Version 0.2.2 LK is available via the download link in the first post. Changes:
- Many improvements to formulas for application of % bonuses from talents, especially Find Weakness.
- Mutilate formula updated to be consistent with reality.
- All items with armor penetration converted to armor penetration rating.
- Armor penetration rating and new Mace Specialization implemented.
- Wrath rank of Backstab implemented.
- Improved Poisons reduced to 3% per talent point, per build 8926.
- Implemented all known major glyphs except Sinister Strike.
If you have any prior versions of the LK sheet, can you try them and let me know if the bug occurs there as well? Alternatively, can you simply screenshot what it looks like when the bug occurs? (edit) And also check for occurrence of any other bugs anywhere else in the sheet, including but not necessarily limited to duplicates of the same bug on the other sheets with the DPS header?
The bug is present back to 0.2.0 LK. I've uploaded screenshots of the affected cells here, here and here. I only see the bug in the "DPS", "Cooldown 1", "Cooldown 2", and "Cooldown 3" sheets, while the other "DPS Calc" sheets seem fine.
Last edited by Havenwood : 09/14/08 at 12:50 PM.
Reason: Added screenshot
I've seen plenty of discussion regarding optimized woltk builds, but I haven't seen much mention of 3.0 builds optimized for lvl 70. Is assassination as overpowered at 70 as it is at 80? is combat competitive?
The definition of Err508 is that there is a missing parenthesis but this does not seem to be the case.
Cells with the error are:
DPS:
B20 - 22, C20 - 22
B39 - 40
B44
B50
Cooldowns 3:
Tons here. The first 2 cells in the Lust Iter column work fine, but the rest don't and use a different formula. The second cell in Lust Timing works fine, but the rest don't and use a different formula. All of Lust Active and Haste % are broken, as well as U1 & X1 (but those are probably chained).
Cooldowns 3 seems to be the root of the problem, with the rest cascading.
Version 0.2.3 LK is available for download via the link in the first post. Changes:
- Fixed an error in "Cooldowns 1", "Cooldowns 2", and "Cooldowns 3" sheets causing Open Office issues.
Turns out that one cell in each sheet was referencing some named cells that no longer existed. As a result, Open Office threw a fit. That's fixed now, should work fine.
I was plugging in a bunch of various hemo and combat builds into the sheet and thought it would be worth sharing the results as I was a little surprised. Knowing that Mutilate is the king right now I wanted to see where other specs would be with the current talents as they are. I lowered the default gear to what I currently have which is end t6 level (armory me for more detail). Weapons used were Vanir's Right Fist, Vengeful Glad. Slicer, BoS, Fang of Kalecgos and Blade of Serration and were changed where appropriate. For reference I included the default talents in the sheet with my changed gear. For the Hemo builds I've shown the estimated contribution for the Raid Debuff and the 3rd column is the personal+raid debuff.
I assume one of the reasons combat is coming in so low is because the Sinister Strike Glyph is not implemented in the sheet yet. Is that still the case? Also, none of the builds I input had Killing Spree since it is not implemented in the sheet yet.
Wouldn't happen to have any numbers between 15/51/5 and 7/51/13 Deep Combat builds? What I'm wondering is Ruthlessness + Lethality vs Serrated Blades. Also are there any numbers how and "when" new Mace Specialization is calculated?
Based on testing in the Combat Ratings at 80 thread, we're pretty sure armor penetration rating and Mace Specialization are applied after flat armor reductions like Sunder Armor and Faerie Fire.
I searched through the spreadsheet and this topic, but I didn't find anything about the racial change of dwarves. Is this already implemented or are you working on it?
it is possible to change focus attacks and poison dmg manually until the next version come out ?
I *think*I hacked the rogue spreadsheet correctly to reflect the new build. Corrected cell are highlighted in red.
As a quick and dirty fix, I set Focused Attacked to 10 talent points to reflect 100% crit
Infectious Poisons obviously has to be zeroed out
On the DPS page, cells O9 and O13 were changed to reflect the lower benefit to proc rate after envenoming
New IP damage and coefficient are in cells O10 and O11.
New DP damage and coefficient is in cell N21.
Edit: Took it down, as we got a new version released and new beta talents anyway.
A lot of the differrences I've noted before are still there, but far less drastic.
Mutilate is back ahead of Shiv, not by a whole lot at 70 with the builds I was plugging in, but Mutilate pulls ahead quite a bit at 80 when you get the points to put opportunity in.
Weapon speed still matters, but not as much: Faster is still definitely better in the offhand but overall Average Hit, stats, and Ilvl have a say in the matter if the speeds are close.
E.g. [Edge of Oppression] is still the best OH but is about equal to or slightly worse than [Messenger of Fate] in the MH, which in turn is about equal or slightly worse than [Fang of Kalecgos]. 1.8 speed still looks like its slow for daggers in the main hand though.
I also took a quick look at gemming: I think AGI is what you want to stack, seems reasonable since AP matters for poison and you have even more crit-reliant talents.
Keep in mind I did all this before crashing last night so please let me know if I'm way off mark on any of this.
