Jump to content


Gun accuracy, is the wiki wrong ?


  • Please log in to reply
42 replies to this topic

OneSock #1 Posted 23 May 2020 - 06:39 PM

    Lieutenant

  • Player
  • 43634 battles
  • 1,827
  • [BBMM] BBMM
  • Member since:
    06-06-2011

Hi,

 

So accuracy in this game is sometimes a bit of a mystery, so I looked up what the wiki says and compared it to the actual game.

 

https://wiki.wargaming.net/en/Battle_Mechanics#Gun_Accuracy

 

Points of note:

 

"The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. "

 

The aiming circle (also called the dispersion circle, dispersion indicator or reticle) describes the area of 2 standard deviations (2σ;) from your aim point.

 

So consider STRV103, which has an accuracy of 0.3. As per the above theory, @356m the diameter of the aim circle should be (2x0.3)x3.56 = 2.14m

 

So look at the attached screen shot of the JT and ~356m. you can clearly see that the aim circle is bigger than the JT. IRL the JT is ~2.8m tall so clearly something doesn't add up (unless the JT is smaller in the game of course).

 

How would this make sense then? Well if the aim circle was actually 3 standard deviations and not 2 as stated then we get an aim circle of 2.14*1.5 = 3.2m, which roughly is what we see. 

 

I have seen some other guides where the aim circle has been stated as 3SD, so maybe the wiki is wrong ? And I have no idea if this lines up with the dispersion distribution !

 

 

Attached Files

  • Attached File   JT at 350m.png   415.98K

Edited by OneSock, 23 May 2020 - 06:39 PM.


General_Jack_D_Ripper #2 Posted 23 May 2020 - 06:51 PM

    Corporal

  • Player
  • 4955 battles
  • 163
  • Member since:
    08-21-2016

I think there was a time in which all guns of all tanks were more precise.

I think it was 1/3, so the wiki may have old text in it, I will look into it.

 

Cannot find it.


Edited by General_Jack_D_Ripper, 23 May 2020 - 07:18 PM.


qpranger #3 Posted 23 May 2020 - 07:02 PM

    Lieutenant General

  • Player
  • 38889 battles
  • 6,401
  • [HAMMY] HAMMY
  • Member since:
    12-25-2013

 



Kdingo #4 Posted 23 May 2020 - 07:07 PM

    Lieutenant General

  • Player
  • 32309 battles
  • 7,625
  • Member since:
    07-05-2011
They meddled with accuracy enough that you can safely ignore the values and just resort to yelling "russia" when you fire. Even kv2 is stronk sniper tank if you just yell loud enough.

unhappy__bunny #5 Posted 23 May 2020 - 07:08 PM

    Brigadier

  • Moderator
  • 21149 battles
  • 4,444
  • [-OC-] -OC-
  • Member since:
    08-01-2012

View PostOneSock, on 23 May 2020 - 05:39 PM, said:

Hi,

 

So accuracy in this game is sometimes a bit of a mystery, so I looked up what the wiki says and compared it to the actual game.

 

https://wiki.wargaming.net/en/Battle_Mechanics#Gun_Accuracy

 

Points of note:

 

"The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. "

 

The aiming circle (also called the dispersion circle, dispersion indicator or reticle) describes the area of 2 standard deviations (2σ;) from your aim point.

 

So consider STRV103, which has an accuracy of 0.3. As per the above theory, @356m the diameter of the aim circle should be (2x0.3)x3.56 = 2.14m

 

So look at the attached screen shot of the JT and ~356m. you can clearly see that the aim circle is bigger than the JT. IRL the JT is ~2.8m tall so clearly something doesn't add up (unless the JT is smaller in the game of course).

 

How would this make sense then? Well if the aim circle was actually 3 standard deviations and not 2 as stated then we get an aim circle of 2.14*1.5 = 3.2m, which roughly is what we see. 

 

I have seen some other guides where the aim circle has been stated as 3SD, so maybe the wiki is wrong ? And I have no idea if this lines up with the dispersion distribution !

 

 


The question should be, who creates the wiki and keeps it update?

 

Answer : The players.

 

Therefore the wiki is only as accurate as players allow it to be.

 

Quote from the "Beginners Guide" on the wiki.

 

Welcome, Editor!

The Wargaming Wiki is a web resource, where players can find useful information on all Wargaming titles. While the platform is provided by Wargaming, it's the community itself that shares their knowledge on the game and the content. So if you're interested in contributing to the wiki yourself, read on to learn the basics in just a couple of minutes!

