Penetration KT vs M-18 - some strange conclusions

[MichaelU]

I've been following the debate on use of tungsten, and how powerful it is versus how available it was. I've no idea about availability, but did a quick check on penetration.

What I did was set up ten King Tigers opposite 10 M-18s on flat terrain, at a range of just under 500 metres. The M-18s had no AP, just tungsten, the weather was dry and clear and all crews were regular.

I then had each tank target the tank opposite and let rip. After the first volley I recorded the results. I did this ten times for a total of 100 shots. I then repeated the experiment, but with both sides hull down.

This led to some interesting conclusions.

On open terrain the M-18s showed a 59% chance of hitting and a kill probability of low. They hit an average of 5.6 times out of 10, with 32% of those being penetrations and 25% kills.

With both tanks hull down, the M-18 was showing a 40% chance of hitting and only a kill probability of rare.

Surprisingly though, although the hit ratio dropped to 46%, the percentage of hits that penetrated rose to 65% and kills 54%.

That means the chance of losing a King Tiger in the circumstances I described is 25% when hull down, and just 14% when in the open.

This is fairly easy to understand when watching the movies. The majority of hits in the open were to the upper hull, with less shots to the turret and even less to the lower hull.

When hull down the majority of shots were to the turret and less to the upper hull. And the King Tiger turret is very bad indeed at stopping 76mm tungsten at 500 metres, while the upper hull does the job a lot better.

The only conclusion is that when facing Allied tank destroyers, get out of hull down position because your survivability is better in the open.

Another strange feature is that even though the chance of penetration about doubled, the penetration went from low to rare, which I assume means it got worse.

[stic.man]

hmmm....interesting...

although i understand most of this post i have a few questions about several things. (im new and have yet to receive the full game)

my questions are:

-What exactly is tungsten?

-What is Hull down?

-What is AP?

thanks

stic

[Doug Beman]

First off, which version of CM are you using? Pre-1.1 have trouble doing hull-down corretly.

Second, which King Tiger were you using? One version has a lot of front turret armor, the other has A LOT of front turret armor. If you were using the former, I think its front turret armor is actually less than front hull armor, which means that making it hull down (only the turret can be hit) would actually increase the chance of a penetration for a hit. If you were using the KT with A LOT of front turret armor, it has more front turret armor than front hull, meaning a hit on the front turret should have a lower % chance of killing.

Third, I think it's logical that a penetration on the turret would have a higher % chance of a kill. There's 3 of 5 crewmen in there, along with all that ammo, etc. A penetration in the front hull could hit the steering gear/tranny (maybe not even immobilizing) or the radio, etc.

Lastly, you mention that "When hull down the majority of shots were to the turret and less to the upper hull." This is what hull down does; it gives the enemy a smaller target to shoot at, and forces them to be hitting the front turret which should be thicker.

DjB

[MichaelU]

I was using 1.1. Actually, I didn't realise there was more than one type of King Tiger, this one had 185mm at 10 degrees on the turret, and 150mm at 50 degrees on the upper hull.

Obviously it's going to be easier to penetrate the turret given this amount of armour. And I agree that a turret penetration is going to kill more often. It just seems strange to me that it is safer being in the open than hull down. Like 100% safer, according to my test. Does this seem right, and would it not lead to unrealistic behaviour like purposely avoiding hull down positions?

Stic.man, in answer to your questions, AP is short for armour piercing, tungsten is special ammo that is better at armour penetration for reasons that have to do with the manufacture and composition of the shell, although I'm hazy on detail here. Hull down is when a tank hides its hull behind some obstacle, normally a hill although I think a stone wall works in CM too, leaving only it's turret and a bit of upper hull visible. This means it can still shoot but is harder to hit. And the fact that we're bringing up small details like this shows how much we love the game.

[Guest grunto]

--Second, which King Tiger were you using? One version has a lot of front turret armor, the other has A LOT of front turret armor. --

Is one turret the Porsche and the other the Henschel design?

Andy

[A Arabian]

Hrm... Seems like a weird probs problem -

Case 1:

56% chance to hit

Say 25% chance to hit turret armor, 75% chance to hit hull (making these numbers up for the sake of teh argument).