Oops! Found some pretty significant errors that needed immediate fixing. Version 0.2.5 LK is now available via the download link in the first post. Changes:
- Fixed several issues with Seal Fate, Glyph of Sinister Strike, and cycle calculations.
Please re-download the new version if you just downloaded 0.2.4. Sorry for the inconvenience.
-----
Version 0.2.4 LK is available via the download link in the first post. Changes:
- Improved modeling of Envenom's effect on Deadly Poison uptime and Envenom's poison proc rate buff.
- Infectious Poisons removed.
- Envenom poison proc rate buff reduced to 15%.
- Instant Poison and Deadly Poison damage and AP contribution reduced per build 8962.
- Focused Attacks increased to 100% chance per crit to gain 1 energy per talent point.
- T7 Bonescythe set bonuses implemented, but the gear will come next version. (If you're clever, you can find a way to fiddle with them right now!)
- Glyph of Sinister Strike implemented.
- Fixed an error in Envenom DPS calculations for Xe/Yr (CttC) cycles.
Deadly Brew is not "unplugged." If you wish to simulate the existing bug where Instant Poison procs Crippling Poison (rather than Deadly), simply remove your talent points from Deadly Brew.
New enchants/gear/etc. are my goal for 0.3.0, so hopefully there won't be any more major changes in beta builds until I get that done.
So according to this post, Focused Attacks does not actually proc on Instant Poison crits, as was speculated. This sheet currently implements Focused Attacks as proccing on Instant Poison crits. To fix this, you can make the following changes yourself:
Format menu > Sheet > Unhide... > DPS Calc 1
Format menu > Sheet > Unhide... > DPS Calc 2
Format menu > Sheet > Unhide... > DPS Calc 3
Go to DPS Calc 1 sheet
Change cell O5:
Original:
=(H30+I30+(N28+N29)*N25)
Modified:
=(H30+I30)
Repeat this change in cell O5 on DPS Calc 2
Repeat this change in cell O5 on DPS Calc 3
That's all. To hide those sheets again, go to each sheet one at a time, Format > Sheet > Hide.
Version 0.2.6 LK is now available via the download link in the first post. Changes:
- Focused Attacks modified to no longer proc off poisons, per updated information.
- Updated all talents per build 8970.
The build included in this is 20/51/0 with Warglaives, providing 3210 DPS. The alternative Combat build, 7/51/13, was shown to provide 3215 DPS (feel free to check this yourself after downloading). The "Saved DPS" amount that I've left in there is for a Mutilate build (51/13/7) using the exact same gear and glyphs and other settings, except with Crux of the Apocalypse and Shiv of Exsanguination instead of Warglaives. Note that the Mutilate build still provides 15% more DPS.
As a bonus, try swapping to Mutilate with the aforementioned weapons -- and then switch the Shiv of Exsanguination for Warp Splinter's Thorn.
I was wondering if there was any way i could personally change the honor among thieves to generate a 1cp/sec to do a rough test of how the dps would possibly be in a raid with a 5s/5r/5e cycle.
On my list of things to implement, Honor Among Thieves is pretty far down near the bottom. The reasons for this are:
- Subtlety has historically been terrible for DPS.
- Aside from Honor Among Thieves and Slaughter from the Shadows, I see no reason to believe Subtlety will surpass either of the other builds.
- Honor Among Thieves will be complicated to integrate into the cycle models due to variability in CP generated, the fact that the model is designed around CP being generated by either finishers or combo moves, and the irregularity of the cycles that would result.
I pretty much have no framework for implementing this talent at the moment, so I can give you no guidance or assistance on trying to hack it in yourself.
So something seems a little odd with the Deadly Poison stack counting.
Her are some results with MH DP, OH IP and a standard 51 points in Assassination (CttC, poison talents etc):
If I set the finisher to Xe CttC with Eviscerate it says 99.23% uptime. That makes sense, it should never drop basically since I'm never getting rid of the stack.
If I change it to Xe CttC with Envenom instead of Eviscerate, it says 99.65% uptime. Even though Envenom will buff my chance for it to land, I would think the uptime would not be that high.
Maybe it really would be that way but I dunno, it seems high.
I tried 0/50/21 specc because i think its better than assa/combat and the spreadsheet reflect this thought,
but it shows the best dps with 3s/5r .
I think its easy possible to do 1s/5r cycle without the energy from combat potency, but with this energy and a switch to 5s/5r/xx cycle for more energy from relentless strikes there should be so much more possible to do with all this combo points.
Is something wrong with the spreadsheet in this way or is 3s/5r really the best cycle and i have to waste so much combo points?
So something seems a little odd with the Deadly Poison stack counting.
Her are some results with MH DP, OH IP and a standard 51 points in Assassination (CttC, poison talents etc):
If I set the finisher to Xe CttC with Eviscerate it says 99.23% uptime. That makes sense, it should never drop basically since I'm never getting rid of the stack.
If I change it to Xe CttC with Envenom instead of Eviscerate, it says 99.65% uptime. Even though Envenom will buff my chance for it to land, I would think the uptime would not be that high.
Maybe it really would be that way but I dunno, it seems high.
So it occurred to me, in originally responding to your question, that my Deadly Poison uptime model was slightly flawed. But let's go through how I've been approaching it.
