First test with Fireflies vs King Tiger

[StellarRat]

Seems like a fair question.

[StellarRat]

I just realized that Vanir's 100% hits test was done at 500m not 1000m. It wouldn't surprise me in the least if he got 100% hits. I'm pretty sure I could hit something that big with a scoped rifle 100% of the time at the same range. Tests would have to be conducted on a smaller target at a longer range to be worthwhile (matching the test range parameters as closely has possible.)

[sburke]

The game engine is now so stable that we are into the territory of tweaking fairly esoteric and sometimes even rather subjective variables.

I get the feeling that the work that could be put in to test out some of these scenarios far outweighs the results of the eventual changes, and it is far from guaranteed that BFC will even bother tinkering with this stuff on a pretty stable and complete game engine.

Its a shame because sometimes these arguably minor issues can change the flavour and even the results of particular games, but none by itself is the sort of thing that will have anyone ragequitting.

LOL kind of where I am at but I have to say I really do love that there are folks who do really care about this level of minutae. The fact that they can even be arguing about this says a lot for what we expect of the game and to a large degree what it delivers. I'd only ask guys to take a deep breath. You may not be getting the answer you necessarily want but I have to say I think Phil is actually doing a very good job of trying to give some guidance on what kind of testing BFC would want to see before they start seriously looking. Just look at some of the other "bugs" that have been posted about recently with little or no testing. Also keep in mind what many of us are hoping - they are hard at work trying to get going on the MG module and do not have spare resources at the moment. The thread is only 3 days old- far too early to get this worked up.

Hell we kept them busy enough with a few dozen threads because we didn't read the recommendation to pull our mods before loading and not making sure our opponents in PBEM games have also upgraded. What did we have a dozen, two dozen threads about Vins mod cause we couldn't be bothered to read the forums or the install notes?

Again though thanks for being this revved up about the details. Seriously. Without this kind of passion CMBN likely wouldn't exist and we'd have settled for something less. One question that comes to mind off hoolaman's musing, what kind of variance would you expect on these tests? What is the variation you are expecting and not seeing that is causing you to run these tests - just curious as I have absolutely no knowledge on the subject.

[Vanir Ausf B]

I broke down and am now running tests against partial hull down KTs at 700m (partial because putting a 2m berm in front of the KT gives partial, a 3m berm blocks LOS and I don't feel like moving them around to get it just right). Early results show accuracy is less than 100%

[akd]

Jon, that is true. However, the US tests at Isigny -- conducted against actual Panther tanks at ranges not exceeding 600m -- are suggestive of a dud-rate-that-has-a-noticeable-effect.

There is an important footnote to that test:

(6) In contrast to the results obtained in this test with 17pdr SABOT, in firing conducted by First U.S. Army at Balleroy on 10 July 44, 5 rounds were fired at the front plate of a Panther tank at 700 yards. Examination of pictures of this firing indicates that the first round struck the mantlet, the second between the track and the nose plate, the third at the junction of the nose and glacis and penetrated. The fourth and fifth were fair hits on the glacis and both penetrated. The conflict between these results and those obtained by the board is explained by Col. A. G. Cole, Deputy Director of Artillery, Ministry of Supply. Col. Cole witnessed part of the test and states that the ammunition lot furnished the board had not been proof fired. He further states that, in his opinion, the lot is of sub-standard manufacture and if proof fired would not have been accepted.

Could be total BS on the part of the Brit officer, but has to be considered.

Two other points on the Isigny tests:

1. 17pdr sabot was also tested at 700m and 800m.