Lets also say that you have a 50% chance to kill with a turret hit, and 20% chance to kill with a hull hit:

Chance to kill=.56*.25*.50 + .56*.75*.20 = 15.4% chance to kill overall (pretty close to the numbers reported above).

But:

Case 2:

46% chance to hit, 100% chance of turret (only thing showing), and still a 50% chance to kill with a turret hit.... = 23% chance to kill.

Ok, so those numbers work out pretty close to the empirical results - but heres the rub - why is case 2's chance to hit 46%? My knowledge of probs/stats is pretty crappy, but shouldnt the chance to hit in that case be the same as the chance to hit the turret in case 1 - i.e. .56*.25 = .14 instead of .46? Or is there some nuance of the probability I'm missing?

Actually, after thinking about it, there should probably be a slightly greater probability than just the chance of case 1 since the opposing armor will be targetting the turret, probably slightly increasing the chance of hitting that in particular (since in case 1 they are targetting the vehicle as a whole rather than the turret proper presumably), but I suspect it wouldnt be to the tune of an extra 30%... Maybe an extra 4-6% making it 18 or 20% rather than 14%.

[This message has been edited by A Arabian (edited 12-12-2000).]

[This message has been edited by A Arabian (edited 12-12-2000).]

[MichaelU]

Although I unfortunately didn't record where the tanks were being hit, the numbers in the open quoted by Arabian sound about right. However, when hull down there was a small chance of an upper hull hit, about 20% perhaps. These tended to ricochet.

The chance to kill with a turret hit was substantially above 50%, although I haven't got my numbers in front of me at the moment.

I assume in a hull down position that the amount of tank showing drops by half, if not more. But the chance of hitting only fell by 16 percentage points, or just under a third in this case. I was wondering why that is.

Any tank grogs out there care to enlighten me.

[A Arabian]

Really!? I would have thought 50% chance to kill would be generous when it is reporting it as "rare".

Good call on the upper hull hits - it would definitely depend on your "hull downedness" (for lack fo a better term). The fact that they actually calculate the visible cross section makes this a much harder discussion since it adds in a bit of a variable. But even that notwithstanding, the numbers don't seem to add up the way you would expect.

[RMC]

Hmm, this is an interesting topic. I don't use either of these vehicles much and so I haven't noticed anything like this. But this thread makes me wonder about the representation of the KT armor. What seems out of kilter to me at first blush is the ease of kills on the turret front. The turret was rather narrow with large portions of the frontal profile being the turret side armor at a very steep angle. The remainder of the turret front should offer a relatively small area of vertical armor that can be hit because of the large gun mantlet. I suspect this is very much like the same issue with the Tiger turret front where the CM engine doesn't account for these variable hit locations and as such perhaps a variable strengthening is in order.

[Guest Rex_Bellator]

Yeah - be careful when ordering tanks into hull down positions. One PBEM game my opponent thought he was being clever by cresting a small rise on the map with 4 PzIVs at the same time. They had a gaggle of Shermans and light British armour to destroy at fairly long range.

In the ensuing firefight he was slaughtered as the turret armour on a PzIV is much worse than the hull and it was all we could hit. The Shermans guns were just strong enough to penetrate the turret but would have bounced off the hull. Therefore - although this may sound silly, I always use tanks with relatively weak turret armour in the open rather than hull down as it makes them much more survivable.

------------------

"We're not here to take it - We're here to give it"

General Morshead's response to the popular newspaper headline "Tobruk Can Take It"

[patboivin]

I am surprised that it is still that easy to hit a hull down tank, given that half its profile has disappeared from view.

Shouldn't it be much harder to target hull down tanks, if all you can see is their turret? I understand that Tiger Ausf B turrets were huge, maybe that's why they didn't gain much by being hull down.

What about hull down StuG IIIs, or the other low-profile tank destroyers? Are they much harder to hit while hull down?

[Jeff Heidman]

<BLOCKQUOTE>quote:</font><HR>Originally posted by Rex_Bellator:

Yeah - be careful when ordering tanks into hull down positions. One PBEM game my opponent thought he was being clever by cresting a small rise on the map with 4 PzIVs at the same time. They had a gaggle of Shermans and light British armour to destroy at fairly long range.

