These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

Warfare & Tactics

 
  • Topic is locked indefinitely.
12Next page
 

Optimal,falloff for artillery
Author
Previous Topic Next Topic
Bethesda Vortarhiat
Godlike Wannabe
#1 - 2011-09-08 08:31:34 UTC  |  Edited by: Bethesda Vortarhiat
Hi,i m having this dispute with a friend about optimal range for turrets.He says that optimal range is calculated in defferent ways for different turrets,something like....optimal for autocannons is different from optimal for artillery...i was like WTF....turrets are turrets...indeed theres differences made by ammo...tracking and other factors....but the formula for calculating the optimal range for a turret is same for all turrets,at least thats how i see it.
His argument comes somehow akward.if he fits autocannons on a maelstrom...he can hit almost anything at 0 distance-oviously with some misses now and there.But if he fits artillery...he cant even hit a battleship if it is in 0 range.I told him that optimal range represents the range where the gun has 100% chances to hit.So if optimal is 10km...you have 100% chances to hit the target if u shoot between 0 and 10km.But then he says he cannot hit another BS if it is in 0 from him..so does optimal range start from 1km for some turrets? or 2km ?

Thanks very much

P.S. Also he says that if optimal range for a turret is 15km for example....he has the most damage when he is at 15km from the target...if he is at 10km he has less damage.That i understand....but is it because he shoots under the optimal limit?or theres other factors involved?
Ypegae Cilamey
Lovely Hitchhikers
#2 - 2011-09-08 08:49:43 UTC
Hi!

The range to target is only one of the things that affect the damage of turrets. There is also tracking and signature.

http://wiki.eveonline.com/en/wiki/Turrets

and

http://wiki.eveonline.com/en/wiki/Turret_damage

Ypegae
Othran
Route One
#3 - 2011-09-08 08:52:40 UTC  |  Edited by: Othran
Optimal and falloff vary between even the same type of turret - eg optimal+falloff on a 200mm AC is further than it is on a 150mm AC.

Autocannon = short optimal, long falloff, good tracking
Artillery = long optimal, short falloff, poor tracking.

He's wrong about the damage and you're wrong about the chance to hit.

Within optimal your chance to hit will be 100% PROVIDED the target and the attacker are completely stationary.

Within falloff your chance to hit drops off with range such that at optimal + falloff your chance to hit is 50%. Once again this assumes both ships are stationary.

Once the ships begin to move then tracking comes into play. Autocannons have good tracking, artillery doesn't. That's why he can't hit another BS inside 10km using artillery but can with autocannons.

HTH.
Bethesda Vortarhiat
Godlike Wannabe
#4 - 2011-09-08 09:03:18 UTC


Within optimal your chance to hit will be 100% PROVIDED the target and the attacker are completely stationary.





Exactly my friends frustration...why does his BS arty misses another BS in 0 to him...and i point that he and the BS target IS stationary ...therefore theres no tracking involved.....is that possible?if it is possible ....why is it?
Othran
Route One
#5 - 2011-09-08 10:32:59 UTC  |  Edited by: Othran
Bethesda Vortarhiat wrote:


Within optimal your chance to hit will be 100% PROVIDED the target and the attacker are completely stationary.





Exactly my friends frustration...why does his BS arty misses another BS in 0 to him...and i point that he and the BS target IS stationary ...therefore theres no tracking involved.....is that possible?if it is possible ....why is it?






The formula doesn't work when distance and transversal are exactly zero.

The formula is :

ChanceToHit = 0.5 ^ ((((Transversal speed/(Range to target * Turret Tracking))*(Turret Signature Resolution / Target Signature Radius))^2) + ((max(0, Range To Target - Turret Optimal Range))/Turret Falloff)^2)


so if Range to target = 0 and transversal speed = 0 then you get :

ChanceToHit = 0.5 ^ ((((0/0))*(Turret Signature Resolution / Target Signature Radius))^2) + ((max(0, Range To Target - Turret Optimal Range))/Turret Falloff)^2)

0 divided by 0 will result in "not a number" if python follows the IEE rules. "Not a number" is simply a mechanism to show an invalid arithmetic operation has occurred. The upshot of which is that everything within the brackets above cannot be calculated.

In theory anything divided by 0 should result in "not a number" but that may have changed in the later revisions of the relevant IEE spec.
Bethesda Vortarhiat
Godlike Wannabe
#6 - 2011-09-08 10:49:27 UTC  |  Edited by: Bethesda Vortarhiat
Perfect,thank you very much...this i understand.But this means he shouldnt hit anything with ACs also...but he does hit in 0 with AC's...thats where the frustration is..

so....you have 2 BS in 0m from eachother.one is the target.both are stationary so no tracking is involved.

