Jump to content


Dispersion values are 3 times worse than claimed.

Proof

  • Please log in to reply
109 replies to this topic

Vela_ZG #1 Posted 11 September 2019 - 12:44 PM

    Warrant Officer

  • Player
  • 27717 battles
  • 735
  • [CBS87] CBS87
  • Member since:
    10-09-2013

I did some measurements and tests and came to a conclusion that the claimed dispersion values are 3 times greater than WG claims. I took the Tiger1 as a reference since it's dimensions are easily accessible on the net. In the attachments you'll find the measurements. I measured with Centurion AX that has 0.29 accuracy against a Japanese Tiger. The model is identical to German Tiger1. 

I parked 100m away from it and let the gun fully aim, then took the screenshot. As you can see Centurion's actual accuracy is around 0.90 give or take, at 100m.

It would be interesting to do some comparisons against different tanks with the same accuracy on paper and see what the difference will be, if any.

In any case, the numbers are wrong with a ratio of 3:1. 

So now we have a reference point, and thats Tiger1's MG port thats roughly 30cm in diameter. Who ever wishes to do the testing is welcome. Maybe we can find out some other things and possibly improve them.

 

Attached Files

  • Attached File   shot_006.jpg   1.56MB
  • Attached File   shot_007.jpg   1.1MB
  • Attached File   WOT dispersion.png   5.35MB


Homer_J #2 Posted 11 September 2019 - 12:57 PM

    Field Marshal

  • Beta Tester
  • 31981 battles
  • 35,125
  • [WJDE] WJDE
  • Member since:
    09-03-2010
Interesting because I've seen this done before with "correct" results.

Jauhesammutin #3 Posted 11 September 2019 - 01:15 PM

    Warrant Officer

  • Player
  • 23607 battles
  • 867
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View PostVela_ZG, on 11 September 2019 - 11:44 AM, said:

I did some measurements and tests and came to a conclusion that the claimed dispersion values are 3 times greater than WG claims. I took the Tiger1 as a reference since it's dimensions are easily accessible on the net. In the attachments you'll find the measurements. I measured with Centurion AX that has 0.29 accuracy against a Japanese Tiger. The model is identical to German Tiger1. 

I parked 100m away from it and let the gun fully aim, then took the screenshot. As you can see Centurion's actual accuracy is around 0.90 give or take, at 100m.

It would be interesting to do some comparisons against different tanks with the same accuracy on paper and see what the difference will be, if any.

In any case, the numbers are wrong with a ratio of 3:1. 

So now we have a reference point, and thats Tiger1's MG port thats roughly 30cm in diameter. Who ever wishes to do the testing is welcome. Maybe we can find out some other things and possibly improve them.

 

The dispersion value is the RADIUS and not the diameter.

 

0.29 accuracy means (according to WG) that the shots will fall 0.29m away from the center at 100m. So the diameter of that circle should be 0.58m.

Also the aim circle only shows 2 standard deviations which leaves ~4,2% of the actual circle unseen.

 

In other words the diameter of the aiming circle you are looking at should be 2*0.29m*0,9545=0,554m.



Vela_ZG #4 Posted 11 September 2019 - 01:16 PM

    Warrant Officer

  • Player
  • 27717 battles
  • 735
  • [CBS87] CBS87
  • Member since:
    10-09-2013
Seriously? A radius?? Hahahhah OMG. :D OK, my bad, stil, its much bigger than it should be.
12:21 Added after 4 minute
Maybe we could talk WG into making it the right size? I mean, 0.554m is almost half of 0.92m. That would be a huge improvement. Why is it wrong in the first place?

Edited by Vela_ZG, 11 September 2019 - 01:18 PM.


Jauhesammutin #5 Posted 11 September 2019 - 01:34 PM

    Warrant Officer

  • Player
  • 23607 battles
  • 867
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View PostVela_ZG, on 11 September 2019 - 12:16 PM, said:

Seriously? A radius?? Hahahhah OMG. :D OK, my bad, stil, its much bigger than it should be.
12:21 Added after 4 minute
Maybe we could talk WG into making it the right size? I mean, 0.554m is almost half of 0.92m. That would be a huge improvement. Why is it wrong in the first place?

 

It's a radius, if the Wiki is correct (https://wiki.wargaming.net/en/Battle_Mechanics#Accuracy_and_Dispersion)

 

 

This is something I've noticed since day 1.

Here is another older example: (https://youtu.be/oKZ95l005wE?t=63)

Klaus is shooting an E-100 from 340m with Leo 1. For simplification let's assume the distance is 400m and the accuracy is 0.3.

 

According to the Wiki the radius of the aiming circle should be 2*0,3*0,9545*4=2,29m.

Actually the aiming circle should be even smaller since the distance is only 340m and Leo 1's accuracy gets a buff from the crew/equipment/consumables.

In the video you can see the aiming circle being BIGGER than the E-100. In other words, E-100 is <2,29m tall.

AMX-13 is 2,35m tall.

