Musa

Yes, I also have a stack of sharpening stones on me for whenever I bump into immune bosses aka hydross, VR. Not that we need it, but I just want highest dps available.

I'd be happy to see them added if you ever reconsider. But as a stone user, I agree they aren't of any use except for these bosses.

Oh, and I do not always have WF available. So I am very confused. How do the 2 stones stack MH/OH?

 

• Aldriana

So, first off: grats on the warglaives. I think I speak for most of us when I say: /jealous

Second: Thanks for taking the time to take some data.

In terms of actual analysis: I show 94 periods of Warglaive uptime in that combatlog, and 5265 total hits. Grinding through the data, we find that of those 94 periods of uptime, there are 19 at least 11 seconds in length, including 5 at least 21 seconds in length, and 1 that is more than 31 seconds in length. Hence, there were, at minimum 122 distinct haste procs in the data set. In reality, there may have been a few more, but lets go with the 122 number for the moment.

122 procs in 5265 total hits works out to a proc rate of roughly 2.32% on the test, with a standard deviation of .21%; hence, if it's a fixed percent proc, the true value lies between 1.91% and 2.72%.

On the other hand, it might be a PPM effect; were it 1 PPM, one would expect to see 156 procs in that data set; since only 122 were observed, I think it somewhat unlikely that it's 1 PPM; it would need to be .8 PPM or so instead.

Unfortunately, due to the nature of the set, it's going to be hard to isolate the hands and figure out which hand procced what; hence figuring out the exact proc mechanics is going to be challenging. I can think of two pieces of data that might be helpful in determining it for certain:

1) If one stands sidewise to a mob, it is possible to find a position wherein you can still land specials on the mob, but not autoattack it. A data set of this sort, where there are no autoattacks present but only SS (or Shiv, your choice) spam, should be able to isolate one hand from the other.

2) Attacking the mob normally, but with using SS and SnD instead of straight autoattack. If it's a PPM effect, SnD will reduce the proc rate; if it's a fixed percent, it will not. This would provide some insight as to what sort of proc we're dealing with.

What we *can* say from this is that the set bonus is still sickeningly powerful, even with the haste nerf.

 

Musa

I have a question about the [Dragonspine Trophy].

I am surprised by the massive dmg boost with this trinket over my romulo's vial.
According to your spreadsheet, upgrading that trinket would grant me way more dps than the upgrade [Spiteblade]>>[Talon of Azshara].

Furthermore, when i compare the result with the other DPS spreadsheet, the results are pretty different, the boost is then higher for the sword upgrade.

 

• Aldriana

This sheet has not yet been updated for the Dragonspine nerf in the last patch. Prior to 2.2, Dragonspine was indeed a larger upgrade than Talon; in fact, Dragonspine by itself provided an upgrade rivaling all SSC + TK loot put together (relative to their T4/Karazhan/Heroic counterparts).

With the nerf to Dragonspine (which will be reflected in 0.8, whenever I get around to finishing it), it will still be quite a significant upgrade over Romulo's, and, hence, among the more important upgrades for a rogue to get; however, it will no longer be far and away the largest upgrade, and it will, for instance, be a smaller upgrade than Talon.

As for why the other sheet differs: I don't know if they have implemented the Dragonspine nerf yet; if they had, that is probably at least part of the difference. Beyond that, I would refer you the Frequently Asked Questions section of the first post in this thread.

 

Musa

Hehe. I am aware you claim this spreadsheet is more accurate than the other one. I like to confront the two spreadsheets anyway.

I just needed details and I indeed forgot that this wasn-t yet updated to the CD nerf of the [Dragonspine Trophy].
Is the haste nerf as well not incorporated? If not, when can we expect a "2.2 ready" spreadsheet?
I know you are working on a lot of new buffs, so I don't want to hurry things, just curious if you have any date we can hope that for.

edit. I am sorry i just paid attention to your line "whenever I get around to finishing it". I suppose you don't want to fix a date yet.

 

• Vulajin

The haste nerf is already available, go to the "Advanced Settings" sheet and turn on the "PTR Haste Nerf" option.

Aldriana gave numbers for the value of the nerfed DST on a post earlier in this thread. I'm pressed for time so I can't find the link directly, but if you search the thread you should find it relatively quickly.

 

• Aldriana

As Vulajin mentions, haste nerf is in; the Dragonspine Trophy nerf is not.

As for when the next version is ready: hard to say. I was hoping to have it out by now, but I got rather busy with work and various other commitments. Looking down my to-do list of changes, there's still several hours of work that needs to be done - it's simply a matter of when I can find time to bang through this list. Sometime early next week is possible; failing that, hopefully next weekend. But, as always: no promises.

 

royaljester

God I love this conjunction of ideas and theorycrafting...man it makes my eye well up in tears sometimes...

Anyways, I tried out the "ghetto" fix to see how much sharpening stones add, and that was a good way to do it, thanks for the tip. Easy and easily seen results.

Grats on the warglaive set.... /bite

Is anyone currently trying out either the Fist/Sword spec (18/43) with Talon of Phoenix OR the maces spec with Rod of Sun King and Swiftsteel Bludgeon? I was wondering, if about equally geared, how you do in actual raiding environments against sword rogues with ToA/Merc. Just looking at possibly going maces or fist/sword after we see some drops and spreadsheet is showing me a slight drop in dps, but not much to cry about. Just wondering.

