a hull down tank"

I suppose the question could be rephrased into what % of the target area is made up by the gun in a head on aspect. Is it 40%? Seems unlikely, on fag-packet calculation I would think it no more than 5%. However, that 5% is smack in the middle of the target area, which does seem to suggest one of two things; the gunners are very good or Redwolf's complex probability model to decide on the precise point of aim is perhaps not as complex or as probabalistic as one might think (and neither, perhaps, is the in-flight balistics model).

The only one of those options a user can test is the crew quality. Do you get the same sort of results with Green, Regular, Veteran etc. crews?

Good luck.

Good point, all my test have been with regular AT gun crews. I will try different experience setups.

I will be very interested to hear from the more educated in the subject, perhaps Battlefront staff might shed some light on the matter. Does these figures match up to their design estimations?

BR

Joakim

[Blackcat]

WineCape,

I don't mind units missing, given the period such things are to be expected (though for a change I wouldn't mind having some of those tanks that people were on about here just a couple of weeks ago, you know, the ones that always scored first round hits whilst on the move). What I was getting at in that post is this business claimed by Steve that gunners have memory. If they do it is on a par with goldfish.

However, Mr. Vulture's explanation that the point of aim is decided upon by a process functionally identical to rolling fair dice makes perfect sense in that it is a theory that fits with experimental results.

Anyway my re-enforcements have turned up in Herr Blucher's "Bloody Dawn" scenario (which is a cracker) and they include a fair number of tanks. So I am off to roll some more dice.

[Magpie_Oz]

This question is more complex than people think. The first debate is whether hitting the gun mantled damages the gun. Or, since it's very thick, getting a hit on the mantled is actually a good thing.

I think this supports the notion that hits to the main gun/mantlet were highly likely. You might find that most tanks heaviest armour was the turret front/mantlet as this was the part of the tank most likely to be exposed.

Then there is the main gun sights. A round at the optics is bad. But the front lens is very small. And many tanks even in WW2 would allow rudimentary aiming with other sights, namely from the commander's position. CM might not allow rudimentary aiming and count the gun as disabled.

You wouldn't need to hit it optics dead on tho' a shot nearby at the very least will put the gunner off his aim and at best rattle the bejeezus out of the optics. I don't think any WW2 era tanks had the ability for the commander to lay and fire the main gun. The only one I know of is the PzIV which had a hull/turret reference indicator that allowed the commander to know the direction of the turret relative to the hull at a glance.

In modern tanks a hit that disabled turret traverse is counted as a weapons kill, but quite obviously if you already are in a battle you'd be more than happy to use hull traverse to somewhat aim (that isn't the same thing as an assault gun, though).

I would think it would be almost impossible to aim the main gun using the tracks. They just aren't that sensitive and the gunner would have to be coaching the driver onto target, running away is a more likely option. In assault guns the gun was not fixed, merely in a limited traverse so the tracks where used to point in the general direction but the fine aiming was done by traversing and elevating the gun, much like an ATG.

[gunnergoz]

Shermans had a commander's vane sight in at least some versions.