Results 1 to 15 of 15

Thread: Disintegration

  1. #1
    Community Member Kinerd's Avatar
    Join Date
    Mar 2010
    Posts
    5,087

    Default Disintegration

    Short Version
    Process rate: 2%
    Average damage: 460
    Dice Distribution: 20d20 + 250

    .

    Long Version

    Process Rate
    I didn't have (or anticipate having) any weapons with Disintegration, so I used a Fleshmaker's Docent and recorded enemy attacks. (The rats never saw it coming.) I verified that the guard could process on a 0 damage hit. I have no way of demonstrating that Disintegration Guard uses the same numbers as Disintegration, but to my mind it would be awfully weird if it didn't. All told, there were 7779 hits and 160 disintegrations. The overall process rate is therefore:
    160 / 7779 = 2.06%
    With standard deviation:
    sqrt (7779 * 2.06% * (1 - 2.06%)) = 12.52

    Therefore, the overall measurement is 2.06% ± 0.48% at 99.7% confidence. I also kept individual recordings along the way in bins of 40:
    Code:
    Hits	Procs
    2834	40
    4142	80
    5677	120
    7779	160
    The same procedure gives possible ranges for process rate for each bin as:
    Code:
    Min	Max
    0.75%	2.08%
    1.63%	4.49%
    1.39%	3.83%
    1.01%	2.80%
    Hence, I conclude that the process rate is 2%.

    Damage Dice
    I recorded the damages done on each disintegration, and they are as follows:


    Using the method detailed in this post, I obtained a standard deviation of 26 and potential dice distributions of:
    Code:
    d6	231.77
    d8	128.76
    d10	81.94
    d12	56.73
    d20	20.33
    d25	13.00
    As with Lightning Strike, the most likely candidate to me appears to be 20d20, in this case + 250 rather than + 400. Following the advice given to me in a private message, I have used kernel density estimation to give a better picture of the data with respect to the distribution:

  2. #2
    Community Member Grenada's Avatar
    Join Date
    Dec 2010
    Posts
    541

    Default

    Yay! More numbers!

    Also, if this is the same for weapons, I think disintegrate does more damage than Lightning strike, in terms of dps.

    Disintegrate:

    0.95 (chance of hitting) * 460 * 0.02 * N/t (N= swings, t = time in seconds)

    =8.74 N/t


    Lightning strike:

    0.95 * 610 * 0.015 * N/t

    =8.6925 N/t
    Quote Originally Posted by Ungood View Post
    You may now return to your WoW bashing.
    Fweebie
    "Crouching Tiger Hidden Nerfbat!"

  3. #3
    Community Member TheDearLeader's Avatar
    Join Date
    Oct 2009
    Posts
    2,217

    Default

    Disintegrate certainly has a wider variety of mobs it can work on. Nice work (yet again).

  4. #4
    The Hatchery Vissarion's Avatar
    Join Date
    Oct 2009
    Posts
    500

    Default

    Love your work, Kinerd.

    I reckon you need an appropriate forum title for your statistical nerdiness.
    Order of the Silver Dragons||Argonessen
    Vratch||~Terevorn~||Ignys||Zofya||Virochka
    Theologians don't know nothin' about my soul.

  5. #5
    Community Member voodoogroves's Avatar
    Join Date
    Oct 2009
    Posts
    8,366

    Default

    Quote Originally Posted by Grenada View Post
    Yay! More numbers!

    Also, if this is the same for weapons, I think disintegrate does more damage than Lightning strike, in terms of dps.

    Disintegrate:

    0.95 (chance of hitting) * 460 * 0.02 * N/t (N= swings, t = time in seconds)

    =8.74 N/t


    Lightning strike:

    0.95 * 610 * 0.015 * N/t

    =8.6925 N/t
    It certainly does, all things being equal. Greensteel-wise though, Lit2 comes with a better T1
    Ghallanda - now with fewer alts and more ghostbane

  6. #6
    Community Member Kinerd's Avatar
    Join Date
    Mar 2010
    Posts
    5,087

    Default

    Thanks all!

    It was suggested to me privately that 15d20 + 300 is the correct distribution. The average value for this distribution is 457.5, so the practical difference between the two is effectively zero, but I think it's still worth looking at.

    Here is the kernel density estimation graph with 15d20 + 300 overlaid in red:

    Visually the 15d20 certainly seems further off. The R^2 values are .9937 for the 20d20 and .9476 for the 15d20. The standard deviation would also have to be smaller: 22 instead of 26, which doesn't seem like a negligible difference to me. In short, the 20d20 is a better fit to the data I have.

    I also meant to cite the software I used to do the kernel density estimation, which can be found at the RSC web site.

  7. #7
    The Hatchery sirgog's Avatar
    Join Date
    Apr 2007
    Posts
    11,175

    Default

    Quote Originally Posted by voodoogroves View Post
    It certainly does, all things being equal. Greensteel-wise though, Lit2 comes with a better T1
    This is the big factor. Corrosive Salt is almost as good as Disintegration and Lightning Strike in raw damage output too, but also has no negative option that holds a candle to Holy or even Goodburst.

  8. #8
    Community Member voodoogroves's Avatar
    Join Date
    Oct 2009
    Posts
    8,366

    Default

    Quote Originally Posted by sirgog View Post
    This is the big factor. Corrosive Salt is almost as good as Disintegration and Lightning Strike in raw damage output too, but also has no negative option that holds a candle to Holy or even Goodburst.
    Yep. Dual shard too ... so painful.