 



Erwin_Von_Braun #6 Posted 23 May 2020 - 07:13 PM

    Lieutenant General

  • Player
  • 45749 battles
  • 7,190
  • Member since:
    01-25-2014

View PostKdingo, on 23 May 2020 - 06:07 PM, said:

They meddled with accuracy enough that you can safely ignore the values and just resort to yelling "russia" when you fire. Even kv2 is stronk sniper tank if you just yell loud enough.


As sad as it may seem, this is just as likely to yield a positive outcome as relying on the guns actual stats :(


Edited by Erwin_Von_Braun, 23 May 2020 - 07:17 PM.


arthurwellsley #7 Posted 23 May 2020 - 07:20 PM

    Brigadier

  • Player
  • 55914 battles
  • 4,338
  • [-MM] -MM
  • Member since:
    05-11-2011

This is not my work it comes from elsewhere, but I think it provides a better explanation of the same thing (just better graphs);

The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. In other words, for a gun with 0.32m dispersion at 100m, 95.45% of all shots will land within 0.32m of the center of your aim at that distance. Dispersion amount increases linearly with distance, i.e. 0.32m effective accuracy at 100m translates to 0.64m at 200m and 1.28m at 400m. 

 

Aiming Circle
Sigma Deviations
Standard_deviation_diagram.png (Picture sauce Thanks Mwtoews ) 

six_sigma_normal_distribution_2.jpg
WG uses a 2 Sigma system now (post 0.8.6) rather than the 3 sigma from before

The aiming circle (or reticle/thing that goes on the Red tanks...) describes the area of 2 standard deviations (2σ;) from your aim point. Based on a standard normal distribution, this used to mean that 4.2% of your shots would fall outside of the aiming circle. Since version 0.8.6 (the sigma rework) any shots falling outside the 2 Sigma circle get a second roll with a flat distribution (pancake) to place them randomly within the circle (prior to this they were placed on the perimeter). Because of this the shots move (deviate) away from your aim point, outwards toward the edge of the circle, but never shoot past its edge. 

(the above partially shameless stolen referenced from the wiki. Thanks Haswell :D )



24doom24 #8 Posted 23 May 2020 - 07:23 PM

    Lieutenant

  • Player
  • 9466 battles
  • 1,656
  • [WW3] WW3
  • Member since:
    10-20-2012

View Postarthurwellsley, on 23 May 2020 - 06:20 PM, said:

This is not my work it comes from elsewhere, but I think it provides a better explanation of the same thing (just better graphs);

The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. In other words, for a gun with 0.32m dispersion at 100m, 95.45% of all shots will land within 0.32m of the center of your aim at that distance. Dispersion amount increases linearly with distance, i.e. 0.32m effective accuracy at 100m translates to 0.64m at 200m and 1.28m at 400m. 

 

Aiming Circle
Sigma Deviations
Standard_deviation_diagram.png (Picture sauce Thanks Mwtoews ) 

six_sigma_normal_distribution_2.jpg
WG uses a 2 Sigma system now (post 0.8.6) rather than the 3 sigma from before

The aiming circle (or reticle/thing that goes on the Red tanks...) describes the area of 2 standard deviations (2σ;) from your aim point. Based on a standard normal distribution, this used to mean that 4.2% of your shots would fall outside of the aiming circle. Since version 0.8.6 (the sigma rework) any shots falling outside the 2 Sigma circle get a second roll with a flat distribution (pancake) to place them randomly within the circle (prior to this they were placed on the perimeter). Because of this the shots move (deviate) away from your aim point, outwards toward the edge of the circle, but never shoot past its edge. 

(the above partially shameless stolen referenced from the wiki. Thanks Haswell :D )

They also changed it so that less shots land in the direct middle of the aim reticle to counteract snap shots. 


Edited by 24doom24, 23 May 2020 - 07:23 PM.


Kdingo #9 Posted 23 May 2020 - 07:36 PM

    Lieutenant General

  • Player
  • 32309 battles
  • 7,625
  • Member since:
    07-05-2011

View PostErwin_Von_Braun, on 23 May 2020 - 07:13 PM, said:


As sad as it may seem, this is just as likely to yield a positive outcome as relying on the guns actual stats :(

 

But its more fun, specially when you have neighbors. :trollface:



Erwin_Von_Braun #10 Posted 23 May 2020 - 07:43 PM

    Lieutenant General

  • Player
  • 45749 battles
  • 7,190
  • Member since:
    01-25-2014

View PostKdingo, on 23 May 2020 - 06:36 PM, said:

 

But its more fun, specially when you have neighbors. :trollface:


They'll end up thinking you're a communist sympathizer...……….:hiding:



MeetriX #11 Posted 23 May 2020 - 08:26 PM

    Brigadier

  • Player
  • 26746 battles
  • 4,741
  • [_ACE] _ACE
  • Member since:
    08-12-2012
Accuracy (and many other things) is directly related to your unicumness.

shikaka9 #12 Posted 23 May 2020 - 10:25 PM

    Captain

  • Player
  • 78403 battles
  • 2,404
  • [-SJA-] -SJA-
  • Member since:
    02-27-2013
wot wiki is well known source for a fake information, never trust it!

