Panzer IV H frontal hull armor seems wrong (test inside) BFC please take a look!

[noob]

Again I'll be happy to send anyone who wants the map and the save I used to do my tests, show me what is different about my test other than it's 1500m vs. 1000m if I'm reading above correctly. It comes out conclusively in favor of the Shermans every time. One other difference in mine is everyone is buttoned.

I have sent you a PM, i will need the map you are using as i am getting the same PzIV dominance at 1000m as i was getting at 1500m, albeit with slightly more PzIV casualties, and these are buttoned and unbuttoned tests.

[noob]

ok noob iam sorry. i`ve just repeated my test using 5 lanes in cmfi and counting retreats and everything. the results are similar to yours. the pzIV spotted faster and showed a much deadlier performance at the range of 1000m distance.

CMFI

setup:

1 platoon of Panzer IV H (early) against 1 platoon of sherman 75mm m4a1 (mid). each pair of tanks seperated with a lane. so we get five single duels each round. i`ve unbuttoned each tank manually (using hotseat) after each minute and deleted paths of movements (just like noob in his tests).

Distance between the tanks: around 1000m.

results:

after 8 rounds

shermans:

retreated: 12

destroyed: 11

knocked out: 1

abandoned: 4

panzer IV:

retreated: 5

destroyed: 3

knocked out: 2

abandoned: 3

my two cents:

well just like earlier tests by noob and other people my test showed that panzer IV optics are really modeled. they are superior over the m4a1s at ranges of around 1000m.

still the question about the availability of m61a1 ammo for the cmfi period remains. because clearly the shermans ingame are using m61a1 according to their penetration performance.

furthermore iam still advocating the increase of panzerIV upper frontal hull (transmission cover) to 22mm and 25mm for the model G. I`ve seen a lot of hts against the transmission cover in the test and all were penetrations

What is the difference between a knocked out tank and a destroyed tank ?......do you mean immobilised for knocked out ?

[Vanir Ausf B]

I have completed 2 tests. The setup is 10 Pz IV h late vs. 10 Sherman M4a1 mid on 10 isolate lanes. Range 1500m. I ran each test 20 times on Iron difficulty hotseat. Each iteration was let run for 3 turns and then I cease-fired.

First test: all tanks buttoned

Pz IV H

Tanks lost: ....... 28

Tanks retreated: 6

Total: .............. 34

Sherman M4a1

Tanks lost: ....... 62

Tanks retreated: 45

Total: .............. 107

Second test: all tanks unbuttoned

Pz IV H

Tanks lost: ....... 68

Tanks retreated: 8

Total: .............. 76

Sherman M4a1

Tanks lost: ....... 64

Tanks retreated: 27

Total: .............. 91

The much larger number of Shermans counted as retreated instead of destroyed is most likely due to its ability to pop smoke, which the Pz IV lacks.

[vossiewulf1212]

Sent off stuff to Noob and Siffo, I'd like it if something was wrong and IVs won at range, as that is what reality would suggest. But even in regular gameplay on my map the IVs consistently lose at ranges >1200mm or so, or at best come close to breaking even.

[noob]

Sent off stuff to Noob and Siffo, I'd like it if something was wrong and IVs won at range, as that is what reality would suggest. But even in regular gameplay on my map the IVs consistently lose at ranges >1200mm or so, or at best come close to breaking even.

After looking at the .bts file you sent, your test is flawed in many ways, firstly if you are testing which tank is dominant at certain ranges, you have to make sure that nothing but the gun optics and armour thickness are determining the results, therefore in my i tests made sure that both sides were fielding similar types, i.e. all the Axis tanks were PzIVH (early) types against M4A1 Sherman (mid) types, all the tanks were on the same terrain, and all the tanks were in the same formation, in your test, you have two types of PzIVG, late, and latest, and two types of M4A1 Sherman, mid and early, tanks on differing slopes, and one side in one formation and one side in another.

So, to get a true tank v tank performance based on gun optics and armour thickness, you need to remove as many other variables as possible, therefore you need to set up a test with tanks of the same type, on a flat plain, with clear weather, dry, and no wind, and a hot seat game so you can see both sides casualties, then the tanks combat effectiveness will be determined purely by gun optics and armour thickness only.