In the ensuing firefight he was slaughtered as the turret armour on a PzIV is much worse than the hull and it was all we could hit. The Shermans guns were just strong enough to penetrate the turret but would have bounced off the hull. Therefore - although this may sound silly, I always use tanks with relatively weak turret armour in the open rather than hull down as it makes them much more survivable.

<HR></BLOCKQUOTE>

This is defintely a problem then.

Logic would dictate that you could *never* be worse off by being hull down; at worst it would be of no benefit (assuming it is impossible to penetrate the armor you are concealing).

The odds of hitting a hull down vehicle (assuming the gunner is not able to select the point on the target he wishes to hit) whould be the same odds of getting a turret hit (or whatever the exposed portion is) had the target not been hull down to begin with, at best!

Jeff Heidman

[MichaelU]

Good point Jeff, if you can only see half the tank should you only have a 50% chance of hitting compared with in the open. I don't know how CM models it though.

One point I did notice, from memory, is that the chance of the KT hitting the M-18 fell by a little more percentage wise than it did for the M-18 hitting the KT hull down. The M-18's chance of hitting fell by 28%, while the KT's fell by 35%. This is probably because the KT has a monster silhouette of about 135 against just 85 for the M-18. Maybe someone with CM open could check these numbers, because I'm at work at the moment and can't run CM here.

Arabian, that also struck me as wierd. The kill probability fell from low to rare, and yet the number of hits that killed about doubled to just over 50%.

[A Arabian]

I think youre right - to a great extent it comes down to the fact that the OVERALL chance to hit is not reduced enough with respect to the loss of visible area.

Using the example you started with, a 56% chance to hit full visible, would in theory drop to 23% chance to hit if 1/2 the vehicle SA is blocked (thats probably hideously oversimplified though, but I'm not sure what I'm missing).

BUT - there is also a complicating element that for each hit you have a greater chance of hitting the turret, so it wouldn't drop your overall kill chance in half as you might think.

It still comes down to the fact that this is sorta suspect.

[Guest Rex_Bellator]

I agree totally. It seems that the computer normally assigns a random hit location once a hit is determined against targets in the open. It sensibly assumes a hit on a H/D target is the on the turret. The benefit for some tanks of a reduced hit chance by being hull down is more than offset by the fact that it is guaranteed to hit the turret, rather than having a (for instance) 70% chance of hitting elsewhere where the armour would be much stronger.

I suppose one answer is to make a big increase in the penalty for shooting at H/D targets. As has already been said, you should never logically be better off firing at H/D tanks.

Another is to use the same chance to hit calculations as a tank in the open. Then randomly determine a hit location from all the possible areas, and treat anything other than a turret hit as a total miss.

Hope this makes sense and I confess I don't know for sure how the computer calculates hits & hit locations, but this is how it seems to a mere avid CM fan.

------------------

"We're not here to take it - We're here to give it"

General Morshead's response to the popular newspaper headline "Tobruk Can Take It"

[RMC]

Anybody remember:

if the DR is less than the TO HIT number and the colored die of the TO HIT DR is < than the non-colored die then the hit is scored against the turret/superstructure armor...

or something like that anyway

ASL did it exactly as Rex B suggests.

[This message has been edited by RMC (edited 12-12-2000).]

[aka_tom_w]

This is a VERY intertesting thread and topic.

I would be interested to see it addressed or discussed by Steve or Charles.

If weaknesses in turret design and armour thincknees indicate that a tank is worse off when it is hull down, (as the turret is more easily penetrated then the hull) then this should really be looked into.

I do think you folks are really on to something here. I have personally found that yes indeed, especially when using tungsten a frontal hit on a turret could lead to one of two bad things, Gun Damaged, or penetration leading to a KO'd tank. If the odds favour a lower hul hit or a upper hull hit, which are both usually more heavily armoured than the turret it would clearly make sense to leave the tank NOT hull down and let the rounds (hopefully) bounce off the upper hull or lower hull as the odds of a round hitting those area's is greater than the chance of a front turret hit, I would assume.