If you fit AC's to one BS and shoot the other one in 0m...u hit almost all the time...but if you fit arty..you wil miss all the time..so if the formula is the same for both ac's and arty...why does arty miss?




Also another thing....my friend also claims that he has the most damage when he is within the optimal limit....EX. if optimal is 15km....he says he has the most damage when he is exactly at 15km....if he is at 10km he says he has less damage....again...is that a fact ? from what i understand ...if you re within optimal you have 100% to hit with full damage provided that the target is stationary....so would u have less damage shootin from 10km than shootin from 15km ?
Othran
Route One
#7 - 2011-09-08 10:56:27 UTC
Bethesda Vortarhiat wrote:
Perfect,thank you very much...this i understand.Also another thing....my friend also claims that he has the most damage when he is within the optimal limit....EX. if optimal is 15km....he says he has the most damage when he is exactly at 15km....if he is at 10km he says he has less damage....again...is that a fact ? from what i understand ...if you re within optimal you have 100% to hit with full damage provided that the target is stationary....so would u have less damage shootin from 10km than shootin from 15km ?


It shouldn't make any difference provided both ships are stationary.

If they're moving then you probably will do more damage at the edge of falloff rather than at half the falloff range but thats to do with tracking - further away you are then the lower the angular velocity between the ships will be and hence your guns will hit better.
Bethesda Vortarhiat
Godlike Wannabe
#8 - 2011-09-08 11:09:58 UTC
Othran wrote:
Bethesda Vortarhiat wrote:
Perfect,thank you very much...this i understand.Also another thing....my friend also claims that he has the most damage when he is within the optimal limit....EX. if optimal is 15km....he says he has the most damage when he is exactly at 15km....if he is at 10km he says he has less damage....again...is that a fact ? from what i understand ...if you re within optimal you have 100% to hit with full damage provided that the target is stationary....so would u have less damage shootin from 10km than shootin from 15km ?


It shouldn't make any difference provided both ships are stationary.

If they're moving then you probably will do more damage at the edge of falloff rather than at half the falloff range but thats to do with tracking - further away you are then the lower the angular velocity between the ships will be and hence your guns will hit better.



Perfect,thank you very much...this i understand.But this means he shouldnt hit anything with ACs also...but he does hit in 0 with AC's...thats where the frustration is..

so....you have 2 BS in 0m from eachother.one is the target.both are stationary so no tracking is involved.

If you fit AC's to one BS and shoot the other one in 0m...u hit almost all the time...but if you fit arty..you wil miss all the time..so if the formula is the same for both ac's and arty...why does arty miss?
Othran
Route One
#9 - 2011-09-08 11:26:12 UTC  |  Edited by: Othran
No idea - perhaps there's a conditional check being made along the lines of if its a short range turret (AC/blaster) then you get 100% hits, if its long-range then you get 0%. No way to tell for sure unless you can see the code.

Its entirely possible that its a bug and they both should hit 100% (or 0%) - depends how CCP handle the "not a number" scenario. Having looked it appears that 0/0 is the only case which should return "not a number", 1/0 should return positive infinity, -1/0 should return negative infinity.

tl;dr need someone at CCP to look at the raw code to find out for sure.

Edit - they shouldn't be behaving differently. If range is zero then Chance to Hit should = +infinity unless transversal is also 0 in which case its entirely dependent on how CCP handle NaN.
KFenn
State War Academy
Caldari State
#10 - 2011-09-08 14:18:29 UTC  |  Edited by: KFenn
Othran wrote:
Autocannon = short optimal, long falloff, good tracking
Artillery = long optimal, short falloff, poor tracking.


Artillery has comparable falloff - it just has higher volley but less overall DPS than Autos. Everything else is true though.

EDIT: Artillery also has far higher fitting requirements.

Commanding Officer of the Treacle Tart Brigade
Othran
Route One
#11 - 2011-09-08 14:24:01 UTC
KFenn wrote:
Othran wrote:
Autocannon = short optimal, long falloff, good tracking
Artillery = long optimal, short falloff, poor tracking.


Artillery has comparable falloff - it just has higher volley but less overall DPS than Autos. Everything else is true though.



It doesn't.

Lets take a 200mm AC2 and a 280mm Howitzer2 (both top of the "small" class) :

200mm AC2 has 1200 optimal and 4800 falloff. The falloff is four times the optimal;

280 Howitzer2 has 12000 optimal and 8750 falloff. The falloff is 0.72 times the optimal.

In short you're talking nonsense so I'd stop now were I you Blink
KFenn
State War Academy
Caldari State
#12 - 2011-09-08 14:28:14 UTC
Othran wrote:
It doesn't.