[siffo998]

What is the difference between a knocked out tank and a destroyed tank ?......do you mean immobilised for knocked out ?

well i do not really know what the term "knocked out" means. iam just using game terms but its certainly not immobilised. the game uses the terms destroyed and knocked out. i think that this is only relevant in campaign games when panzers could be repaired between scenarios. i think a knocked out tank could be repaired between the scenarios a destroyed tank is smashed beyond repair...

i will give noobs file a try after work... but if it really is like noob said than the ground elevation could play a large factor in the outcome of the battles. the ground elevation most likely support the already sloped armor of the sherman which gives this side a favor. thats another problem in the CM engine: gun elevation is not taken into account. this way a tank can fire at an enemy while standing in an impossible angle and using the extra sloping of the armor.

[noob]

well i do not really know what the term "knocked out" means. iam just using game terms but its certainly not immobilised. the game uses the terms destroyed and knocked out. i think that this is only relevant in campaign games when panzers could be repaired between scenarios. i think a knocked out tank could be repaired between the scenarios a destroyed tank is smashed beyond repair...

i will give noobs file a try after work... but if it really is like noob said than the ground elevation could play a large factor in the outcome of the battles. the ground elevation most likely support the already sloped armor of the sherman which gives this side a favor. thats another problem in the CM engine: gun elevation is not taken into account. this way a tank can fire at an enemy while standing in an impossible angle and using the extra sloping of the armor.

You have to create a level playing field if you want to isolate certain vehicle attributes for comparison, as any deviation from that spoils the experiment, and therefore the results, so once the results of combat between different types of vehicles is established in "laboratory" conditions, the players then have a benchmark figure to as a reference to take into combat and a probable non level playing field, after that, it's skill and experience that will allow the player to augment that number with terrain effects (partial / full hull down and optimum ranges), or, with a lack skill and experience, diminish that number by bad positioning (non hull down, numerical inferiority, etc).

[Vanir Ausf B]

Pz IV J vs. M4a1 @ 1500m. Unbuttoned.

Pz IV

Tanks lost: ....... 55

Tanks retreated: 11

Total: .............. 66

Sherman

Tanks lost: ....... 59

Tanks retreated: 36

Total: .............. 95

[noob]

Pz IV J vs. M4a1 @ 1500m. Unbuttoned.

Pz IV

Tanks lost: ....... 55

Tanks retreated: 11

Total: .............. 66

Sherman

Tanks lost: ....... 59

Tanks retreated: 36

Total: .............. 95

What are the specs of the test ?

[Vanir Ausf B]

Exactly the same as the previous tests I posted further up on this page, except that the IV Hs are swapped out for Js.

[noob]

Exactly the same as the previous tests I posted further up on this page, except that the IV Hs are swapped out for Js.

Ok, but rather than have each tank in a separate lane, which is highly unlikely in a CM game, remove the walls, and let the tanks slug it out in platoon sized groups (5 v 5), that way you have a benchmark figure for a more likely game situation, i did company v company (15 v 15) to simulate the scale of the tank battles in one of my forthcoming operations.

[Vanir Ausf B]

That introduces a snowball effect without telling us anything we can't learn from isolated lanes. At times I have used the type of test you refer to when testing large groups of infantry in order to generate huge sample sizes, but when the number of units are small enough be make it practical I prefer isolated lanes.

[noob]

That introduces a snowball effect without telling us anything we can't learn from isolated lanes. At times I have used the type of test you refer to when testing large groups of infantry in order to generate huge sample sizes, but when the number of units are small enough be make it practical I prefer isolated lanes.

But the snowball effect is something you get in a battle, so i would want to know which type of one sides tanks can achieve that effect against what type of the other sides tanks, then you will have a reference point when you get into a similar situation in a battle.

However the lanes test gives you an idea who will win if there is one tank versus one tank in a battle, which is a common thing too, and one type of test should inform the other type. which it does in the case of PzIV's v Shermans at 1000 - 1500m, so either test type will do.

[siffo998]

Ok, but rather than have each tank in a separate lane, which is highly unlikely in a CM game, remove the walls, and let the tanks slug it out in platoon sized groups (5 v 5), that way you have a benchmark figure for a more likely game situation, i did company v company (15 v 15) to simulate the scale of the tank battles in one of my forthcoming operations.

a one on one situation is much more likely ingame than a 5 v 5 or 15 v 15 situation. and again: by testing 5v5 you have to many variables ingame apart from the performance of the tank... for example leadership (platoon level) etc...