This is a very interesting observation.

Anyone else? Comments?

Steve?

Charles?

Matt?

Dan?

-tom w

[aka_tom_w]

More calculations:

The round is fired....

First Calculation->

did the round hit the tank?

If yes Where?

If no Where did it impact the gound?

NOW

if hit = yes then where hit?)

if the AVF presents it frontal aspect?

Chance to hit track? (should be small ~5%)

Chance to hit Lower Hull? ( what%? 20 - 25%?)

Chance to hit upper Hull (40 - 50% ?)

Chance to hit Gun to cause Gun Damage? (~5% ?)

Chance to hit Turret (should be small 10- 20%)

Chance to hit Turret Ring (does the game do that? 5%)

Run the random number generator?

Determine where the round hit

with the NEW proposal Like ASL it should go somthing like this:

Where did it hit?

If the hit is a track hit or a lower hull hit

Calacualtion should Ask "Is the Tank Hull down"?

If yes then the round strikes the ground short immediatly in front of the tank. If he tank is not hull down then the round hits the tank int that area.

This way keeps the chance to hit percentage on the turret (say 10- 20% in the example above) the SAME for hull down tanks and exposed tanks.

I guess what we are asking is HOW does the game handle the location of the hit if the tank it hull down and how does it calculate the location if the tank is not hull down?

-tom w

P.S.

Legal Disclaimer:

I don't have a clue what I'm talking about.

I invented, out of thin air, ALL the numbers and hit locations percentages noted above as a purely FICTIONAL example from my knowledge of how a similiar target chance to hit system works in Avalon Hill's old game Tobruk, and they have nothing to do with how the game really works and I have NO idea how these location calculations in CMBO actually work. I posted the numbers above just an example of how the previuosly proposed ASL model of determinig hit location might work.

[This message has been edited by aka_tom_w (edited 12-12-2000).]

[Jeff Heidman]

<BLOCKQUOTE>quote:</font><HR>Originally posted by A Arabian:

I think youre right - to a great extent it comes down to the fact that the OVERALL chance to hit is not reduced enough with respect to the loss of visible area.

Using the example you started with, a 56% chance to hit full visible, would in theory drop to 23% chance to hit if 1/2 the vehicle SA is blocked (thats probably hideously oversimplified though, but I'm not sure what I'm missing).

<HR></BLOCKQUOTE>

Logically, it should work like this.

Take a tank in the open. lets assume that if you hit the tank, there is a breakdown in hit location as follows:

15% Lower Hull

35% Turret

50% Upper Hull

I have no idea what those numbers actually are, but lets pretend that I am correct for arguments sake.

The chance of a fully hull down vehicle being hit should then be 35% of the chance of it being hit otherwise. If that is the case, then the final chance to obtain a kill *cannot* be greater due to being hull down.

The point is that no matter what the real numbers are, the decrease in chance to hit should be equal to the chance of hitting the exposed surface if the vehicle was in the open. This would always result in a decrease (or, at worst the same if the covered up armor is impervious) in the overall chance of getting a kill result.

However, in the example cited, the chance of obtaining a kill is, in fact, *greater* when the KT was hull down than when it was not. This would seem to be an error, if not a bug.

But I could be missing something...

Jeff Heidman

[Jeff Heidman]

<BLOCKQUOTE>quote:</font><HR>Originally posted by RMC:

Anybody remember:

if the DR is less than the TO HIT number and the colored die of the TO HIT DR is < than the non-colored die then the hit is scored against the turret/superstructure armor...

or something like that anyway

ASL did it exactly as Rex B suggests.

[This message has been edited by RMC (edited 12-12-2000).]<HR></BLOCKQUOTE>

Ahhh, them were the days!

You know, CM is such a great improvement over ASL, but I still miss something about sitting down for 12 hours and fighting out the desperate effort to take the Bread Factory with actual counters on an actual map against an actual human sitting across the table from me...

But who has the time anymore?

