| Pages: 1 2 3 [4] :: one page |
| Author |
Thread Statistics | Show CCP posts - 2 post(s) |

Akita T
Caldari Caldari Navy Volunteer Task Force
|
Posted - 2007.05.03 12:09:00 -
[91]
Originally by: dalman Read above in this thread, and I posted the basics of how hitchance also determine the quality of hits and expected average damage.
Originally by: Akita T Just re-stating the obvious (and already mentioned a bit higher on the thread)...
I said that too 
Originally by: dalman
Originally by: Akita T If {x<z} Then {damage = quality x expected base damage}
However, the code they suggested does not account for the difference in the 3 categories of misses which all do 0 damage. My thoughts does (by having intervals < 0 < 0.25 < 0.50), which makes me wonder if their code really is closer to exactly how it's done.
Well, the "If {x<z} Then {damage = quality x expected base damage}" part could then be expanded into "If {x<z} Then {damage = quality x expected base damage} elseif {miss_message_generating_sequence(x,z)}" that takes those ranges into account too. Practically, for average DPS calculation purposes, it really makes no difference whatsoever.
Originally by: dalman Regardless, a good enough estimate of the forumla for average damage has been provided - and it's been shown that "wrecking hits are broke". And the tracking guide should be updated to reflect this.
I wouldn't quite say that wrecking hits are broken. Weird and counter-intuitive, but not broken.
They still make up AT MOST 1% of shots (in case hit chance is 1% or higher)... in case the "to hit" chance drops below 1%, all shots are indeed wrecking shots, but they're no longer 1% of total shots. If you only have a 0.1% to hit chance, you hit only 0.1% of the time (so one out of 1000 shots, not 1% or 1 out of 100 shots), but indeed, each one of those "one in 1000 shots" is a wrecking shot, for an average of 0.3% from the base DPS (not 3%).
The tracking guide doesn't even reflect the basics of tracking (misrepresenting what signature radius and tracking value actually means), let alone true average DPS, with hit quality factored in. _ MySkills | Module/Rig stacknerfing explained |

dalman
Finite Horizon
|
Posted - 2007.05.03 17:30:00 -
[92]
Edited by: dalman on 03/05/2007 17:48:23
Originally by: Akita T
Originally by: dalman
Originally by: Akita T If {x<z} Then {damage = quality x expected base damage}
However, the code they suggested does not account for the difference in the 3 categories of misses which all do 0 damage. My thoughts does (by having intervals < 0 < 0.25 < 0.50), which makes me wonder if their code really is closer to exactly how it's done.
Well, the "If {x<z} Then {damage = quality x expected base damage}" part could then be expanded into "If {x<z} Then {damage = quality x expected base damage} elseif {miss_message_generating_sequence(x,z)}" that takes those ranges into account too. Practically, for average DPS calculation purposes, it really makes no difference whatsoever.
Yea, for average DPS it doesn't have any purpose - only for trying to establish exactly how it's done (and to get the very exact formula for DPS). What I meant was that we have ("hit description" in "damagemod intervals"): miss< 0 < close miss < 0.25 < glancing miss < 0.5 < scratch < 0.625 < light hit < 0.75 < hit < 1.0 < well aimed < exellent < 1.49 If you have a hitchance of 10% (0.1), your shots will be distributed in a damagemod [-0.4, 0.59]. And you'll have 40% misses, 25% closely misses, 25% glancing misses and 10% scratching hits. (well, 1% will be wreckings, so remove 1% from one of these intervals - it's not really certain which one) Same way, if your hitchance is 90% you'll never ever get a "miss" or "closely misses". The 10% missed shots will all be "glancing misses".
Goumin simply used if x>z then shot = miss. Without any way that explains the 3 types of misses.
And wreckings are broke, as they said it was changed to 1% of all hits (it was 1% of all shots before). If the random number is generated with a whole lot of decimals, it should be extremely easy to fix. All that would be needed is for the "wrecking interval" to not be 0.01 as it is today. As in using > 1-hitchance/100 if it's cropped at top, or <0+hitchance/100 if it's cropped at the bottom. Am I forced to have any regret? I've become the lie, beautiful and free In my righteous own mind I adore and preach the insanity you gave to me |

Kartissa Elentari
|
Posted - 2007.05.03 18:27:00 -
[93]
Originally by: Akita T Edited by: Akita T on 03/05/2007 11:08:41
For the umpteenth time, this is what the tracking guide MISLEADS you to believe. In fact, it's as simple as this: sigradius ratio (target sig / gun sig) is just a tracking value MULTIPLIER, nothing more, nothing less.
In other words, if a target has a sig 2 times smaller as you gun sig, it's as if your turrets have only half tracking. If a target has sig 2 times larger as your gun sig, it's as if your turrets have two times better tracking as they have listed.
And that's IT. Nothing else.
I stand corrected, although I was basing my explanation mainly on real-world physics, rather than the tracking guide. (Which I just discovered is linked to in the first post on this thread! Do I feel a fool...?)
I suppose I should have known better than to rely on the real world for an explanation of in-game mechanics. I already worked out that Eve doesn't use the usual distance for 1 Astronomical Unit.
|

