Accuracy of main battle tanks

[Redwolf]

Originally posted by Vanir Ausf B:

The comparative accuracy of guns in CM is simply a function of muzzle velocity.

In addition to issue below, people keep claiming BTS models improving accuracy for the 17pdr over the months, or only the APDS shot of the 17pdr. I have never seen BTS saying this, can anyone confirm or deny it?

CM does seem to model the fact that larger rounds are more accurate at long range than smaller rounds, even with lower initial velocity. For example, the 50mm gun experiences a greater decrease in accuracy from 100m to 2000m than the 88mm. So, the 88 should become more accurate relative to the 76 as range increases. However, the difference does not appear to become significant until ranges outside of those seen in a typical CM game. How realistic this is I don't know.

Very realistic - the heavier the projectile is, the less impact a given side wind has. I guess that is what BTS did in the patches that bumped long-range 88mm accuracy.

[KNac]

Well...

First of all, we must see if that was an individual case or is the rule. To test this, design a map (for better conditions it should be plain, no cover of any type, and without other parameters, like unit experience, that could modificate the output) and put both tanks in specific positions in the range ou want to check out (ie 250m) then put the facing each other. And run the game X times when > is X better accuracy on the figures and stadistic (ie X = 100). To make easier the job you can do with a pair of guys. Each time you run the game you must keep the the units static, and write in a table the number of shots that takes each time the Tiger to take out the Sherman. With this we can see which are the probabilities. I can make some nice tables and functions fast for that and turn it out to a Gauss bell if needed.

With all the data we can check if it´s enought accurate by historical data and/or simple logics. If the number of round that take the Tiger to destroy the Sherman is >2 the 90% of the times (with a regular crew) it should be strange (that 90% can change by 80% or 95 %depending on the historical data we have, in which case we should base it on technical data and specific reports and not in individual and personal quotes).

However we must abstract the human factor with some logic. In this case we have both regular crews, one in charge of a Tiger I and other in a Sherman. In a logical way we should say that the Sherman crew should be scared and should have quite a few fear of death making them more nervous and though less accurate (but however it have been demostrated that adrenalina can help in combat situations sometimes) in the other side we have the Tiger crew, with the moving pillbox and aware of the situtaion, but not more scared than the Sherman crew, making not them more nervous. As we were saying, those tanks were at point blank range, or at least at short range (>250 m). Well if you have played a good WWII tank simulator you should see how is the aiming system on those tanks, and personally I think that for a trained crew it shouldn´t be very hard to aim a 250m away tanks, even a small tank, I will use WWIIOL as an example (I have quit the game because is almost incomplete in a lot of things, but in the tank sim and aiming systems is very accurate) with the optics of 1940 PzIII F you can aim easy to an A13 or a R-35 that are for sure smaller than a Sherman at 250 m without problems, and saying that the Tiger I optics should be more similar to the Flak36 optics than the PzIII F, I don´t think that it should be a problem (I´ve destroyed Somuas S-35 at 2000m with the Flak 36).

Having said that the "scared" factor is not a modificator for the most of cases (say 90%-95%) with the consitions that we have written for the scenario, the graphic and data we could archive by testing sould say us if aiming at 250m for the Tiger I in the specified conditions (aka perfect conditions, that are neither 100& realistic in game or in real life, but can give us in some way a good conclussion) for more reliable info we should do the same with other more normal conditions, and then mix all the data we have gained to come with a conclussion.

Off course you can do this with other tanks.

Just my 0.2 euros

[Andreas]

Originally posted by Vanir Ausf B:

IIRC American forces did not encounter the Tiger in combat in Europe until the Ardennes offensive in December (at least in Northern Europe, they may have in Italy and they did in N Africa as well). I'm sure some grog will slap me around if I'm wrong.

[slap]They did encounter Tigers in Normandy, some of them (very few even Koenigstiger variant.[/slap] A number of sPzAbt were present in Normandy.

[Andreas]

Originally posted by Mattias:

Missed some of the finer points of that as I took it from memory...

Fleisher/Eiermann claims in their book "Die deutsche Panzerjägertruppe" that the 75mm L/70 mounted on the JgPz IV/70 had 80-90% chance of hitting a 2,5 x 2 meter target at 1000 meters under "kriegsmässigen Bedingungen", combat conditions.

That was the quote I was thinking of..

M.