    I've got an "oops" Unholy I made a while back, but honestly with that I'm more likely to go TTS than Disintegration.
    Ghallanda - now with fewer alts and more ghostbane

  9. #9
    Time Bandit
    Join Date
    Sep 2009
    Posts
    141

    Default

    By the way I sort of advise against assuming that guard proc rate = weapon proc rate until there's been confirmation of that for a few effects, since it's possible that they go under different sections of code. One is an attacker -> victim subroutine, while the other is a victim -> attacker subroutine; it's possible that the programmers just do lightning_strike(victim, attacker) or lightning_strike(attacker, victim) as appropriate, but it's also possible that they wrote separate code for each. On the other hand, guard proc rates are still important for tanks, so they're worthwhile to know as well in their own right. I was looking for guard proc information back when deciding which GS HP goggles to make and wasn't very satisfied with what's out there, so ended up just doing lightning strike on a guess.

    So it doesn't mean that they're not the same, just that the assumption should be tested out on a few effects first. I have lightning guard on my barbarian so I can test that out I suppose. Depending on your server, people also may or may not be willing to donate GS mats to make test weapons.

    P.S. You mentioned earlier in your lightning strike thread that you were going to test the radiant blast effect on the Mabar wraps. Not sure how far along you are with it, I finally capped my monk so I'll be releasing some numbers on it soon. EDIT: numbers here.

  10. #10
    Founder LeLoric's Avatar
    Join Date
    Feb 2006
    Posts
    2,903

    Default

    Quote Originally Posted by sirgog View Post
    This is the big factor. Corrosive Salt is almost as good as Disintegration and Lightning Strike in raw damage output too, but also has no negative option that holds a candle to Holy or even Goodburst.
    True however for consideration of alchemical it puts earth tier two really strong for twf. Especially if using offhand smallblade or even tier 3 spare hand to cover your double strike.

    Many monks in guild are going water tier two actually as they are using spare hand and the combo of crushing wave and corrosive salt is excellent dps with most of it not subject to any resist. Throw in +2 alchemical wisdom and it's almost a no brainer.
    Ghallanda Rerolled
    LeLodar LeLothian LeLoki LeLoman LeLonia LeLog

  11. #11
    The Hatchery sirgog's Avatar
    Join Date
    Apr 2007
    Posts
    11,175

    Default

    Quote Originally Posted by LeLoric View Post
    True however for consideration of alchemical it puts earth tier two really strong for twf. Especially if using offhand smallblade or even tier 3 spare hand to cover your double strike.

    Many monks in guild are going water tier two actually as they are using spare hand and the combo of crushing wave and corrosive salt is excellent dps with most of it not subject to any resist. Throw in +2 alchemical wisdom and it's almost a no brainer.
    Yeah T2 water is really solid for any build that cares about Wisdom.

  12. #12
    Community Member Kinerd's Avatar
    Join Date
    Mar 2010
    Posts
    5,087

    Default

    Quote Originally Posted by Vanshilar View Post
    By the way I sort of advise against assuming that guard proc rate = weapon proc rate until there's been confirmation of that for a few effects, since it's possible that they go under different sections of code. One is an attacker -> victim subroutine, while the other is a victim -> attacker subroutine; it's possible that the programmers just do lightning_strike(victim, attacker) or lightning_strike(attacker, victim) as appropriate, but it's also possible that they wrote separate code for each. On the other hand, guard proc rates are still important for tanks, so they're worthwhile to know as well in their own right. I was looking for guard proc information back when deciding which GS HP goggles to make and wasn't very satisfied with what's out there, so ended up just doing lightning strike on a guess.

    So it doesn't mean that they're not the same, just that the assumption should be tested out on a few effects first. I have lightning guard on my barbarian so I can test that out I suppose. Depending on your server, people also may or may not be willing to donate GS mats to make test weapons.
    Yes, I would much rather have tested with a Disintegration weapon, but such is life.
    P.S. You mentioned earlier in your lightning strike thread that you were going to test the radiant blast effect on the Mabar wraps. Not sure how far along you are with it, I finally capped my monk so I'll be releasing some numbers on it soon. EDIT: numbers here.
    When the challenge weapons came out I TRed my monk, as I was going to make a Mournlode weapon for my ranger anyway (and I really wanted to TR my monk). With nothing else to test I went with Disintegration Guard, but now there's a contradiction between my data and the information I've received from several anonymous sources, so I'm probably not going to get to Radiant Blast anytime soon.

  13. #13
    Hero dTarkanan's Avatar
    Join Date
    Jan 2010
    Posts
    388

    Default

    Are these sources reliable? Thanks for all your hard work.

  14. #14
    Community Member Delt's Avatar
    Join Date
    Jun 2006
    Posts
    0

    Default

    Nitpick time...why is "process" being used in lieu of proc?

  15. #15
    Community Member Kinerd's Avatar
    Join Date
    Mar 2010
    Posts
    5,087

    Default

    Quote Originally Posted by dTarkanan View Post
    Are these sources reliable? Thanks for all your hard work.
    You're welcome!

    I don't understand your question. The data in my first post is all empirical: get hit, disintegration goes off, write down the number, keep track of how many hits and processes. I don't know who or what the original source for the "15d20 + 300" I mentioned is, but due to other information I consider it reliable. I collected about as much information as I put in the first post to see what would happen, and the curve I got was smack dab in the middle of 15d20 + 300 and 20d20 + 250, so I said the heck with it, 15d20 + 300 is close enough.
    Quote Originally Posted by Delt
    Nitpick time...why is "process" being used in lieu of proc?
    I like the sound of it better. I don't always subvert etymology, but when I do I use aesthetics.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

This form's session has expired. You need to reload the page.

Reload