Matrix Games Forums

Just wondering is their any special reason why the type1 medium has an anti armor value of 80 ? It has the 47mm At gun which has anti armor 52. It also has armor 50 which seems a bit high considering it is no real sloped armor.

The Sherman in the latest update has armor 45, but in reality it has a sloped front, so should be higher. T34 should be higher too it has all around slope. Sherman 75 has anti armor now of 62, yes the 75mm had a better HE shell, but still the 75 AP should be a bit higher esp. later war...would VERY surprised if the 47mm of type1 would be better in armor pen than the 75mm US AP ? Is there a source for this?

YES, these values of the updated scen 1/2 are better than the insane values of many tanks in the old orig scen, but was wondering if there are special reasons given for above discrepancies or are they an oversight?

Also the late war SOVIET Sherman should have the 76 HV gun, but in game has the same low value as the 75mm? See here:
tm.asp?m=4274438

Well, those values are simply wrong - nothing to really say about that.
A 1944 M4 Sherman had frontal armor of 63 mm - then sloped at 47 degrees which realistically yields armor protection of 105mm on the glacis plate. The 75mm M3 gun had a penetration value of 66 to 90 mmm at 500 yards [depending on calculation methods and armor type] but realistically was good for over 90mm armor penetration at normal combat angles, that is a 20 degree or less deflection shot).
They should be updated in the next patch - if they aren't already.
The Japanese 47mm AT gun was nowhere in the same league as the 75mm M3 gun or soviet 76.2mm F-34 gun.

Still the Sherman we have now is "better" than the orig one... it had iirc anti armor 120 (!!) and armor 90 or so... (and the T34 was worse in early scen game than early Sherman in armor ! LOL) and no discernment between the later one which had also wet storage (which should in game translater to perhaps some more armor) and the early ones with bad armor quality and bad ammo storage.. check the values of my silver mod (in scen forum) these were the best I could come up with and were a compromise between the both the orig and updated scens (both wrong, but still we are now a bit better of with the updates).

Perhaps early scen was made by "AFB" and updated by "JFB" ? [:'(] Consider in game a good IJ player with good industry and VEH points can upgrade most type 97 to type 1 (which ofc was not the case in reality - iirc less than 300 were made at all, just check Wiki). Wiki gives tpye1 max armor 50, so would be even correct in game, BUT it had not a good slope anywhere. Another site said it had the 50mm only at some frontal places..other places were more thinner like T97

I found pics of Shermans however where it said, destroyed by 47mm fire (side shots ofc!)

Well, suffice to say, the game was not optimized for realistic tank vs tank combat.
Anti-tank round vs armor plate penetration and vehicle destruction is a very complicated subject, and there are no simple comparisons.

A whole lot of what is out there on the internet and even publications is oversimplified, urban legend, and/or just plain wrong - that is - without any hard historical backing/analysis.

The M4 medium in the Asia-Pacific theater was as dominating as the PZ VI Tiger was to the ETO in general.... ie. very tough to take on, but not invulnerable absolutely.
The Soviet T-34 series were as tough as the M4 Medium generally...though different.

Japan made some notable and effective machines (naval and aircraft) but never seriously challenged European or American AFV's in battlefield dominance.

B - do you have updated values for the allied AFV's that you can share? I would like to include in my mod.

ORIGINAL: PaxMondo

B - do you have updated values for the allied AFV's that you can share? I would like to include in my mod.

I'm sure I do - let me check and I'll get back to you [8D]

Hello Pax, what seemed like an easy request on the surface turns out to be much more difficult than I thought (but certainly not impossible - I just never delved into it)..

First, after looking - I don't have what you asked - pure and simple - I didn't really think about it before - I can see now.
Second, I'll look into it - should have been done long ago...

Third, just to let you understand, it isn't (or shouldn't be) just as quick and simple as just looking at numbers (as you would think).
Many years ago (before home computers) in the days of Avalon Hill's Squad Leader - I got interested in this topic. That led me to go to Cal State Fullerton's Physic's Dept to seek some answers, as well as writing to the US Army's Historical Division to get some information - through my Congressman.
The upshot answer is that I got "some" idea of how it all really works from a physics POV, and what the Army knew after WW2.