Thanks, very interesting. Are 'kriegsmaessige Bedingungen (which I translate as 'combatlike conditions', but not 'combat conditions') for the IVL70 primarily ambush conditions by any chance? Seeing that it was a Panzerjaeger.

Edit - while a native speaker, I am not certain of the vernacular in this case. I don't proclaim to know what kriegsmaessige Bedingungen meant in an official report in 1944.

[ April 30, 2002, 02:04 PM: Message edited by: Andreas ]

[kappa]

I have had exactly the same experience as jack carr, generally german tanks are better, but in situations were there should be no doubt that the german tank makes the kill e.g. ambush; 200m etc., he never dose.

:confused:

[Vanir Ausf B]

Originally posted by redwolf:

Very realistic - the heavier the projectile is, the less impact a given side wind has. I guess that is what BTS did in the patches that bumped long-range 88mm accuracy.

I think it also has a lot to due with the total amount of initial energy.

I should have been more clear. I am well aware that the principle is realistic, but the differences between the 50mm and 88mm at 2000m are not as large as I would have expected in my totally unscientific opinion. So it was actually the amount to which it is modeled that I was questioning.

Originally posted by Germanb-, er, Andreas:

They did encounter Tigers in Normandy, some of them (very few even Koenigstiger variant. A number of sPzAbt were present in Normandy.

I knew they were there, but I could have sworn I had read somewhere that the Germans made the decision to commit them against the British.

Bah. I know nothing.

[ April 30, 2002, 02:32 PM: Message edited by: Vanir Ausf B ]

[Redwolf]

Originally posted by Vanir Ausf B:

[QB] </font><blockquote>quote:</font><hr />Originally posted by redwolf:

Very realistic - the heavier the projectile is, the less impact a given side wind has. I guess that is what BTS did in the patches that bumped long-range 88mm accuracy.

I think it also has a lot to due with the total amount of initial energy.

</font>

[Redwolf]

Originally posted by KNac:

First of all, we must see if that was an individual case or is the rule. To test this, design a map (for better conditions it should be plain, no cover of any type, and without other parameters, like unit experience, that could modificate the output) and put both tanks in specific positions in the range ou want to check out (ie 250m) then put the facing each other. And run the game X times when > is X better accuracy on the figures and stadistic (ie X = 100). To make easier the job you can do with a pair of guys. Each time you run the game you must keep the the units static, and write in a table the number of shots that takes each time the Tiger to take out the Sherman. With this we can see which are the probabilities. I can make some nice tables and functions fast for that and turn it out to a Gauss bell if needed.

Why all the tests? CMBO prints the hit probablity.

In addition, that is simply a wrong test because when target and shooter stand still, you get zeroing in, but the slightest TacAI move or target drop/retargetting resets it. So you don't even get the data you want unless you look very closely.

With all the data we can check if it´s enought accurate by historical data and/or simple logics. If the number of round that take the Tiger to destroy the Sherman is >2 the 90% of the times (with a regular crew) it should be strange (that 90% can change by 80% or 95 %depending on the historical data we have, in which case we should base it on technical data and specific reports and not in individual and personal quotes).

However we must abstract the human factor with some logic. In this case we have both regular crews, one in charge of a Tiger I and other in a Sherman. In a logical way we should say that the Sherman crew should be scared and should have quite a few fear of death making them more nervous and though less accurate (but however it have been demostrated that adrenalina can help in combat situations sometimes) in the other side we have the Tiger crew, with the moving pillbox and aware of the situtaion, but not more scared than the Sherman crew, making not them more nervous. As we were saying, those tanks were at point blank range, or at least at short range (>250 m). Well if you have played a good WWII tank simulator you should see how is the aiming system on those tanks, and personally I think that for a trained crew it shouldn´t be very hard to aim a 250m away tanks, even a small tank, I will use WWIIOL as an example (I have quit the game because is almost incomplete in a lot of things, but in the tank sim and

aiming systems is very accurate) with the optics of 1940 PzIII F you can aim easy to an A13 or a R-35 that are for sure smaller than a Sherman at 250 m without problems, and saying that the Tiger I optics should be more similar to the Flak36 optics than the PzIII F, I don´t think that it should be a problem (I´ve destroyed Somuas S-35 at 2000m with the Flak 36).

No, this hearsaying stuff is most probably nonsense. The last accuracy thread had a few range shooting numbers quoted, and they support BTS base hit probablities. I have the Panzerjaeger autobio I intend to write a review for and it supports it very well.

Isolated comments stating how impressive this and that was are worth almost nothing.