[Vanir Ausf B]

If you are going to design or play a large armor heavy scenario or QB I can understand the utility of testing the way noob is. But I agree that in a more typical combined arms CM game armor engagements tend to play out as a series of 1v1 or 2v1 encounters over the course of the game rather than all the tanks on each side opening fire on each other at the same time.

[noob]

a one on one situation is much more likely ingame than a 5 v 5 or 15 v 15 situation.

Given the bulk of the CMFI scenarios are no greater than medium this is true, the road test is better for those games.

and again: by testing 5v5 you have to many variables ingame apart from the performance of the tank... for example leadership (platoon level) etc...

Incorrect, in my platoon and company sized tests, the crews are all regular, with normal motivation, and 0 for leadership, which gives a base line to work from when having to play with a more mixed set of crews in a typical tank platoon.

[vossiewulf1212]

That introduces a snowball effect without telling us anything we can't learn from isolated lanes. At times I have used the type of test you refer to when testing large groups of infantry in order to generate huge sample sizes, but when the number of units are small enough be make it practical I prefer isolated lanes.

Lanes is testing to isolate single variables which are the source of the problem. I and noob are doing integration testing, testing what players actually see in game. The snowball effect is very real in game, and therefore should be part of a test to see what the real outcomes the player can expect.

I understand what Siffo said about my test above, but none of the factors he mentioned seem to me to be anything that should radically effect the outcome. IMO seeing 50% in highly unrealistic, controlled lanes 1v1 tells you much less about what will really happen in game than what I did, which is to take two regular tank companies, whatever the game decided that makeup should be, and place them face to face in totally equivalent but not table-top flat terrain and let them go at it, just as they would in game.

My results and as I said earlier actual playtesting with this map shows me that the IVs almost always lose, and if they have a very good day, break even. I really don't care that under completely controlled 1v1 lanes they come out even or the IVs slightly ahead, because that will never happen in game. I do care that in what seem to be very even but realistic game situations, the IVs always lose in my tests.

[Vanir Ausf B]

Needless to say I think your approach is completely wrong and will end up going nowhere. However...

My results and as I said earlier actual playtesting with this map shows me that the IVs almost always lose, and if they have a very good day, break even. I really don't care that under completely controlled 1v1 lanes they come out even or the IVs slightly ahead, because that will never happen in game. I do care that in what seem to be very even but realistic game situations, the IVs always lose in my tests.

If you think that your testing has demonstrated a problem tell us what the problem is exactly and what needs to be changed.

[siffo998]

Lanes is testing to isolate single variables which are the source of the problem. I and noob are doing integration testing, testing what players actually see in game. The snowball effect is very real in game, and therefore should be part of a test to see what the real outcomes the player can expect.

I understand what Siffo said about my test above, but none of the factors he mentioned seem to me to be anything that should radically effect the outcome. IMO seeing 50% in highly unrealistic, controlled lanes 1v1 tells you much less about what will really happen in game than what I did, which is to take two regular tank companies, whatever the game decided that makeup should be, and place them face to face in totally equivalent but not table-top flat terrain and let them go at it, just as they would in game.

My results and as I said earlier actual playtesting with this map shows me that the IVs almost always lose, and if they have a very good day, break even. I really don't care that under completely controlled 1v1 lanes they come out even or the IVs slightly ahead, because that will never happen in game. I do care that in what seem to be very even but realistic game situations, the IVs always lose in my tests.

ok i`ve just checked your map and first: really nicely done. iam looking forward to do a armored engagement PBEM on that map.

second: after doing what you said in the mail: hitting the go button 10 times and then hitting cease fire and look at the results. the game showed me 10 dead shermans and 9 panzer IV.

but i think i`ve found another flaw in this whole testing. i guess you used the german side vs the ai. this way by only hitting the go button without giving orders, your hit panzer IV will just retreat behind cover and never support the rest of the forces. the ai shermans on the other hand will move forward and put huge pressure on the remaining panzer IVs...

i suggest that you use hotseat for testing and cancel after each go all movements and unbutton the tanks. also you should move retreated tanks into contact so that they can attack the enemy again...

still there are a lot of UPPER FRONTAL HULL PENETRATIONS that lead to destruction or retreat. i would strongly suggest that BFC increases the UFH to a historical level... 22 and 25mm... that should give the pzIV a little bit more survivability...