As you might guess the real answer is "messy", but kinda' decipherable, ... there are many factors.
Factors of distance (usually quoted from 500 and 1000 yards for measuring points), Angle of obliquity (of attack), plate thickness, Brinell Hardness Factor of plate and shell, Face Hardened or Homogeneous armor (and there is NO BEST quality here .. it all differs in circumstance), vehicle weight (If properly cemented - the projectile must overcome the weight of the entire vehicle - not just the area of attack) , method/quality of of attaching plates to structure,.... the damn process of determining is almost endless -as well as the fact that changing some variables changes the answer (such as caliber of projectile to thickness of plate - ..up to a point - and that is affected by velocity as well) - ...makes it all tough to come up with ONE rule.

Any way - give me some time to try to come up with something good.
But in WITP:AE terms - that shouldn't be too difficult.

B[8D]

EDIT - None of this even takes into account the combat effectiveness of what those units can and will do.
For example, Does the unit all have radios? What is the combat doctrine of the unit in question?( such as US Tank Destroyers in Normandy where apparently under-armored M-18's engaged heavily armored PZ V's, they should be decimated by the Panthers - but instead their doctrine had them effectively hit and run from cover in the opposition's under-defended flanks), etc... and the air assets ability to inform ground units on the vulnerability of the target's placement, etc...
How does one really account for the combat effectiveness of vehicles/units in question?

ORIGINAL: Big B

ORIGINAL: PaxMondo

B - do you have updated values for the allied AFV's that you can share? I would like to include in my mod.

I'm sure I do - let me check and I'll get back to you [8D]

B-
Was aware of most of this which is why I have not (cannot) do it. Not only is there all the empirical data, but never having seen the code, there is no way I can know (or even guess) how the code actually treats the values input. Gary has to have some alogrithms to approximate a lot, but what those are and how they relate to reality .... yeah. I have to defer to someone like you or John ...

Anything you can do/suggest will be welcome. Like others, I have long thought that the IJA AFV's were 'generously' graded, but without understanding how to fix it, I have left it alone.

Thanks for whatever you can do.

Well, looks like I have a project to look into [8D]

ORIGINAL: PaxMondo

B-
Was aware of most of this which is why I have not (cannot) do it. Not only is there all the empirical data, but never having seen the code, there is no way I can know (or even guess) how the code actually treats the values input. Gary has to have some alogrithms to approximate a lot, but what those are and how they relate to reality .... yeah. I have to defer to someone like you or John ...

Anything you can do/suggest will be welcome. Like others, I have long thought that the IJA AFV's were 'generously' graded, but without understanding how to fix it, I have left it alone.

Thanks for whatever you can do.

If I can be of assistance in number crunching, I am happy to help. I am pretty good at that ...

My goodness. This sounds awful like many a robust conversation held over on the Steel Panthers site 17 years ago. [X(]

Shame we don't have more synergy between the game titles so all that info isn't lost... [:'(]

I would tend to agree Reg!

If you guys come up with something useful that most agree with, I will also implement into my Mods.

Another informal source of info might be the Army's museum at Aberdeen Proving Grounds in Maryland. They have a lot of WWII armored vehicles there. As a youngster our Scout troop went there once a year (I grew up near Baltimore) and they let us climb all over the tanks. Now days, of course, that is verboten. While I was in college I worked there as an engineering co-op student, and got to see some testing of the GAU-8 (A-10 Thunderbolt II cannon) and the Bushmaster (Bradley IFV cannon). I'm sure they have changed things since those bygone days, but they used to have a display about armor penetration, and some statistical data. I don't know what their source was, but I'm assuming it was from some Army database.

At the risk of using one game's model to work out another games values, someone with a real interest in tweaking the numbers might want to check out War Thunder. They have a good gun penetration/armor model which is generally based on actual army texts and ballistics performance test reports. Free to play, and you can take a look at all the penetration/armor data without actually playing through the tank preview. Game includes pretty much all of the US, German, British, Soviet, and Japanese armor.

Mike

P.S. The game itself is great fun, too. Includes most/just about all the significant air planes for the same nations, plus the Italians.

I had a go at armor.

Personal disclaimer: I have no experience what so ever in ballistics or tank warfare – so please be gentle on me. I’m however (as Pax put it) pretty good at number crunching.

Basic philosophy: We don’t know what the engine does with the armor and penetration values anyway, so there is no point in going into complicated thinking what it might do with them. We know from the editor that armor is supposed to be a thickness value in mm. Hence I tried to compress all the armor of a tank into a “weighted effective armor thickness”.