Jauhesammutin #13 Posted 23 May 2020 - 10:54 PM

    Second Lieutenant

  • Player
  • 24287 battles
  • 1,208
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View PostOneSock, on 23 May 2020 - 05:39 PM, said:

Hi,

 

So accuracy in this game is sometimes a bit of a mystery, so I looked up what the wiki says and compared it to the actual game.

 

https://wiki.wargaming.net/en/Battle_Mechanics#Gun_Accuracy

 

Points of note:

 

"The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. "

 

The aiming circle (also called the dispersion circle, dispersion indicator or reticle) describes the area of 2 standard deviations (2σ;) from your aim point.

 

So consider STRV103, which has an accuracy of 0.3. As per the above theory, @356m the diameter of the aim circle should be (2x0.3)x3.56 = 2.14m

 

So look at the attached screen shot of the JT and ~356m. you can clearly see that the aim circle is bigger than the JT. IRL the JT is ~2.8m tall so clearly something doesn't add up (unless the JT is smaller in the game of course).

 

How would this make sense then? Well if the aim circle was actually 3 standard deviations and not 2 as stated then we get an aim circle of 2.14*1.5 = 3.2m, which roughly is what we see. 

 

I have seen some other guides where the aim circle has been stated as 3SD, so maybe the wiki is wrong ? And I have no idea if this lines up with the dispersion distribution !

 

 

The difference comes from perspective. Try to aim at a rock right in front of you and then at a mountain 1km away. The aiming reticle on your screen (when fully aimed) will be the same size. If the aiming reticle would adjust with the distance it would be quite disorienting, so instead it's fixed at some distance. 



ZlatanArKung #14 Posted 24 May 2020 - 10:26 AM

    Major General

  • Player
  • 1537 battles
  • 5,758
  • Member since:
    12-20-2014

View Postarthurwellsley, on 23 May 2020 - 07:20 PM, said:

This is not my work it comes from elsewhere, but I think it provides a better explanation of the same thing (just better graphs);

The accuracy value for a gun is given in meters at a range of 100m. The lower the value the more accurate your gun is. The value describes 2 standard deviations σ from the center of your aim. In other words, for a gun with 0.32m dispersion at 100m, 95.45% of all shots will land within 0.32m of the center of your aim at that distance. Dispersion amount increases linearly with distance, i.e. 0.32m effective accuracy at 100m translates to 0.64m at 200m and 1.28m at 400m. 

 

Aiming Circle
Sigma Deviations
Standard_deviation_diagram.png (Picture sauce Thanks Mwtoews ) 

six_sigma_normal_distribution_2.jpg
WG uses a 2 Sigma system now (post 0.8.6) rather than the 3 sigma from before

The aiming circle (or reticle/thing that goes on the Red tanks...) describes the area of 2 standard deviations (2σ;) from your aim point. Based on a standard normal distribution, this used to mean that 4.2% of your shots would fall outside of the aiming circle. Since version 0.8.6 (the sigma rework) any shots falling outside the 2 Sigma circle get a second roll with a flat distribution (pancake) to place them randomly within the circle (prior to this they were placed on the perimeter). Because of this the shots move (deviate) away from your aim point, outwards toward the edge of the circle, but never shoot past its edge. 

(the above partially shameless stolen referenced from the wiki. Thanks Haswell :D )

In some new update like 9.18 WG remade distribution in such a way that a shot landing at the edge have a 0 probability of happening.

Earlier,  you had many shots landing on the edge, since any shot calculated to land outside was moved to the edge.

But this have been changed, and now you get a distribution inside circle with a 0% chance of landing at Edge. 



arthurwellsley #15 Posted 24 May 2020 - 12:27 PM

    Brigadier

  • Player
  • 55914 battles
  • 4,338
  • [-MM] -MM
  • Member since:
    05-11-2011

View PostZlatanArKung, on 24 May 2020 - 09:26 AM, said:

In some new update like 9.18 WG remade distribution in such a way that a shot landing at the edge have a 0 probability of happening.