2. Hits were recorded as hits to the aiming point (glacis plate), not hits on the targeted tank. Many "misses" still struck the targeted tank. This in no way undermines the poor performance of the 17pdr SABOT relative to the other rounds in this particular test, but I've often seen the results of this test interpreted as 17pdr sabot only hitting a hull-up panther with 57% of the shots fired.

http://wargaming.info/1998/us-army-1944-firing-test-no3/#more-89

[Vanir Ausf B]

I've ran 30 iterations so far. Overall hit rate is much lower. I will need a much larger sample size and I plan to do more testing over the weekend. But I will throw out what I have so far. Again, this is 5 Firefly VCs, regular/normal/+0, firing at 5 partially hull down KTs at 700m. The KT crews are regular/fanatic/+0 and have short cover arc to prevent return fire. I reloaded save games except that I did restart the scenario every 10 iterations. I understand this is not up to Phil's standards. I don't care. I value my sanity more than this game

AP

Hits: 12

Misses: 73

14.1%

APDS

Hits: 25

Misses: 40

38.5%

I'm surprised at the numbers. Not only that there is a significant difference, but that it's in the opposite direction of what I would have expected.

I will update the numbers this weekend as I get more results.

[StellarRat]

It's the opposite of what I would expect too based on your documentation. There might still be a problem. Everything I've read says WWII APDS was very inaccurate. I am relieved to see misses though.

[Vanir Ausf B]

Yep, but don't get too excited yet. Those numbers could change a lot as the sample size increases. I plan on doing a minimum of 120 iterations, eventually. That way we can compare how much variation there is between each group of 30 iterations.

[JonS]

Everything I've read says WWII APDS was very inaccurate.

You're reading the wrong things then.

It was imprecise when the sabot didn't work as advertised, but accuracy wasn't in doubt.

[StellarRat]

Maybe my "readings" took that into account in their numbers. Admittedly its been a very long time since I've read a lot about WW II.

[JonS]

accuracy != precision

[hoolaman]

Of course that's true, but it seems a bit silly to say that a tank gun is imprecise but accurate. It only needs to be imprecise enough to miss most of the time and nobody would be calling it accurate.

A shotgun firing at a tennis ball 100m away could be accurate but imprecise too.

[JonS]

but it seems a bit silly to say that a tank gun is imprecise but accurate.

Statistics. It's what helps us understand the world.

[Vanir Ausf B]

Data dump. 30 more iterations. Conducted the same as the last 30.

AP

Hits: 12

Misses: 77

13.5%

APDS

Hits: 25

Misses: 36

41%

Results are very close to the last batch. So close that I doubt we will see much difference with further testing, but I will do 30 more and see what they look like.

The cumulative first shot hit % after 300 shots fired is 13.79% for AP, 39.68% for APDS.

First results post

[akd]

Very interesting. Can you also attach the test scenario you are using to the thread (or share by dropbox if too big)?

I was also playing around with a test and found that you can also place a wall in front of the KT to get an almost perfectly hull-down target presentation. I ran this setup (800m range) several times just to get a feel for how it was working. Didn't record stats, but I did note that one first-round APDS shot that missed did so 500m short of the target, which is the most extreme error I've every seen with tank guns firing at this range. Could just be a fluke range error, but might be worth tracking whether or not any of the APDS shots that miss do so with extreme errors.

[Vanir Ausf B]

If you go in under scenario author test and use the Fireflys' target tool it says partial hull down, but that is a 2m berm in front of it and although I was not tracking hit location I don't recall ever seeing a hit anywhere other than on turrets. Given the very low hit % for AP I would not want them any more hull down anyway.

http://www.2shared.com/file/LvlU5jcE/gunnery_APDS_700m.html

This includes the last save game file I made so you can see exactly how I ran it.

[Vanir Ausf B]

So, any word on this? Did anyone take a look at the test? Do I need to run more iterations? 300 shots is a solid sample size, and the lack of variation between the first and second batch of 150 suggests more testing is unlikely to reveal markedly different results, but I will do it if needs be.

[akd]

The sample size is only 300 if you are looking at first-round accuracy broadly. For purposes of this discussion, it is a comparison of 126 samples to 174 samples. Regardless, I doubt the numbers would change much with more tests and I think it is safe to say that when firing on a target the size of a King Tiger turret at 700m, the APDS ammo has a significantly higher chance of a first-round hit than regular AP.

[JonS]

I think it is safe to say that when firing on a target the size of a King Tiger turret at 700m, the APDS ammo has a significantly higher chance of a first-round hit than regular AP.