Approach:
1. Collect armor thickness and angles (from the vertical) from internet sources for the front, side and rear. Better sources give separate values for different areas of the front, side and rear.
2. Calculate the effective armor thickness of an area by the formula
effective_thickness = raw_thickness / cos (angle)
The effective armor is hence the length of the path of a horizontally travelling projectile through the armor. For example: The front armor of an early war Sherman is 51 mm at 56 ° equals 91 mm effective thickness.
3. Calculate an “averaged direction thickness” by averaging all values for a direction (front/side/rear) evenly weighted.
4. Calculated the final armor rating by averaging the front and side values (and rounding to the nearest integer). I discarded the rear value. For example an early war Sherman has a side armor of 38 mm, so its armor rating is (91+38)/2 ~ 65 which is nicely between the old database’s inflated value of 90 and the new 45.

Some justification and room for improvements:
- Concerning point 2: This formula most likely understates the ability of angled armor to bounce a projectile off.
- Concerning point 2: Angles are rarely constant. The round nose of the Sherman for example contains every angle between 56° and 0°. In cases like these I generally used the higher angle here, partially hoping this would cancel with the last point.
- Point 3 could be improved by weighting the individual areas based on their size. I partially did this by discarding some areas for which I found values but which I considered unimportant and/or extremely small (for example, front – under the nose).
- Point 4 is very arguable and would appreciate feedback here. What percentage of casualties were front/side/rear kills? One practical reason I discarded rear armor was simply: Outside the American tanks, it was almost impossible to find values.
- Totally excluded turret armor, but I checked that the turrets were roughly matching the bodies. Is that reallistic? How common was a "turret kill"?
- Totally excluded the influence of material type/quality (as I’m a material scientist – this really hurts!) nor the structural attachment of the armor to the rest of the tank and many other points mentioned by B.

Results on the tanks that sparked the discussion:
- The American mediums all got values which are somewhere between the old and new scenario databases and sound sensible at first glance. As mentioned above the M4 comes at 65 (old: 90 new: 45), the M3 at 51 (65/38), M26 at 99 (135/76).
- The T34/85 comes to 74 (90/50)
- For most Jap armor I got the exact same values as the databases (though this might be partially due to the abysmal data situation – see below). This includes 50 for the Type 1 medium.

More results:
- The only difference for the IJA was in the Type 98 and Type 2 light which both got 22 instead of 40 (in both DBs). The maximum armor thickness is given as 16 mm by several sources and though their angles are arguably the best of all IJA armor, I see no way to translate a raw 16 into an effective 40 as the databases do it. Also consider: The Type 98 light has almost the same value as the M4 Sherman in the new database, which sounds improbable to say the least.
- The commonwealth mediums are also between the two databases: Matilda II: 86 (120/70), Valentine III: 66 (100/60) though the super heavy Churchill VII comes to 122 (120/95)
- Most of the American lights (M3, M5 and M24) come to the same value of 34, this is in contrast to the databases where the M5 and M24 are noticeably better than the M3. Looking at the raw armor thickness and angles I can’t find an obvious reason for the databases claim.
- I got differing values to the DBs for the American TDs though these were hardly changed between the two DBs. I will recheck my raw data here.
- According to my research the LVT(A) 1 was unarmored (12 in the DBs) and the LVT(A) 4 had only optional armor (which decreased the carrying capacity). Using the optional armor I get 10 as armor value instead if the DBs’ 22.

Outstanding: I haven’t done any Soviet tank other than the T34/85 nor any of the armored cars.

Sources:
- American: Used the great site http://afvdb.50megs.com and http://www.wwiivehicles.com as the main sources. Info was plenty and easy to find.
- T34/85 as the T34 is a minor legend you find the values pretty easily
- Commonwealth: Used http://www.wwiivehicles.com and http://www.historyofwar.org as main sources. Info was ok, except for the Valentine III were info was scarce.
- IJA: I desperately scrounged any number from http://www.wwiivehicles.com http://www.historyofwar.org and Wikipedia. In most cases the only info I found was like “6-16 mm” without any details on distribution. Angles I guessed myself based on pictures. In the one case where I found detailed distributions (Type 97 medium) I realized that the Japanese designers must have used the same formula as myself to create a homogeneous effective armor front. After that I simply assumed they did the same on the ones where detailed area info was lacking. This ended in most tanks effectively getting their maximum armor thickness as armor value and except for the lights mentioned above those values are exactly the same as the ones already in the DB. Perhaps the original designers also just took this value?
- I found 2 vehicles which could be meant to be the “Type 4 SP-Gun”: The “Type 4 Ho-Ro” or the “Type 4 Ha-To” (it’s worth googling the later just to see a picture, looks like something out of Mad Max).
- I have no idea what the “Type 2 CS-Tank” is supposed to be. Found no "Type 2" vehicle which seemed to match. Any input would be appreciated.