Earlier,  you had many shots landing on the edge, since any shot calculated to land outside was moved to the edge.

But this have been changed, and now you get a distribution inside circle with a 0% chance of landing at Edge. 

 

"In some new update like 9.18 WG remade distribution in such a way that a shot landing at the edge have a 0 probability of happening."  FALSE. Patch 9.18 did no such thing. Here are the patch notes = https://worldoftanks...18-updatenotes/

 

"Earlier,  you had many shots landing on the edge, since any shot calculated to land outside was moved to the edge."  Partly true. Patch 8.6 rework means shots that would have gone outside the circle get a second roll that places them randomly within the circle. Prior to patch 8.6 shots outsider the perimeter were allocated to the perimeter line.

 

"But this have been changed, and now you get a distribution inside circle with a 0% chance of landing at Edge. "  FALSE. Prior to 8.6 shots would be allocated to land on the perimeter. Post patch 8.6 as above there is the second roll to see where they randomly land within the perimeter. Because it is a random roll there is still a chance that the shot will land at the edge.



evilchaosmonkey #16 Posted 24 May 2020 - 01:09 PM

    Captain

  • Player
  • 18891 battles
  • 2,268
  • [EIGHT] EIGHT
  • Member since:
    05-04-2013

View Postarthurwellsley, on 23 May 2020 - 06:20 PM, said:

but never shoot past its edge. 

 

 

My account would like to beg to differ on this point.



ZlatanArKung #17 Posted 24 May 2020 - 09:24 PM

    Major General

  • Player
  • 1537 battles
  • 5,758
  • Member since:
    12-20-2014

View Postarthurwellsley, on 24 May 2020 - 12:27 PM, said:

 

"In some new update like 9.18 WG remade distribution in such a way that a shot landing at the edge have a 0 probability of happening."  FALSE. Patch 9.18 did no such thing. Here are the patch notes = https://worldoftanks...18-updatenotes/

 

"Earlier,  you had many shots landing on the edge, since any shot calculated to land outside was moved to the edge."  Partly true. Patch 8.6 rework means shots that would have gone outside the circle get a second roll that places them randomly within the circle. Prior to patch 8.6 shots outsider the perimeter were allocated to the perimeter line.

 

"But this have been changed, and now you get a distribution inside circle with a 0% chance of landing at Edge. "  FALSE. Prior to 8.6 shots would be allocated to land on the perimeter. Post patch 8.6 as above there is the second roll to see where they randomly land within the perimeter. Because it is a random roll there is still a chance that the shot will land at the edge.

No.

 

WG do not use a normal standard distribution that gets rerolled once when calculating where shots end up on circle..

You don't get the distribution of 9.15 or whatever patch is with rerolling the shot once.

Look at image to verification.

 

https://www.google.c...he-sansbox/amp/

 

 

So the idea of rerolling a shell outside reticule once is thus false.

With singular reroll you would still get shots at edge, something we don't see in picture.

You could have infinite rerolled until shot lands within reticule, but that is just a incredibly stupid way of dealing with this. Much easier to use an altered standard distribution that makes shots on edge and outside impossible.

 


Edited by ZlatanArKung, 24 May 2020 - 09:31 PM.


Tom_Deekanarry #18 Posted 25 May 2020 - 09:13 AM

    Warrant Officer

  • Player
  • 17256 battles
  • 603
  • Member since:
    06-28-2012

I have been left with the impression the aiming circle is designed to put the shot into any part that has no tank in it 50% of the time,  it's also programmed to home in on any obstacle no matter how small even if it only covers one road wheel or corner of the target!

 

 



Jauhesammutin #19 Posted 25 May 2020 - 10:03 AM

    Second Lieutenant

  • Player
  • 24287 battles
  • 1,208
  • [KANKI] KANKI
  • Member since:
    11-05-2013

View Postevilchaosmonkey, on 24 May 2020 - 12:09 PM, said:

 

My account would like to beg to differ on this point.

And again, the aiming reticle doesn't follow perspective. It's fixed somewhere between you and your target. It's not "painted on" your target like with arty.



evilchaosmonkey #20 Posted 25 May 2020 - 01:00 PM

    Captain

  • Player
  • 18891 battles
  • 2,268
  • [EIGHT] EIGHT
  • Member since:
    05-04-2013

View PostJauhesammutin, on 25 May 2020 - 09:03 AM, said:

And again, the aiming reticle doesn't follow perspective. It's fixed somewhere between you and your target. It's not "painted on" your target like with arty.

 

What on earth has that got to do with shots going outside the aiming circle?  Which does happen.  Still. Despite what patch notes say.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users