Gentzen
Izanagi Alliance
|
Posted - 2007.05.04 13:42:00 -
[94]
Edited by: Gentzen on 04/05/2007 13:40:25
Originally by: Akita T Edited by: Akita T on 03/05/2007 11:08:41 ___
ALSO, there's the "hit quality" issue. According to recent research (not done by me, but by Kazuo Ishiguro), you have a situation like this:
x = random number between 0 and 1 z = ((1/2)^((((Transv/(Range*Tracking))*(Sig_Res/Sig_Rad))^2) +((max(0,Range-Optimal))/Falloff)^2)) z is the "To-Hit Chance", a number between 0 (0%) and 1 (100%)
If {x<0.01} Then {quality = 3} else {quality = x + 0.49} might be x+0.5 for all I know, or even x+0.495, no big difference If {x<z} Then {damage = quality x expected base damage} __
You only deal the "lower quality" hits when hit chance is lower, and the higher quality hits get discarded... so that would imply average DPS decreases a bit sharper as hit chance decreases (i.e. not linearly). It also means that at exactly 1% hit chance, you start ONLY dealing wrecking shots, for a 3% of "expected average" DPS.
A 50% hit chance would mean 1% wrecking shots and 49% normal hits linearly distributed in the *0.5 to *0.99 (or *1.00?) damage range, for a final DPS of a bit under 39.75% instead of the "expected" ~52% DPS. ____
Just re-stating the obvious (and already mentioned a bit higher on the thread)...
In other words, average DPS (simplified form for chance to hit at or ABOVE 1%) is:
AvgDPS = BaseDPS * [ 0.03 + (ChanceToHit-0.01)*(0.5 + ChanceToHit/2) ] Where ChanceToHit is 0.5^((((Transv/(Range*Tracking))*(Gun_Sig/Target_Sig))^2) +((max(0,Range-Optimal))/Falloff)^2)
Simplified form for chance to hit at or BELOW 1%, AvgDPS = BaseDPS * 3 * ChanceToHit
very interesting. though I don't know to what extent is that equation correct, I added that equation to Turret Calculator experimentally.
Turret Calculator
----- [tool]Turret Calculator, admin of evewiki.info |

Soltueur
|
Posted - 2007.05.05 18:00:00 -
[95]
The link to the turret tracking guide takes you to the missile guide...
|

Gentzen
Izanagi Alliance
|
Posted - 2007.05.05 22:32:00 -
[96]
Originally by: Soltueur The link to the turret tracking guide takes you to the missile guide...
thanks. fixed. ----- [tool]Turret Calculator, admin of evewiki.info |

DaJokr
Carebear Collective
|
Posted - 2007.05.09 21:56:00 -
[97]
I was planning on moving out of Missle Specialization into gunnery, but after reading all this, I'm more confused now then when i started...it's the whole concept of the transversal. I'm young and simple...thats just over my head. Anyone dumb this down to me so that a toddler could understand?
|

Stephanie Var
|
Posted - 2007.05.11 04:49:00 -
[98]
The Link for this guide is still taking you to the missiles .. after a while of randomly picking numbers in the address bar I found it.
The last page is really awsome for picking weapons
http://www.eve-online.com/guide/en/g25.asp
Happy Shooting
|

Swinton Wolsoncroft
Caldari GoonFleet GoonSwarm
|
Posted - 2007.05.25 21:53:00 -
[99]
If one follows the guide below you will be able to track and destroy Polaris frigs
[ 2007.03.20 06:30:15 ] Orange Species > isd reporters are not to interfere with gameplay [ 2007.03.20 06:30:22 ] Orange Species > that includes getting too close [ 2007.03.20 06:30:25 ] Orange Species > to our ships [ 2007.03.20 06:30:32 ] D4kkon > petition her [ 2007.03.20 06:30:57 ] Orange Species > nah no petition [ 2007.03.20 06:31:00 ] Orange Species > msn chat is enough [ 2007.03.20 06:31:05 ] D4kkon > rgr
[ 2007.03.20 06:30:15 ] Orange Species > isd reporters are not to interfere with gameplay [ 2007.03.20 06:30:22 ] Orange Species > that includes getting too close [ 2007.03.2 |

ZedBias
|
Posted - 2007.06.02 17:16:00 -
[100]
Originally by: Gentzen
Originally by: Soltueur The link to the turret tracking guide takes you to the missile guide...
thanks. fixed.
Its not fixed for me, still taking me to missle guide. 
|

Ararius
e X i l e FREGE Alliance
|
Posted - 2007.06.10 07:54:00 -
[101]
It links to the missle guide...
Originally by: Alski It depends how much CCP learn from past user interface mistakes, i'm sure they'll do very well, but if not i'll be right clicking on random people and selecting "Orbit at 0.5m" 
|

Jared Byar
Minmatar Brutor tribe
|
Posted - 2007.06.19 07:13:00 -
[102]
Yes indeed leading to the missle section. Not that it matters much since you can clicky on the tracking guide from there.
|

Overclocker BG
|
Posted - 2007.07.23 22:10:00 -
[103]
mhm
|

Exuscon
Amarr
|
Posted - 2007.11.01 04:45:00 -
[104]
For such a high tech setting for a game you think these Idiot races would invent a gyroscope?
|

Ancy Denaries
Caldari Isseras Manufacture
|
Posted - 2007.11.01 08:38:00 -
[105]
Most necroed thread in the world award?
|
| |
|
| Pages: 1 2 3 [4] :: one page |
| First page | Previous page | Next page | Last page |