Reasons: most don't state the number of misses. Take for example the Elefant company example quoted in this very thread. OK, so they hit the Soviet tanks at distances never known before. Can you imagine how many shots 10 to 16 vehicles get off within a few minutes? What hit probablity does that end up with?

Without further qualification, I do not believe any test results like the IV/70 results posted here. What have been written here about that test could very well be shots from a stationary shooting position, not doing a new range estimation and gun targetting each time. Please post more info on the exact circumstances.

And I don't want the results of truely gifted guys to be taken into account, like Wittman or the Stuart guys who supposely made a sports out of kill Panther drivers by exploiting the mantlet shot trap. I want realistic gameplay. CMBO misses a few factors in the hit probablity model, I agree. But I don't want that covered up with a lameass fix by just bumping hit probablity to levels that are atypical to start from, blown out of proportion by bad reporting and omissions like number of misses, or outright fairy tales.

Thanks.

[Terence]

I have a question for Soddball. When you measure accuracy, do you think the size of the target is significant at all?

It seems to me that you have been testing Tigers vs Shermans and Shermans vs Tigers. Is it possible or worthwhile for you to test Tigers Vs. X and Shermans Vs. X, where X is a vehicle that exists in both OOB and is of similar size/profile.

If you have been doing that, pls ignore my post.

[Soddball]

Holien said:

Looking at those figures the total 1st shots (missed and hit) for each tank type do not match?

Why?

Surely if you are going to get an accurate comparison the total number of shots should match?

H

The reason is that in some cases, one tank fired and killed another before that other had a chance to fire its first round. It was invariably a Sherman which got its shot off first, killing the Tiger before the latter had fired its first shot.

Terence said:

I have a question for Soddball. When you measure accuracy, do you think the size of the target is significant at all?

It seems to me that you have been testing Tigers vs Shermans and Shermans vs Tigers. Is it possible or worthwhile for you to test Tigers Vs. X and Shermans Vs. X, where X is a vehicle that exists in both OOB and is of similar size/profile.

If you have been doing that, pls ignore my post.

I do think that the size of the target may factor in somehow, but I don't know how. I will be doing other tests (although not today, it's 8pm and I've just got home from a 12-hour day at work) and will let the interested parties know how those tests pan out.

My suspicion is that what I will see is a gradual increase in accuracy for the Sherman as range decreases, whilst the for Tiger accuracy will increase, then drop suddenly, as range decreases.

Redwolf said:

And I don't want the results of truely gifted guys to be taken into account, like Wittman or the Stuart guys who supposely made a sports out of kill Panther drivers by exploiting the mantlet shot trap. I want realistic gameplay. CMBO misses a few factors in the hit probablity model, I agree. But I don't want that covered up with a lameass fix by just bumping hit probablity to levels that are atypical to start from, blown out of proportion by bad reporting and omissions like number of misses, or outright fairy tales.

I couldn't agree more. As I'm sure you're aware, I haven't been bleating "BTS please fix or do somefink!" about CM, but I want to find out why it appears that a supposedly accurate weapon is inaccurate at short ranges.

[Holien]

Originally posted by Vanir Ausf B:

The comparative accuracy of guns in CM is simply a function of muzzle velocity. The gun with the higher MV will be more accurate all else being equal.

88L56 MV: 773 mps

76L54: 793 mps

The 76 has a higher MV and therefore is more accurate in CM. The fact that the Tiger is a bigger target than the Sherman adds to it.

This would explain the issue you have Sodball and I have learnt something new.

H

P.s. I see we posted at same time. Just seen your response, thxs for the info.

[ April 30, 2002, 03:20 PM: Message edited by: Holien ]

[Vanir Ausf B]

In order to factor out the much larger size of the Tiger, you need to fire the Sherman at a target about the same size as itself. So if you use a Wirblewind instead of the Tiger (silhouette value 105 compared to the Sherman 76's 104) you get the following first hit probablities for the 76L54 vs. the 88L56:

250m

88: 75%

76: 77%

500m

88: 54%

76: 57%

1000m

88: 29%

76: 29%

1500m

88: 15%

76: 14%

2000m

88: 8%

76: 6%

As you can see the differnces at short range are really quite small. I strongly believe these numbers show that the Sherman's greater effectiveness vs. the Tiger at short range is due mostly to higher rate of fire rather than greater accuracy.