 



Vela_ZG #6 Posted 11 September 2019 - 01:47 PM

    Warrant Officer

  • Player
  • 27717 battles
  • 735
  • [CBS87] CBS87
  • Member since:
    10-09-2013
Yes, that's why I did this also. The accuracy is completely wrong. 

Nethraniel #7 Posted 11 September 2019 - 01:52 PM

    Major

  • Player
  • 14521 battles
  • 2,565
  • [T-D-U] T-D-U
  • Member since:
    05-30-2012

Are the dimensions of tanks in WoT actually correct with regards to distances? Had the impression they are not to scale. So, this would mean, that real dimensions from technical tank drawings are not the same as ingame. 



Vela_ZG #8 Posted 11 September 2019 - 01:55 PM

    Warrant Officer

  • Player
  • 27717 battles
  • 735
  • [CBS87] CBS87
  • Member since:
    10-09-2013
This could be the case, but in any way, at least the numbers should be corrected. If not the accuracy.

mjs_89 #9 Posted 11 September 2019 - 01:55 PM

    Warrant Officer

  • Player
  • 8242 battles
  • 795
  • [95137] 95137
  • Member since:
    08-16-2017

Have you taken the FOV-settings into account? I'm in no way an expert, but as I understand it anything but the standard 95° probably messes something up here. Would be an interesting thing to try out imo.

 



Balc0ra #10 Posted 11 September 2019 - 02:00 PM

    Field Marshal

  • Player
  • 72037 battles
  • 20,304
  • [WALL] WALL
  • Member since:
    07-10-2012

View PostVela_ZG, on 11 September 2019 - 01:16 PM, said:

Seriously? A radius?? Hahahhah OMG. :D OK, my bad, stil, its much bigger than it should be.

 

Even then, do it a few times, and results may vary to hit the mark exactly. It's not a preset locked value pr 10 or 100 shots either if you will.

 

View PostNethraniel, on 11 September 2019 - 01:52 PM, said:

Are the dimensions of tanks in WoT actually correct with regards to distances? Had the impression they are not to scale. So, this would mean, that real dimensions from technical tank drawings are not the same as ingame. 

 

Some tanks have been laser scanned to get models that are 100% the same size. Others have not. So there results will vary regardless.



Jauhesammutin #11 Posted 11 September 2019 - 02:11 PM

    Warrant Officer

  • Player
  • 23607 battles
  • 867
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View PostNethraniel, on 11 September 2019 - 12:52 PM, said:

Are the dimensions of tanks in WoT actually correct with regards to distances? Had the impression they are not to scale. So, this would mean, that real dimensions from technical tank drawings are not the same as ingame. 

How would you test it? The game has buildings, but are those buildings in scale?

You could try it in a relatively flat map. Count how many times a tank X fits between the edges of the map. We know the dimensions of the maps (1000*1000m or 800*800). 

Or you could measure the distance with your gun, let's say 50m, and see if X amount of tanks fit between.

 

Nonetheless it doesn't matter which scale is wrong. It's either the aim circle or the tank scale.

13:13 Added after 2 minute

View PostBalc0ra, on 11 September 2019 - 01:00 PM, said:

 

Even then, do it a few times, and results may vary to hit the mark exactly. It's not a preset locked value pr 10 or 100 shots either if you will.

 

 

 

We aren't talking about the standard deviation of the shots. We are talking about the accuracy of the gun. How many times you hit the bulls-eye has nothing to do with the accuracy.

 

View PostBalc0ra, on 11 September 2019 - 01:00 PM, said:

 

 

Some tanks have been laser scanned to get models that are 100% the same size. Others have not. So there results will vary regardless.

 

It doesn't matter. If I scan a 20cm ruler and scale it down to 15cm in the game, it's not in the correct scale no matter how accurate my laser scanner was.

13:14 Added after 3 minute

View Postmjs_89, on 11 September 2019 - 12:55 PM, said:

Have you taken the FOV-settings into account? I'm in no way an expert, but as I understand it anything but the standard 95° probably messes something up here. Would be an interesting thing to try out imo.

 

This could be interesting. Does the FOV affect sniper view though?



Dava_117 #12 Posted 11 September 2019 - 02:23 PM

    Brigadier

  • Player
  • 22277 battles
  • 4,671
  • [T-D-U] T-D-U
  • Member since:
    12-17-2014

Interesting! Being the dispersion expressing the radious and not the diameter, there still is a roughly 50% error than the listed stats. Is something that would be nice if WG could dig in. Doubt they will, but considering it may also be a bug, I would try to post this also in the bug report section of the forum.

 

Also cheers for opening a thread with some actual documentation! Its quite rare!  :)


Edited by Dava_117, 11 September 2019 - 02:24 PM.


Nethraniel #13 Posted 11 September 2019 - 02:37 PM

    Major

  • Player
  • 14521 battles
  • 2,565
  • [T-D-U] T-D-U
  • Member since:
    05-30-2012

I wouldn't call it wrong. Just calculating it through gives you unexpected results. For the use ingame and as value for comparison the WoT accuracy stat is ok.