[JonS]

I do care that in what seem to be very even but realistic game situations, the IVs always lose in my tests.

Fair enough.

The obvious question, though, is why is that? What factors are in your test that are absent from Vanir's, and what is the causal mechanism that changes a significant advantage in Vanir's test to a significant disadvantage in your test?

Edit: sniped. What Vanir and Siffo said.

Edit2: And the causal mechanism seems clear. Vanir has shown that the PzIVs have a distinct technical advantage (especially at long range when buttoned, and possibly also when unbuttoned), but you've put them at an overwhelming tactical disadvantage. In your test the tactical seems to outweigh the technical, but that tells us nothing about what the technical should be. It just tells us that tactics are important.

[noob]

I understand what Siffo said about my test above, but none of the factors he mentioned seem to me to be anything that should radically effect the outcome.

You are using a form of the arguement from ignorance fallacy:

http://en.wikipedia.org/wiki/Argument_from_ignorance

Because you cannot see why there is anything that could radically effect the outcome apart from the type of tanks used, you decide their cannot be anything, no other explanations are valid, which is irrational given the amount of variables you have introduced into the test.

For example, take the formations, because the PzIV's are tightly packed, the Sherman's only have to move their turrets slightly to move from target to target, whereas the PzIV's have a more widespread group of targets to engage, therefore they need to make bigger adjustments, thus taking longer, and when those differences are compounded, they "could" make a difference.

However one cannot say for certain this is the case, therefore the differences in the formations of the opposing tanks must be eliminated to remove that uncertainty, the same goes for the terrain, and the particular model of the tanks on the same side.

Therefore all your test shows is the outcome if two players positioned their mixed type tanks on that map in the way you have positioned them, there's no way to "know" for certain if the advantages one group have over the other are to do with optics, terrain, tank type, or formation, that is why one needs to create a flat featureless surface with the tanks of the same type per side, lined up facing each other in the same formation, that way the only thing that could possibly effect the initial volleys are the gun optics and the penetration effects, and nothing else.

One deviation from this is when the Sherman's pop smoke, when that happens the playing field is not level any more, but in all the tests, the smoke does just delays the inevitable, as the victor of the test is the group that gets the most opening hits before the smoke, which are the PzIV's, and the reason is definitely the gun optics, as all other factors are equal.

[JonS]

What you can also do is, instead of two live tanks shooting at each other, set up lanes so that a bunch of live tanks are shooting at inert targets. They can be made inert by either giving them short covered arcs, or by bailing out the crew which has the handy side effect of preventing smoke being discharged and also preventing the target from retreating or moving in any way.

Then you would count the number of target tanks KO'd in a period of time, or the total time to knock out all the target tanks, or the average time to KO a tank, or the number of rounds, or whatever you want, really.

Then compare the figure you deem worth measuring (or the basket of measures) for the Shermans with the equivalent for the PzIVs.

To really mix things up, you could also look at Shermans shooting at Shermans, and PzIV shooting at PzIV, along with the traditional Sherman vs PzIV.

[c3k]

^^^

When TARGETING is what is tested, using the SAME TARGET is important. To do so, getting the crews OUT of the target tank prevents smoke, retreating, return fire, etc.

Hotseating allows you to get the damage details from the unmanned tanks.

Platoon vs. platoon tests have little validity vis a vis targeting. It is a test merely of platoon combat worth vs. platoon combat worth.

[noob]

Platoon vs. platoon tests have little validity vis a vis targeting. It is a test merely of platoon combat worth vs. platoon combat worth.

True, Platoon v Platoon tests evaluate a combination of the accuracy of the gunners and the effects of the hits in a group conflict at a certain range, therefore, one can only ascertain from the test, which group will dominate the other in a combat situation at that range, the performances of certain tasks for individual tanks are somewhat lost in the mix.

[JonS]

Hotseating allows you to get the damage details from the unmanned tanks.

I use Scen Author Test Mode for that - it's much quicker than constantly flicking back and forth between sides. Depedning on exactly what I'm doing I then use either RT or HTH. The big problem with RT is the loss of replay. If exact nose counts don't matter then set it to RT and go make a cup of tea. If you are counting seconds or rounds you'll probably want to use H2H vs the AI.