As it should be. MV of APDS is over 33% faster than APCBC or AP. Yes; MV isn't the only factor to consider, but it is an important one. If (and granted that's a big 'if') the APDS petals work as designed then APDS should have a significantly higher chance of a first-round hit than regular AP.

[akd]

I am making no comment on the numbers themselves, just on the validity of the test.

I am currently looking for a WO report that had a comparison of 17 pdr APCBC and APDS first-round chances to hit, but I can't find the original reference (seems I failed to bookmark it!) The chance to hit for 17 pdr APDS was higher than APCBC at 1000m, but the spread between the two was much less. If I recall, it gave 15% for APCBC and 19% for APDS.

[c3k]

(The chance for second round hit would, I think, strongly favor APCBC. It was rare for anyone in the tank to "sense" the fall of APDS when it missed. APCBC was more easily seen when it missed. Now I just need to find the references where I picked that information up.)

[PzKpfw 1]

Forgive me for resurecting this thread. We covered 17lb APDS accuracy issues a long time ago or so i thought till i found this thread.

Concerning 17lb accuracy Vanir is correct, it should be very innacurate above 500yrds. This is supported by fact. Ie, 12th Army Groups live fire tests in July August 1944, @ 700-800yrds 2, APDS rounds out of 5 fired hit the Panther Ausf A glacis out of 5 fired the other impacts were all over the place, the gunners were rated as 'superior'. APDS accuracy or lack of did not impress the US team observing the tests, the innacuracy was blamed on bad ammo. Their were 4 joint LF tests conducted in July August 1944 vs Panther Ausf As. The most faumous being Insigny.

For further documentation on accuracy we also have: W/O 291/1263, & W/O 165/135 dated 22.09.44 on the Sherman VC Firefly's 17lb firing APC & APDS useful range was 900yrds for APC & 450yrds for APDS. The documents detail live fire test hit% results vs an standard British 6' x 6' target, below are the results:

400yrds:

APC - 90.5%

APDS - 56%

600yrds:

APC - 73.0%

APDS - 34.2%

800yrds:

APC - 57.3%

APDS - 21.9%

1000yrds:

APC - 45.3%

APDS - 14.0%

1500yrds:

APC - 25.4%

APDS - 7.1%

These reports were written after the LF tests. Their should not be any 20 out of 20 hits etc firing APDS. APDS accuracy should never be better then APC/APCBC. Their is no need for a large statistical sample, etc. APDS innacuracy is well documented in fact its the reason the US & Soviets stuck with composite sub-calibre penetrators into the 1950s.

Regards, John Waters

[Vanir Ausf B]

Vanir is correct

I'll resist the temptation to add this to my sig. Thanks for the input.

[c3k]

Thanks for resurrecting this. The eternal flame and all that...

I've run across SECONDARY references regarding those WO results, but they are buried somewhere under other tomes.

The problem with the APDS rounds was sabot separation characteristics. Getting the sabots to peel off reliably and consistently and simultaneously was quite hard. (I hope I'm not repeating upstream information.)

Please don't let this thread die out...

Ken

[PzKpfw 1]

Since were discussing APDS, I was finaly able to track down some production numbers on German PzGr.40 ammunition. Anyone familiar with this from the past discussions on modeling it in CM or includeing it, at all. the main contention against it, was no one had any hard data. I found PzGr.40/44 used in some German AA reports concerning the Pzkpfw V, & VI, & refrence to PzGr. 40/45 but nothing concluseive for game terms.

Below are PzGr.40 production numbers etc i could find:

1942 - 800 rounds

1943 - 8,900 rounds

Concerning PzGr.40/43 for the KwK.43, 5,750 rounds were produced. Of the 8,900 PzGr.40 produced for all guns, in 1943, 5,570 were expended in combat, 1,670 rounds were returned to factories & later used in the manufacture of machine tools. Their is no data to date, i can find, on how many of the 5,750 KwK.43 PzGr.40/43 rounds were expended, or returned to factories etc.

The source listed for the data is: Fritz Hahn's Waffen und Geheimwaffen des Deutschen Heeres 1933-1945.

Regards, John Waters