Sure, you could mathematically min/max use of equipment and consumables if you would know the exact numbers, but on the other hand this is not a simulator and you know that a gun with more accuracy is in fact more reliable in hitting smaller targets. 



Dava_117 #14 Posted 11 September 2019 - 02:41 PM

    Brigadier

  • Player
  • 22277 battles
  • 4,671
  • [T-D-U] T-D-U
  • Member since:
    12-17-2014

View PostNethraniel, on 11 September 2019 - 02:37 PM, said:

I wouldn't call it wrong. Just calculating it through gives you unexpected results. For the use ingame and as value for comparison the WoT accuracy stat is ok.

Sure, you could mathematically min/max use of equipment and consumables if you would know the exact numbers, but on the other hand this is not a simulator and you know that a gun with more accuracy is in fact more reliable in hitting smaller targets. 

 

Well, while generally this is not a very important problem of the game, as affect all guns the same, I would still like that the info showed by WG being correct.



If_I_Die_You_Die_Too #15 Posted 11 September 2019 - 02:50 PM

    Warrant Officer

  • Player
  • 10769 battles
  • 992
  • Member since:
    07-14-2016

WoT numbers are like government statistics

 

They can make them up to suit themselves, there is no proper oversight and they can never be independently confirmed

 

You should just shoot tonks


Edited by If_I_Die_You_Die_Too, 11 September 2019 - 02:50 PM.


Mannanan #16 Posted 11 September 2019 - 02:55 PM

    Major

  • Player
  • 28654 battles
  • 2,943
  • [KHIL] KHIL
  • Member since:
    07-28-2017
I would be interested if fully aimed circle is in relation to 1 map square of 100m. Not that easy to measure though. I would find flat corner on a 1km map then drive along one side of a map square and count how many times fully aimed circle fits the other side of map square, using minimap to roughly find out where 100m ends on a map. It there is 3 times difference I would be enough to find out it does not fit. If it fits then tank models are out of scale and have different dimensions in relation to map distances. FoV settings might skew it too, would set them to 95%.
13:58 Added after 3 minute
Ah, it would mean to do 100 steps, if it is radius, with 0.5 dispersion gun:( But based on minimap one could guess at 50th step if at middle of the square or it is way off.

Edited by Mannanan, 11 September 2019 - 03:03 PM.


Karasu_Hidesuke #17 Posted 11 September 2019 - 03:10 PM

    Brigadier

  • Player
  • 16893 battles
  • 4,425
  • [I-S-L] I-S-L
  • Member since:
    12-03-2013
Well, there definitely is something strange there, because even when playing with the same tank it varies a lot from battle to battle. I know this is my 'tin foil hat moment', but I could swear there is some random generator that determines your overall RNG values for each separate battle meaning sometimes you get good RNG, and sometimes you just get royally shafted.

Jauhesammutin #18 Posted 11 September 2019 - 03:11 PM

    Warrant Officer

  • Player
  • 23607 battles
  • 867
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View PostNethraniel, on 11 September 2019 - 01:37 PM, said:

I wouldn't call it wrong. Just calculating it through gives you unexpected results. For the use ingame and as value for comparison the WoT accuracy stat is ok.

Sure, you could mathematically min/max use of equipment and consumables if you would know the exact numbers, but on the other hand this is not a simulator and you know that a gun with more accuracy is in fact more reliable in hitting smaller targets. 

 

View PostDava_117, on 11 September 2019 - 01:41 PM, said:

 

Well, while generally this is not a very important problem of the game, as affect all guns the same, I would still like that the info showed by WG being correct.

 

The accuracy stat doesn't really matter as long as you can compare different guns. But why does WG go all the way trying to explain the accuracy stat in-game if it's just a bunch of nonsense.



Dorander #19 Posted 11 September 2019 - 03:13 PM

    Major General

  • Player
  • 21025 battles
  • 5,741
  • Member since:
    05-07-2012

You might be doing double work here though you can certainly run your own experiments but you may find this informative too: http://forum.worldof...k-maths-inside/


Edited by Dorander, 11 September 2019 - 03:14 PM.


TheDrownedApe #20 Posted 11 September 2019 - 03:28 PM

    Lieutenant General

  • Player
  • 49068 battles
  • 6,285
  • [IDEAL] IDEAL
  • Member since:
    03-27-2013

View PostKarasu_Hidesuke, on 11 September 2019 - 02:10 PM, said:

Well, there definitely is something strange there, because even when playing with the same tank it varies a lot from battle to battle. I know this is my 'tin foil hat moment', but I could swear there is some random generator that determines your overall RNG values for each separate battle meaning sometimes you get good RNG, and sometimes you just get royally shafted.

 

I witness this all the time. You take out your...PTA and fire 10 shots in one game, 2 hit the centre and the other 8 go wild. Next game your same 10 shots all find their mark.


 

I'm not saying it's "evening out" your RNG over so many games but it is far, far from consistent and this is what bugs me.


 


 







Also tagged with Proof

25 user(s) are reading this topic

1 members, 24 guests, 0 anonymous users


    Space_Vato