Battleship Sigma Values.

[1nv4d3rZ1m]

Blatantly stolen from: https://www.reddit.com/r/WorldOfWarships/comments/4jw6i5/bb_sigma_values_0551_live/

 

This caught my eye because it seems to be the first time that a lot of sigma values have been shown. Interesting values to note are the 2.0 for the Nikolai, 2.1 for the Yamato, and that the Colorado changes sigma values with the hull. 

 

The sigma represents the vertical distribution of shells in the ellipse according to sources: https://www.reddit.com/r/WorldOfWarships/comments/4jspal/0551_live_056_pt_changes/d398528

 

Larger values are more accurate in game, however since higher sigma values mathematically represent populations less concentrated on the center the sigma given has to be something else.

 

Some Maths:

 

 

As you can see the lower sigma would be considered more accurate in game because you have more of the measured population near the center. The theory I think is best currently say that the sigma represents the sigma value the dispersion is attached to. 

 

 

So a 2 sigma would contain more of the shots than a 1 sigma and so would be more accurate. If someone has a better theory I am all ears...

 

 

 

Tier	Ship	Sigma
2	Mikasa	1.8
3	Kawachi	1.8
4	Myogi	1.8
4	Ishizuchi	2.0
5	Kongo	1.8
6	Fuso	1.5
7	Nagato	2.0
8	Amagi	1.8
9	Izumo	1.8
10	Yamato	2.1
3	South Carolina	1.8
4	Wyoming	1.5
4	Arkansas Beta	1.5
5	New York	1.8
5	Texas	1.8
6	New Mexico	1.5
7	Colorado A	1.8
7	Colorado B/C	2.0
8	North Carolina	1.8
9	Iowa	1.8
10	Montana	1.8
8	Tirpitz	1.8
6	Warspite	2.0
4	Imperator Nikolai I	2.0

[10T0nHammer]

Higher the number, the better the accuracy I assume.

 

Some how, a WW1 Relic (Nikolai) is more accurate than radar carrying USN BBs. Seems legit. I'm surprised at the Fuso though, I do more damage with my Fuso than my Nagato..

[scruffycavetroll]

what are those #'s?

[Terrible_Turtle]

what are those #'s?

 

[Zampy]

These numbers seem to be the opposite of the dispersion value used in WoT.  Here, they are greater than 1, in WoT, they are 1/sigma. Glad to finally see them, though.

[Madwolf05]

I don't understand how the New Mexico is on par with WWI battleships.

 

I don't understand how the NC, Iowa, and Montana are on par with the New York, yet worse than the Colorado.

 

How is the Yamato so much better than all of them?

[1nv4d3rZ1m]

Higher the number, the better the accuracy I assume.

 

Some how, a WW1 Relic (Nikolai) is more accurate than radar carrying USN BBs. Seems legit. I'm surprised at the Fuso though, I do more damage with my Fuso than my Nagato..

 

More shells, since its all a probability of shells going toward center, having more shells gives you more chances for a good rng roll and makes up for having a worse sigma. 

[10T0nHammer]

 

More shells, since its all a probability of shells going toward center, having more shells gives you more chances for a good rng roll and makes up for having a worse sigma. 

 

Makes sense, quantity has a quality of its own

[1nv4d3rZ1m]

I don't understand how the New Mexico is on par with WWI battleships.

 

NM is a ww1 BB.

 

 
I don't understand how the NC, Iowa, and Montana are on par with the New York, yet worse than the Colorado.

How is the Yamato so much better than all of them?

 

Because balance comrade. 

 

[Tedster_]

I don't understand how the New Mexico is on par with WWI battleships.

 

I don't understand how the NC, Iowa, and Montana are on par with the New York, yet worse than the Colorado.

 

How is the Yamato so much better than all of them?

 

because that's how the US built guns?  IJN went for extremely tight dispersion, but if the FCS is off, you miss completely.  A wider spread increases chances of a hit when the FCS is not perfectly accurate.  Works IRL, but in game where we have perfect FCS's, is a disadvantage.

 

@OP, my working theory is that the sigma value defines what point on the bell curve the listed dispersion is at, as that's how WoT does it (the listed value used to be for 1.3 sigma, then it was changed in 0.8.6 to 2 sigma)

[TenguBlade]

All I care about these numbers is that it finally shuts people up who say that the New Mexico is so much more inaccurate than the Fuso.

 

As Tedster pointed out a greater dispersion/sigma value makes dealing damage easier if you can't line up your target to a tenth of a kilometer, but if you're good at it then there's an inherent disadvantage.

[1nv4d3rZ1m]

 

@OP, my working theory is that the sigma value defines what point on the bell curve the listed dispersion is at, as that's how WoT does it (the listed value used to be for 1.3 sigma, then it was changed in 0.8.6 to 2 sigma)

 

Yeah sounds like what I said in my spoiler. 

[Seaman_Rookie]

Well, that explains why I like the Wyoming and the Fuso, but am having trouble stepping up the Colorado and Nagato...  It's more fun to hit stuff, than see your shells land in front of, and behind, your target...

 

[DoomStomper]

This is exactly why only listing a ship's maximum dispersion value isn't enough to tell the whole story.

 

Wargaming needs to find a way to incorporate this factor into the port stats so that we can more effectively compare overall accuracy.