Jeff Heidman

[MichaelU]

Rex, tom_w, I think you may be on to something here. I've played a couple of wargames where you first calculate whether you hit, and then where, and if the area is covered you are assumed to hit the cover. The only danger I can see here is that tanks would be plowing too many rounds into the ground ahead of the hull down target, rather than winging them over the top which would seem more likely as they would be aiming higher.

[Jeff Heidman]

It is always important to understand what basic assumptions are made in any argument.

We all seem to have agreed upon something here, but all of our arguemnts (mine included) are based on one fundamental assumption:

*****************************************

The odds of hitting a hull down target are linearly lessened in direct proportion to the decrease in target area.

In other words, if the turret makes up 35% of the target, then if only the turret is visible, the chance of hitting the target is 35% of what it would be otherwise.

*****************************************

This may not be true.

Presuambly, we can assume that a gunner is going to aim for the center mass of his target. If the target is hull down, that aim point is different, .i.e. center mass is the center of the turret, as opposed to the center of the hull. While the turret is smaller, and hence there is less margin for error, this is NOT the same as aiming for the center of the hull!

Indeed, a savvy gunner,knowing that he cannot penetrate the hull anyway, would aim for the turret regardless of hull down, trading a decrease in hit percentage for an increase in chance to hit a vulnerable spot. I doubt CM models this however.

Alternatively, think of it this way:

Lets say we have a tank where 50% of randomly distributed hits will strike the turret. Now, if (under our assumption) the odds of hitting the turret HD are always 50% the odds of hitting the vehicle, what are the odds of hitting the turret of a HD tank at 15 meters? At BEST it would be 50% (say 100% chance of a hit on a non-HD tank, 50% of that gives a 50% chance for a hit). I would claim that this is not supportable in reality, since in reality at 15m you could easily hit a turret sized object mch more than 50% of the time.

Of course, what I am saying is that the facing hit is not always random at all. Which might just open a whole 'nother problem.

Jeff Heidman

[JoePrivate]

I did a quick check. A M4A3's hit chance against a PzIVH at 500m is 47%, when the PzIVH is hulldown then the hit chance is only reduced to 30%. At 1000m the M4A3's hit chance is 22% and 13% respectively. I agree, on the surface that seems strange but there may be other things not taken into account here.

[Doug Beman]

I'm pretty sure that CM does not assign hit locations randomly. Since each shell is tracked in real time physics, the engine can actually calculate where an incoming shell would hit.

DjB

[aka_tom_w]

Well I would beg to dissagree with Doug

If "each shell is tracked in real time physics, the engine can actually calculate where an incoming shell would hit."

Then the round would not be able to travel THROUGH Live and dead AFV's that are not burning. It is my opinion that the game does not model physcis to the degree of accuracy you suggest.

I think Some random number generator spits out a value that determines the location of the hit.

how could it possibly model the accuracy to "track" where the round was actually aimed and where it subsquently hit, without the use of some form of a random chance variable being generated.

In other games like ASL and Tobruk we used to roll dice and go through a series for combat result look up tables to determine

first the result of the hit

hit or Miss?

if hit then where did it hit? (rool the dice)

when the location of the hit was determined a llook up table was referenced to determine the result of the hit at that location.

We have been told that in CMBO very complicated and state of the art algorythms replace these look up tables and can account for ANY percentage of chance or random varible between 0% and 100% including decimals and exact angles for determining penetration results.

I am of the strong opinion that the location of exactly where the round hits, once a round has been determined to have hit, it does a random number calculation based on percentages of exposed areas' (measured with painstaking accruacy I'm sure ) meaning that if for given tank the exposed area of the lower hull is deemed to be 20% and the exposed area of the upper hull is deemed to be 50% and the exposed area of the turret is deemed to be 30% of the TOTAL exposed armour that the round could hit, then THOSE are the percentages that are used to determine the chances of a round striking any particular exposed armour surface on the tank.

but that's ALL just opinion and conjecture on my part as most of the REALLY interestoing Math calculations and inner workings of the armour penetration algorythms are closed guarded trade secrets so Says BTS when we've asked about them in the past anyways.

-tom w

[This message has been edited by aka_tom_w (edited 12-12-2000).]