Jump to content


Photo

Accuracy


  • Please log in to reply
7 replies to this topic

#1 ACBelMutie

ACBelMutie

    Crunchie

  • Members
  • Pip
  • 14 posts
  • Gender:Male

Posted 22 February 2017 - 0640 AM

Hello. Sorry about my english. I'm trying to make a comparison between soviet and german accuracy, but it's almost imposible to do it because of the different criteria. There are a few topics about accuracy, but I didn't find what I was looking for.

 

I know soviets use median dispersion and Germans use 50% zone dispersion. I have been reading this well-known page: http://www.panzer-war.com/page34.html

 

The autor says it's necesary to multiply by 2 the soviet 50% figures to be able to compare them with german's. But i'm not sure about that. For example, we can take one of the D-25T tests:

 

IS%2Baccuracy.jpg

 

The results are as follows:

 

is2%2Bprecision.PNG

 

We can see the 100% zone (a circle, not a square) and the 50% is very different in both tests. If we multiply by 2 the 50% radius to get the 50% zone, the results are very different. Let's compare that test with the IS-3 test:

 

is-3-gunnery-1.png

 

The result:

is3%2Bprecision.PNG

 

So my question is, is correct to multiply by 2 the R50% soviet dispersion get the german 50% zone?

 

I mean, you can get the 50% zone if you have the R100%, but not if you have R50% because R100% isn't double R50%.

 

Greetings

 

PD: Making some tests about accuracy, I got this two figures:

 

comparativaC_zpszchw7kxw.png

 

In the first figure, I got the real dispersion between the shots to get the maximum dispersion between them. In the second figure I used the maximum dispersion, multiplied it by 2 (Radius - Diameter) and got a very different result.

 

PD: las image modified (error)


Edited by ACBelMutie, 22 February 2017 - 1639 PM.


#2 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,139 posts
  • Gender:Male
  • Location:Southern California
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 22 February 2017 - 1154 AM

Those posts with the circles drawn in after a shoot are interesting but are don't tell us much about accuracy overall.

That may be of a particularly good shoot

 

This is taken from a 122mm firing table.

br471bmedian.jpg

For example Soviet tests of the Lend Lease 3" M7 gun consisted of 415 shots.

http://tankarchives....gun-trials.html

They are going to test their own guns far more than this.

 

The Soviets define their deviation zone as an oval not a circle.  It has an x and y axis, their numbers are a deviation from the center.

p35img1.jpg

 This is different than the German deviation 50% rectangle zone which also has an x and y axis.  

 

Also, note the UK defined their gun deviations with x and y axis. 

http://worldoftanks...._Hatch_Firefly/

 

So you can divide the German 50% zone by 2 to get a Russian deviation or multiply the Russian deviation by 2 to get a German deviation.

Once that is done the x,y deviation numbers shouldn't matter.  What does matter is the way an accuracy to-hit percentage is calculated from these deviations.  The German method converts to standard deviations in along each axis and compares that to target size in along each axis to find a ratio.  With some work the to-hit percentage is found after multiplying the ratios together.  That is the rectangular method.

 

The Russians use a different method to find the to-hit percent their oval represents.   There are Russian ballistics math sites on the internet and with Google translate you can probably figure out how they do it.  Unless you are Russian then no problem.

 

Now, post WWII the US released a number different studies on methods of finding the to-hit percentage (CEP) of an oval zone. 

The one I use is from Cornell University but there are others studies residing at the dtic site.  

http://www.dtic.mil/...t/u2/697697.pdf

 

This produces to-hit percentages that differs from the German rectangular method under about 1000 meters.  Beyond that there is little difference.  

The biggest anomaly in the rectangular method is that in many cases the rectangular method produces 100% accuracy under 300-500m (even with other errors like range estimation and gun jump thrown in),  With the CEP oval method it is only 100% at near 0 m range (well at least under 200 yds as that is where my calculations stop). 

 

Though a good thing about the rectangular method is that tanks are rectangular in shape.  There is more work trying to find how much of a rectangle is within an oval.


Edited by Mobius, 22 February 2017 - 1513 PM.


#3 KV7

KV7

    Crew

  • Members
  • PipPip
  • 862 posts

Posted 22 February 2017 - 1724 PM

Both are using the same methodology (taking the median) but for the USSR it is median from the centreline, for Germans it is median of the spread, for large samples they will converge to German measure = 2* Soviet measure.


Edited by KV7, 22 February 2017 - 1726 PM.


#4 ACBelMutie

ACBelMutie

    Crunchie

  • Members
  • Pip
  • 14 posts
  • Gender:Male

Posted 27 February 2017 - 1328 PM

Thanks for the answers.

 

So if we compare the accuracy of the D-25T and the KwK.43, the german gun is more acurated than the soviet one: http://www.panzer-war.com/page34.html

 

Is that correct? I'm using the 50% zones in both guns, so they are supposed to be the same figure and the D-25T has 0,6/0,6 al 1000m, when the KwK-43 has 0,3/0,5.

 

Greetings



#5 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,139 posts
  • Gender:Male
  • Location:Southern California
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 28 February 2017 - 0708 AM

Thanks for the answers.

 

So if we compare the accuracy of the D-25T and the KwK.43, the german gun is more acurated than the soviet one: http://www.panzer-war.com/page34.html

 

Is that correct? I'm using the 50% zones in both guns, so they are supposed to be the same figure and the D-25T has 0,6/0,6 al 1000m, when the KwK-43 has 0,3/0,5.

 

Greetings

As far as i can determine from firing tables, yes.   BTW, that is my website.  

 

Cheers.



#6 ACBelMutie

ACBelMutie

    Crunchie

  • Members
  • Pip
  • 14 posts
  • Gender:Male

Posted 04 March 2017 - 1623 PM

As far as i can determine from firing tables, yes.   BTW, that is my website.  

 

Cheers.

 

 

Nice web.

 

However, I have the same question as before, why multiply by 2 the soviet tables? I don't get why 2 is correct and 1.45 isn't right for example. I have a webpage too (in spanish) and I'd like to make a nice article about accuracy. I have some tables and a few firing tests, but I can't compare the data because of the different parameters used to get the numbers.

 

I'll translate the article about the 17 pounds accuracy in the future, because it's quite interesting.

 

Greetings



#7 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,139 posts
  • Gender:Male
  • Location:Southern California
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 04 March 2017 - 1802 PM

 

As far as i can determine from firing tables, yes.   BTW, that is my website.  

 

Cheers.

 

 

However, I have the same question as before, why multiply by 2 the soviet tables? I don't get why 2 is correct and 1.45 isn't right for example. I have a webpage too (in spanish) and I'd like to make a nice article about accuracy. I have some tables and a few firing tests, but I can't compare the data because of the different parameters used to get the numbers.

I multiply by 2 to get the Russian deviation into the same form as the German deviation.  It is easier to use in my program if the data is in the same form.

But it isn't used by the equations of the formula as 50% deviation it is used as standard deviations.  So it is converted to standard deviations by dividing by 1.349.

Here is the basis of the probability formula for an ellipse.

 

tohitformula.jpg



#8 Mobius

Mobius

    Member

  • Members
  • PipPip
  • 5,139 posts
  • Gender:Male
  • Location:Southern California
  • Interests:wargamming
    programming
    ballistics
    3D game development

Posted 11 March 2017 - 1116 AM

Here is an alternative formula.  It is from WWII Armor and Ballistics by Bird and Livingston.

There may be some errors in it.

ProbtohitWWIIballistics.jpg






0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users