To start with, I want to know the likelihood of being at zero stacks of Deadly Poison at time T seconds. Well, to be at zero stacks, there are two things that could have occurred: I didn't proc Deadly Poison for a while, or else I performed Envenom recently. To be more specific, we know that Deadly Poison lasts 12 seconds, so it will fall off if I fail to proc it for 12 seconds. As for Envenom, we know that if I perform Envenom at time T, then I will be at zero stacks. What if I performed Envenom at time T-2? I would only still be at zero stacks if I failed to proc Deadly Poison in those 2 seconds. What if I performed Envenom at time T-12? Well, that Envenom doesn't technically affect my current situation, because even if I hadn't Envenomed at that time, if I had failed to proc Deadly Poison for the next 12 seconds it would have dropped.
So essentially we are concerned with the two conditions:
1) Did Deadly Poison fail to proc for 12 seconds?
2) Did an Envenom occur at time (T - X), followed by Deadly Poison failing to proc for X seconds, where 0 ≤ X ≤ 11?
Here's where I had erred. I had set up the formula to calculate the chance that, for every second over a period of 12 seconds, either the player did not proc Deadly Poison or he performed Envenom. The problem with this approach is that it does not acknowledge the fact that an Envenom at time (T - X) eliminates the importance of any Deadly Poison procs or non-procs in the range of time [(T - 12), (T - X)].
What I actually needed was a power series to add the chance of an Envenom 0 seconds ago and 0 seconds of Deadly Poison no-procs plus an Envenom 1 second ago and 1 second of Deadly Poison no-procs plus an Envenom 2 seconds ago and 2 seconds of Deadly Poison no-procs plus ... plus 12 seconds of Deadly Poison no-procs. Once I set this up, the calculated Deadly Poison uptime decreased substantially; in the provided build, the average number of Deadly Poison stacks decreased from nearly 5 to roughly 3.70.
One hour passed after typing that previous paragraph, and what I realized in that time was that simply using the conjunction of "Envenom X seconds ago" and "no Deadly Poison procs for X seconds" wasn't enough, because this would doubly count possibilities such as Envenoms occurring at both times T-10 and T-7, with no Deadly Poison procs from T-10 onwards. Thus, the proper conjunction would be "Envenom X seconds ago" with "no Deadly Poison procs or Envenoms for X seconds." Using this proper value increased the average Deadly Poison stacks to 3.80.
In the end, the DPS difference resulting from this error came out to be roughly 165. Thanks for encouraging me to go back and take a closer look at it.
(edit) Using this improved model, I'm able to get Mutilate DPS of 3644.94, and Combat swords DPS of 3286.15, a difference of about 10%.
Turning off all buffs with the same builds, I see Mutilate DPS of 1653.54, and Combat swords DPS of 1422.34, a difference of about 14%.
(edit 2) Ninja-updated the download link in the first post with the Deadly Poison change. If you downloaded 0.2.6 LK before 2:39 AM PDT this morning, please download it again now.
Cool. I didn't follow you completely (power series were never my strong suit), but the new uptime calculations definitely seem far more in line with what I've been seeing in the field. I'm glad I was able to help, if questioning really counts as helping!
So the interesting thing is now with the changes they've pushed through this week in addition to having more accurate DP numbers, you really start to see a number breakdown which I suspected was what they were shooting for. With my current gear (which is basically half substantially better than Sunwell, and half substantially worse) and the theoretical top DPS build, you see that 4+e 4+r is substantially better than just Envenom spam (though the point Aldriana makes about Rupture not actually fading in time merits looking at). In fact, in my case I would do over 6% DPS if I went with just Envenom spam.
Additionally, you see percentages which I think are very reasonable and in line with what they want:
White DPS - ~32%
Mutilate DPS - ~23%
Poison DPS - ~19%
Rupture DPS - ~9%
Envenom DPS - ~5.4%
Now what I'd really like to do is be able to test this in some raid encounters this weekend, but unless they hotfix the "miss rating" bug it will be impossible to get a valid data set.
I see a smal anomaly with the latest LK 0.2.6 spreadsheet (the one after the ninja update of the link)
I use the gear in my armory profile with some changes:
a) weapons are Shard of Azzinoth and Fang of Kalecgos.
b) Ring 1 Angelistas Revenge instead of Stormrage
c) trinket 1 Shard of Contempt
d) trinket 2 Ashtongue
Talent build is this.
Race human, Cycle 4+e/4+r (Cttc) with energy pooling set to 35/40.
If I now deselect ruthlessness I see an increase in dps by about 8 dps. This looks counterintutive to me.
Note if I increase energypooling to 60/60 (resulting in a net loss of 3 dps in the full build) deselecting Ruthlessness results in a loss of DPS as expected.
It's possible that's not a bug -- if you had Envenom activated. Performing Envenom more often doesn't necessarily mean more DPS, since it means your Deadly Poison stack drops more often. The increased uptime of the Envenom poison proc rate buff might not offset that.
I'll double check the calcs tomorrow, but I think they're actually right on this one.
09/14/08, 4:58 AM
Similar Threads