[ April 30, 2002, 03:57 PM: Message edited by: Vanir Ausf B ]

[Redwolf]

Originally posted by Soddball:

I do think that the size of the target may factor in somehow,

It is already reflected in the hit probablity that the LOS or target tool prints.

A tank with silhuette 120 has a 20% (not percent point) higher chance to get hit than a silhuette 100 one, it is plain and simply linear.

[Terence]

Originally posted by Vanir Ausf B:

Vanir's excellent post

AHA!!!!

[Holien]

Mystery solved.

I await the release of CMBB to see my AT guns live longer.

Thxs very much for your time explaining and clearly showing the issues involved in simple statements of X misses Y too often.

H

[Vanir Ausf B]

Originally posted by redwolf:

A tank with silhuette 120 has a 20% (not percent point) higher chance to get hit than a silhuette 100 one, it is plain and simply linear.

Actually silhouette counts for less than that, and less than I thought as well. For example, at 500m the 76L54 has a 59% first hit chance against a Tiger (120 silhouette) and a 52% first hit chance against a Pz IVH (92 silhouette). This is only a 13% difference despite a 30% difference in silhouette values.

EDITED for bad math.

[ April 30, 2002, 04:18 PM: Message edited by: Vanir Ausf B ]

[Andrew H.]

I think the MV and silhoutte sizes explain things nicely. There are a couple of points to keep in mind, the first being that the Tiger I was really a huge tank - the modified T-34 (I think) in SPR looked almost exactly like a Tiger...but it was much smaller. The second point to keep in mind is that despite the relatively large number of 76mm Sherms made, they were not around at a lot of critical situations. For one, they were not around in the first month or so following D-Day, which is where a lot of the negative comments about Sherms come from. That is, the tankers are talking about 75mm Sherms, which have a significantly lower MV than the 76mm Sherms. Somewhat randomly, there was almost a complete absence of 76mm Sherms at the Battle of the Bulge. I'm not sure why; I've always sort of assumed it was because they weren't given to the Ind. Tank Batts. supporting infantry. These guys mostly had M-10s or towed AT guns (which they also called Tank Destroyers).

Edit - I don't know how CM uses the silhouette in affecting the hit probability, I think it's generally understood that a taller vehicle is easier to hit than a broader (but shorter) vehicle, even if both vehicles have the same area. That's because ranging errors (which cause a shot to go high or low) are much more common than whatever kind of error would make a shell go to one side or the other. Having a larger target is always better, of course, but in calculating hit probability, the vehicle's height should probably be given more effect. If this is not already modelled, it would probably have the effect of making a Sherm easier to hit, since it's a pretty tall tank.

[ April 30, 2002, 04:43 PM: Message edited by: Andrew Hedges ]

[Vanir Ausf B]

If you're looking for some way in which CM short changes the Tiger it can be found in the undermodeling of the shatter gap of 76mm ammo vs. the Tiger's armor (according to Rexford) and the probable over-availability of tungsten. Hopefully (probably) fixed in CMBB.

[wodasini88]

Ok I have got a few extreme examples of stuff that regularly happens to me in games. Panther 20m away from a firefly shoots and misses.....twice. That is not possible period. Luckly the conscript firefly did nothing and the 3rd shot was a hit. This type of thing happens a lot to me, first shot missed at 80meters and under the gunner would have to purposefully aim the gun at the ground or in the sky to miss.

[KNac]

Well. I was asking for good and reliable info for the test Redwolf, nos aislate cases, sorry if I cannot express myself better, but english is not my language. I think I posted somewhere "technical and historical data " and "not individual or personal info" or something like that.

Also I´m not trying to atack BTS or something evil like that, I´ve better things to do and I do not win nothing doing that. I love the game and I´m waiting for CMBB so you (refering to anyone) don´t need to use "fanboi" face with me, LOL.

Second I was suggesting the test for those who are conplaining, if you are not, you don´t need to do it. The test is not about gun accuracy, you must understand that it is for proving that the Tiger shooting at point blank ranges just doesn´t work (or it works). In the second part of my post I was talking about how to delimitate which sould be the percentage of first (or second as much) kill shots, and how much the rest, however I wasn´t giving any percentage, just examples.

Finally, I would like to say that those kind of tests seem stupid but that way is how you test games, and other things in RL, so maybe thy aren´t that stupid. You think testing things is fun?

However, returning to the topic: if it happens 2 times of 100, no problem; but if it happens 90 times of 100 it´s a problem (for that I was suggeting the test). So individual examples are not a worth...