Lets take a 200mm AC2 and a 280mm Howitzer2 (both top of the "small" class) :

200mm AC2 has 1200 optimal and 4800 falloff. The falloff is four times the optimal;

280 Howitzer2 has 12000 optimal and 8750 falloff. The falloff is 0.72 times the optimal.

In short you're talking nonsense so I'd stop now were I you Blink

Wait, what? Now you're talking optimal relative to falloff, instead of just Arty falloff vs. Auto falloff. 8750 > 4800, so in that regard it does have comparable (well, far superior) falloff.

I believe it is you, good sir, who is talking nonsense.

Commanding Officer of the Treacle Tart Brigade
Alua Oresson
Aegis Ascending
Solyaris Chtonium
#13 - 2011-09-08 14:54:36 UTC
Othran wrote:
Bethesda Vortarhiat wrote:


Within optimal your chance to hit will be 100% PROVIDED the target and the attacker are completely stationary.





Exactly my friends frustration...why does his BS arty misses another BS in 0 to him...and i point that he and the BS target IS stationary ...therefore theres no tracking involved.....is that possible?if it is possible ....why is it?






The formula doesn't work when distance and transversal are exactly zero.

The formula is :

ChanceToHit = 0.5 ^ ((((Transversal speed/(Range to target * Turret Tracking))*(Turret Signature Resolution / Target Signature Radius))^2) + ((max(0, Range To Target - Turret Optimal Range))/Turret Falloff)^2)


so if Range to target = 0 and transversal speed = 0 then you get :

ChanceToHit = 0.5 ^ ((((0/0))*(Turret Signature Resolution / Target Signature Radius))^2) + ((max(0, Range To Target - Turret Optimal Range))/Turret Falloff)^2)

0 divided by 0 will result in "not a number" if python follows the IEE rules. "Not a number" is simply a mechanism to show an invalid arithmetic operation has occurred. The upshot of which is that everything within the brackets above cannot be calculated.

In theory anything divided by 0 should result in "not a number" but that may have changed in the later revisions of the relevant IEE spec.




The problem with that is that a calculator sometimes doesn't understand the realities of math. The client does take into account when people are stationary what the damage is. 0/0 = 0 and anything that is multiplied by zero can be discarded. So, the calculation you would get would be:

ChanceToHit = ((max(0, Range To Target - Turret Optimal Range))/Turret Falloff)^2)

Which, if your target is for 0 to optimal your chance to hit will be 100%. If you are past optimal you will do damage while in falloff. When you reach the falloff range you will do 40% of the damage.

Basically, your friend must have done something wrong when doing the test. The ships have to be absolutely still for artillery to hit as its tracking is VERY low. At low ranges it can barely hit anyghing. That's what keeps AC's viable.

http://pvpwannabe.blogspot.com/
Othran
Route One
#14 - 2011-09-08 15:26:59 UTC  |  Edited by: Othran
The problem is that youre not dealing with REAL mathematics here, you're dealing with various maths functions in programming languages. They are not accurate.

Clearly raising a number to the power of zero in REAL maths results in an indeterminate result. In many programming languages it does not if you call the wrong function - you'll get a result of 1 for pwr just to give an example. Edit - likewise 0/0 may give you NaN, +infinity, -infinity or 1 depending on the maths library your calls reference.

I suspect there might be a bug regarding whether the last recorded transversal is "positive" or "negative". This might be what you're on about regarding residual movement.

It isn't likely to get "fixed" as the chances of it occurring in real combat are unlikely. Better they put the (limited) resources elsewhere IMHO.
Othran
Route One
#15 - 2011-09-08 15:35:20 UTC
KFenn wrote:
Othran wrote:
It doesn't.

Lets take a 200mm AC2 and a 280mm Howitzer2 (both top of the "small" class) :

200mm AC2 has 1200 optimal and 4800 falloff. The falloff is four times the optimal;

280 Howitzer2 has 12000 optimal and 8750 falloff. The falloff is 0.72 times the optimal.

In short you're talking nonsense so I'd stop now were I you Blink

Wait, what? Now you're talking optimal relative to falloff, instead of just Arty falloff vs. Auto falloff. 8750 > 4800, so in that regard it does have comparable (well, far superior) falloff.

I believe it is you, good sir, who is talking nonsense.


I guess it could be taken both ways but I'd like to see you argue it in a HAC gang Blink

For me I look at the figures on the gun. You fit for falloff on AC and optimal on arty.