[teamoldmill]

All I care about these numbers is that it finally shuts people up who say that the New Mexico is so much more inaccurate than the Fuso.

 

As Tedster pointed out a greater dispersion/sigma value makes dealing damage easier if you can't line up your target to a tenth of a kilometer, but if you're good at it then there's an inherent disadvantage.

 

I never understood why people think the Fuso is a shotgun. Just as accurate a any other BB in the middle tiers.

[1nv4d3rZ1m]

This is exactly why only listing a ship's maximum dispersion value isn't enough to tell the whole story.

 

Wargaming needs to find a way to incorporate this factor into the port stats so that we can more effectively compare overall accuracy.

 

What they need to do for both shell velocity and accuracy is have a visual GUI element with a slider bar for range that shows flight time, penetration, and the distribution of shells for the range. The accuracy also needs to be a plot with colors representing the probability of shells hitting that area and units so people can see how likely they are to get a hit on a ship at a range. 

[kerensky914]

All I care about these numbers is that it finally shuts people up who say that the New Mexico is so much more inaccurate than the Fuso.

Part of that perception is range.  If two ships have the same dispersion at "max range", and one has a max of 20km, while the other has a max of 15km, and they are both shooting at a target 10km away, the first ship will seem 'more accurate' because its shells are only at half their dispersion, while the second ship will be at 3/4 of its max dispersion.

[10T0nHammer]

 

What they need to do for both shell velocity and accuracy is have a visual GUI element with a slider bar for range that shows flight time, penetration, and the distribution of shells for the range. The accuracy also needs to be a plot with colors representing the probability of shells hitting that area and units so people can see how likely they are to get a hit on a ship at a range. 

 

It would also be nice if they added pen chances and damage probabilities of hitting with certain shells on certain parts of the ship along with that slider

[YamatoA150]

I wonder how much of an effect the Accuracy Mod affects the values.

 

Either way, this explains why my Fuso can bracket the target but fail to hit it perfectly with a solid lead but Nagato can not only strike the target, but often exactly where you want it.

[vak_]

Hmm. So it seems that each gun has a certain dispersion, and then they can modify total dispersion for each particular ship, so that ships can have different accuracy despite having same guns? Okay.

 

However, I still don't understand what these numbers represent, given that increasing values increases accuracy.

 

1nv4d3rZ1m, on 18 May 2016 - 04:08 PM, said:

Spoiler

 

 

So a 2 sigma would contain more of the shots than a 1 sigma and so would be more accurate. If someone has a better theory I am all ears...

 

 

I don't have a better hypothesis, but your suggestion doesn't make sense to me.  2 sigma does have more shots than one sigma, however without truncating at boundaries accuracy is the same, since it's the same distribution; with truncation one sigma is more accurate since 32% of shots will land at the boundary of +1 sigma

[VTAdmiral]

What I find interesting about this is that the sigma values for New York/Kongo, Fuso/New Mexico, Nagato/Colorado B, Amagi/North Carolina, and Iowa/Izumo all match and yet players claim USN battleships are far less accurate.

[DoomStomper]

What they need to do for both shell velocity and accuracy is have a visual GUI element with a slider bar for range that shows flight time, penetration, and the distribution of shells for the range. The accuracy also needs to be a plot with colors representing the probability of shells hitting that area and units so people can see how likely they are to get a hit on a ship at a range. 

 

Exactly.  I've suggested something like that numerous times in the past.

 

It shouldn't be that difficult to incorporate, but there will always be opposition from the shortsighted who don't care about understanding the intricacies of the game mechanics and write it off as useless information.  These are often the types who assume that any improvements to the UI would be done by the same programmers who handle balance, level design, and 3D modeling, as if it would result in a delay of new content.

[teamoldmill]

What I find interesting about this is that the sigma values for New York/Kongo, Fuso/New Mexico, Nagato/Colorado B, Amagi/North Carolina, and Iowa/Izumo all match and yet players claim USN battleships are far less accurate.

 

Range.

[Tedster_]

Hmm. So it seems that each gun has a certain dispersion, and then they can modify total dispersion for each particular ship, so that ships can have different accuracy despite having same guns? Okay.

 

However, I still don't understand what these numbers represent, given that increasing values increases accuracy.

 

 

I don't have a better hypothesis, but your suggestion doesn't make sense to me.  2 sigma does have more shots than one sigma, however without truncating at boundaries accuracy is the same, since it's the same distribution; with truncation one sigma is more accurate since 32% of shots will land at the boundary of +1 sigma

 

not sure if you played WoT, but this happened.  20% of your shots would hit edge of reticule pre 0.8.6

[vak_]

Some how, a WW1 Relic (Nikolai) is more accurate than radar carrying USN BBs

 

Why is everyone assuming that the number in the table represent the final accuracy that we see in game? To me that's ridicolous (and I have data-mined stats to back myself up).

 

I can repeat, I think this is just some dispersion modifier attached to a particular ship (and, perhaps, hull type). It allows to modify dispersion for a particular battleship without touching the gun itself (which would modify dispersion for all other ship with that rifle).

 

All I care about these numbers is that it finally shuts people up who say that the New Mexico is so much more inaccurate than the Fuso.

 

It IS more inaccurate than Fuso, at least at comparable range (10km).