[ April 30, 2002, 06:32 PM: Message edited by: KNac ]

[Doug Beman]

Ok I have got a few extreme examples of stuff that regularly happens to me in games. Panther 20m away from a firefly shoots and misses.....twice. That is not possible period.

Upon what do you base your emphatic statement? Have you read every battle account?

Well if you have played a good WWII tank simulator you should see how is the aiming system on those tanks, and personally I think that for a trained crew it shouldn´t be very hard to aim a 250m away tanks, even a small tank, I will use WWIIOL as an example (I have quit the game because is almost incomplete in a lot of things, but in the tank sim and aiming systems is very accurate) with the optics of 1940 PzIII F you can aim easy to an A13 or a R-35 that are for sure smaller than a Sherman at 250 m without problems, and saying that the Tiger I optics should be more similar to the Flak36 optics than the PzIII F, I don´t think that it should be a problem (I´ve destroyed Somuas S-35 at 2000m with the Flak 36).

Ummm...tell me again why we should use one computer game as a benchmark against which we judge another computer game? In particular, the gun in the Tiger was NOT identical to the FlaK36; George Forty's book states

This gun, known as the KwK36, should not be regarded as a development of the FlaK18 and 36 guns, but a parallel development with the 7.5cmKwK40...the only similarity to the FlaK36 lies in the ammunition and ballistics.

The gunner's sighting gear was a binocular telescope (probably TZF 12a, also mounted in Panther), not quite identical to the rangefinders used in FlaK gunnery.

DjB

[ April 30, 2002, 11:17 PM: Message edited by: Doug Beman ]

[illo]

In my last game my veteran panther got 2 stuarts and 1 sherman while it was reversing along road at full speed. It was funny to see panther back up fast, jumping up and down (road was anything else but level) getting off 3 shots and 3 KOs at ranges over 600m.

And yes, also seen few times my crack/veteran panther miss 2 or more shots in row from under 50m ranges.

For me it seems there is maybe even too much of variation.

Both incidences (which have happened several times)feel just plain wrong.

[ May 01, 2002, 01:57 AM: Message edited by: illo ]

[wodasini88]

exactly illo, I mean sometimes I feel like my gunners in the tank are saying "humm well I got 2 first shot hits while moving at full speed just a minute ago so I guess I will miss this point blank shot in the sake of probobility" that is what I mean by too linear.

Doug I base my empathic statement on my turret visibly turned toward the firefly and my gun also leveled at it. Then firing and seeing the round hit the ground on the other side firefly. Well lets see how did that get there? I dont recall ever hearing about a round coming out of a tank's barrel then magically going through the enemy tank without penetrating it...twice. Then again I've never hear of a panther going full speed in reverse over hilly terrain and getting 3KOs at 600m. See my point?

[JonS]

Originally posted by wodasini88:

... weird round behaviour ... Well lets see how did that get there? I dont recall ever hearing about a round coming out of a tank's barrel then magically going through the enemy tank without penetrating it...twice ...

Maybe that was the prototype for the round that got Kennedy?

[Vanir Ausf B]

Originally posted by illo:

For me it seems there is maybe even too much of variation.

Both incidences (which have happened several times)feel just plain wrong.

There is a tendency to underestimate the frequency of rare events. Rare events happen all the time, both in real life and the game. You should expect to see them often. There is nothing wrong with this as long as it isn't the same event over and over. Then there could be a problem, but even then testing would have to be done to be sure.

A few weeks ago someone posted a documented case of a German 88 AT gun missing a Soviet tank at 50m. In real life, not the game. **** really does happen and one of the best things about CM is the fact that it reflects this. However, there are a lot of people who don't like seeing their units perform outside of what they consider the norm, and get all worked up when it costs them a game.

That's not to suggest everything in CM is 100% accurate. It's not. It's just that when you say "I saw my Panther miss twice at 20m. That should be impossible!" it doesn't really mean much.

wodasini88:

Doug I base my empathic statement on my turret visibly turned toward the firefly and my gun also leveled at it. Then firing and seeing the round hit the ground on the other side firefly. Well lets see how did that get there? I dont recall ever hearing about a round coming out of a tank's barrel then magically going through the enemy tank without penetrating it...twice.

If a round misses, where the little shot graphic hits is semi-randomly determined. Under the right circumstances you can see shots go through buildings and hills. Call it a bug if you want. It doesn't mean anything.