Interesting thread for once. Not sure it actually helps anyone who flies outside zero orbit though Big smile
Alua Oresson
Aegis Ascending
Solyaris Chtonium
#16 - 2011-09-08 15:54:36 UTC  |  Edited by: Alua Oresson
Othran wrote:
The problem is that youre not dealing with REAL mathematics here, you're dealing with various maths functions in programming languages. They are not accurate.

Clearly raising a number to the power of zero in REAL maths results in an indeterminate result. In many programming languages it does not if you call the wrong function - you'll get a result of 1 for pwr just to give an example. Edit - likewise 0/0 may give you NaN, +infinity, -infinity or 1 depending on the maths library your calls reference.

I suspect there might be a bug regarding whether the last recorded transversal is "positive" or "negative". This might be what you're on about regarding residual movement.

It isn't likely to get "fixed" as the chances of it occurring in real combat are unlikely. Better they put the (limited) resources elsewhere IMHO.



This could also be the case. You are right about sometimes computer languages not calculating correctly. It could be a combination of being at zero range and zero movement. The point is that it doesn't really matter because in actual fighting you aren't going to be sitting still relative to whoever you are shooting.


EDIT:

Also that calculation isn't directly from CCP (At least I seem to remember reading that at one point). I believe they stated at one time that the calculation isn't exactly what they use, but is close. So, there could be something in the actual calculations used that is causing the artillery to not hit such as a variable thrown into the calculation to always have some movement so that the computers don't have a coronary trying to divide 0/0.

http://pvpwannabe.blogspot.com/
Othran
Route One
#17 - 2011-09-08 16:19:26 UTC  |  Edited by: Othran
Yeah without knowing exactly what the code says and the maths libs involved it could be something as simple as the wrong function being called.

I do have to say that I remember similar complaints about not hitting at zero several years ago.

Its a rare condition anyway.

What would be interesting is if two (skilled in flying terms) corpies got on ships at zero and flew at exactly the same speed aligned to a planet say. Then one slowed by say 10m/s. Then bring it back up to same speed.

Same guns similar skills so all you're altering is transversal.

Would be interesting to see your results.
whaynethepain
#18 - 2011-09-09 01:14:27 UTC
I know exactly what you mean, which makes a change.

Optimal, and optimum are two different meanings.

within and at, are the closest English words I can think off, no doubt some foreign person will "correct" this definition, add some numbers and letters, and call me a fool, but it is that simple.

CCP used the within option.

Getting you on your feet.

So you've further to fall.
Bethesda Vortarhiat
Godlike Wannabe
#19 - 2011-09-09 08:27:19 UTC
First of all i want to thank you all who wrote here,everyone had very interesting things to say and proved that you know what this game is about.
@waynethepain : I don t know what you meant by optimum and optimal...and a foreign person "correcting" this...i am also foreign..but i understand the difference between words,even tho they might look and sound same.

Anyway...let me give you an inside of the chat i had with my friend...the chat that brought me here to ask for your advice.

Friend : can you undock your onyx so i can do a small test with my arty maelstrom ?

Me : no problem - i undock my onyx...friend is in 3km - we both stationary

Me - ok go ahead

Friend - wait a second...i need to get into optimal range

Me : optimal range ? you already are in optimal range.....

Friend : nope...optimal is at 15km

Me : lol....optimal range is 0 to 15km

Friend : nope...could start somewhere under 15km...but definitely not from 0....cos i never hit anything in 0

Me : m8....optimal range for any kind of turret is a RANGE between 0 and xx km ...so if you shoot within that range you have maximum chance to hit...and do full damage etc....i have to point out that you have to be stationary...if you miss within optimal range its because of tracking..

Friend : nope...not true...optimal range for arty doesnt start from 0...maybe it starts from 0 for AC's because i can hit in 0 with them....but not for arty...i cannot hit anything in 0 even if me and target are stationary....cant even hit a BS

Me : ok...so...i m totally lost now..
Cyniac
Federal Navy Academy
Gallente Federation
#20 - 2011-09-09 11:40:12 UTC
The optimal vs optimum range is an interesting one.

Optimal range - whatever the guns say it is, in practice it's an attribute in range

Optimum range - the ideal fighting range for a ship. Note that ideal range will factor in not only damage output, but also speed of the ship vs tracking of guns and the range of other combat effects (e.g. want to be in or out of scram range?) but also something as simple as where a pilot is comfortable fighting. (Some like to be up close and in your face, others like to be farther out)

I tend to just call it the fighting range but whatever - yeah they are different.
12Next page