Final words:
Any comments, criticism or ideas on the methodology would be highly appreciated. I’m also willing to share the collected raw data and all results with anybody interested.

Update after reading Panther Bait:
War Thunder also uses the same formula for effective armor thickness for sloped armor. In addition the game wiki lists many values I didn’t find (details for the IJA!). I could use it to fill some gaps if the approach is judged worthwhile.

That is a good method Rafid. It is much wiser than simply seeking the biggest number one can find. The code is very simplistic and if armor must be characterized by a single value, that value must necessarily represent an ‘average’ or ‘effective’ value.

Babes scenarios use a very similar technique, a box rule using a standard rectangle. A turret side is 1 rectangle, an upper hull is 2 rectangles side-by-side, a lower hull is 2 rectangles side-by-side, an upper front hull is 1 rectangle, a lower front hull is 1 rectangle, and a turret front is 2/3 of a rectangle. We usually say the rear hull is 1 rectangle and turret rear is 2/3 of a rectangle, but they can be ignored, as de minimis, as you suggest. So the ‘effective’ armor for a vehicle is:

Enter Turret side armor x 2 (Left and Right sides)
Add Upper hull side armor x 4 (2 rectangles for L and R each)
Add Lower hull side armor x 4 (2 rectangles for L and R each)
Add Upper front armor x 1
Add Lower front armor x1
Add Turret front armor x 0.6666

Divide by 12.6666

As one can see, it’s easy to add the rear armor (rear x 1 + turret rear x 0.6666) and divide the result by 14.3333. Most good sources give armor values for each of the listed areas. Your suggested sites are very reasonable. There is very detailed information available on Japanese AFVs. Unfortunately, for most, it is in Japanese. There is as much there for IJ tanks as for everybody else, so get a Japanese buddy (there are quite a few around) and a Japanese keyboard emulator and go at it. Or you could consult Akira Takazawa, aka Taki, who is very well versed in the genre and participates often in the better discussion forums. He is very responsive to reasoned requests and has much primary source material at his fingertips.

Matt

Thanks for the reply and info on the babes scen's calculation, Matt.

You're right that my data collection can only be considered a rough beginning if the task was to be truly attempted. Thanks for your suggestions of additional sources.

I would summarize, that there seems to be decent agreement how to calculate the influence of sloped armor and that some sort of averaging of the zones is required.
An outstanding point however would be the weighting of the zones. The babes’ approach seems to be solely focused on area size as weight. This suggests that a shot fired at a tank has an equal chance to hit it in any square inch of its surface area (excluding the rear). Since the sides have more surface area than the front, this favors the sides as targets.
Babes: 79% of shots are fired on the side of a tank and 21% are fired on the front.
If this was correct, the perfect tank design would have a much more evenly distributed armor than was actually the case. However we must keep in mind that tanks were designed for the mobile massive tank warfare of the ETO and used very differently in the PTO which maters to us here. For comparison: My averaging assumed a 50/50 % chance (but I would need to include the turret). I wonder if there are any suggestions or even better on the data on the subject: Which area was most likely hit (and we’re talking about any hit, not just fatal ones)?

ORIGINAL: US87891

Or you could consult Akira Takazawa, aka Taki, who is very well versed in the genre and participates often in the better discussion forums. He is very responsive to reasoned requests and has much primary source material at his fingertips.

Matt

+1

Sometimes its more than just the numbers themselves. All 50mm of steel are not the same and all shot are not the same. And taking into account design flaws and design bonuses. Relative numbers may have to be off in order to give real world percentage results.

Big tank = big target, small tank = small target.

The Type 95 Ha-Go could be disabled with a rifle bullet. Although highly maneuverable and able to go into rough places Japanese tanks were most often used as pillboxes.

https://www.youtube.com/watch?v=ao11xwciDyI


And the rare cave tank: https://www.youtube.com/watch?v=jyJKXDKgVPg

The gun was elevated and depressed by the gunners shoulder?!!

Failed Tanks! Episode 39: The Type 97 Chi-Ha


https://www.youtube.com/watch?v=MrFD9NlqO2k