 

Myrix

I recently collected the cards to make the [Darkmoon Card: Crusade], but since I have a few days to think it over before the Faire arrives, I figured I'd come get some opinions.

After 2.2, has the value of the Abacus dropped below that of the Card? Shadowpanther lists both Abacus and Hourglass a bit higher than the card, so if I can just get those and actually be better off, then I wont bother using the Deck and just sell it.

 

• Aldriana

For purely sustained fights, Crusade beats Abacus by about 10 AP. However, Crusade takes about 10 seconds to stack up, meaning that for fights that are less than a minute in length - or fights where you suffer interruptions 10 seconds in length at least once a minute, on average - Abacus wins, even neglecting the fact that Abacus is improved by interrupted fights. So if you're doing 25-man raid bosses that are frequently this length or longer, Crusade is probably better than Abacus. Of course, if you're doing 25 man raid bosses, you should have access to trinkets that are a fair amount better than either. But if you're just doing 5-mans or Karazhan, Crusade is probably less worthwhile.

So, all in all: if you have the money to spend, there's certainly a place for it. But it's not a huge upgrade, and, depending on what you're doing, it may not be an upgrade at all.

P.S. You probably want to fill out your WoW Profile at some point, as it makes it easier to assess what your other options might be. Also, the powers that be get grumpy if you don't. Just a friendly tip .

 

uhohzombies

Hi Ald. I did another test for you, this time using SnD (I got lazy halfway through and switched from 3s/5r to 5s/5r because SnD was dropping when I got unlucky with ruthlessness and relentless strikes, hopefully that doesn't affect anything though) and for 28 minutes as opposed to one hour due to time constraints.

Combat log: Free file hosting by Savefile.com



Right before I halted the test, I noticed the proc refresh itself an incredible eight times before expiring. =/

Hopefully this data will be of use to you in figuring out where it's ppm or percentage based. I'll try do the sideways SS only test for you at some point as well, if necessary after this data.

edit: Uploaded the wrong log, the link should be correct now. 8:52pm EST.

second edit:

Free file hosting by Savefile.com

Here is the log for a 20 minute no auto/sinister strike only test. 8 procs in 20 minutes, for what it's worth.

p.s. The logs are a double whammy because I also had Madness of the Betrayer equipped, so you can also parse the data for Forceful Strikes to find out if anything changed with it or not.

 

• Aldriana

I seem to have left the program I wrote to parse these Warglaive data sets at work. Rather than rewrite it from scratch, I'm just going to wait until Monday to grind through those data sets. Thanks for taking them, though.

 

uhohzombies

No problem at all. Thank you for all the time and work you put into the sheet.

 

koaschten

If we would assume 1ppm for [Formula: Enchant Weapon - Executioner] ("occasionally ignore 840 armor") what average benefit might we see from it? Did you have a look at that already Aldrianna?

 

• Aldriana

When it was first announced I ran some numbers on it and concluded that it was probably going to be slightly inferior to mongoose for rogues (making in vastly superior for dps warriors and shaman, one would think). However, the newest revision of the sheet is going to change some things that will affect their relative values, which may close the gap somewhat; when I finish 0.8, I'll revisit this question.

Also, I have another mechanics question for you fine people: does anyone know how Improved Hunter's Mark works? The spell description on thottbot reads:

"Places the Hunter's Mark on the target, increasing the ranged attack power of all attackers against that target by 110 and by an additional 11 each time they are struck by a ranged attack, up to a maximum of 440."

The talent reads:

"Causes 100% of your Hunter's Mark ability's base attack power to apply to melee attack power as well"

So my question is: is "base attack power" just the 110? Or do we get the benefit up to 440? And in the latter case, how long does it take to stack up (aka, how frequently do hunters land ranged attacks? 1/second? More than that?)

Finally a question for uhohzombies: I could figure it out myself, but in the interest of saving time: were you wearing 2/5 T6 for that test? i.e., is that 30% SnD, or 35% SnD?

 

• Vulajin

The stacking portion of Hunter's Mark does not apply to melee attack power. Only the 110 AP applies.

 

Alexsiss

Ald yes he was. was hanging out there with my druid for lotp

 

koaschten

Does that imply you will add it on 1ppm base for 0.8 or that you will have a look at it once its available on ptr/live and the procc chance is known?

 

• Aldriana

Once 0.8 is created, I will whip up a ballpark estimate of it's value assuming 1 PPM and post it in this thread. I am not going to add it to the sheet until the proc rate is actually known.

To Vulajin and Alexsiss: thanks for the info.

 

Citrus

Has anyone checked out this Shiv/Envenom theorycraft? So far the results have been pretty impressive.

WoW Forums -> Shiv/Envenom is viable in PvE. (WWS data)

 

Tosa

Surely the 1.4 millon difference in white damge in the 2 reports affected something.

Hold shift for focused movement.

 

Citrus

I assume you're looking at the full reports and not the individual fights. Being that the full reports include trash mobs.

 

• Aldriana

I haven't done much analysis on this myself, but the other spreadsheet thread has some discussion of the topic; the salient points of which are: Shiv has significantly lower damage per energy efficiency than Sinister Strike, and the increased rate of finishers is not generally sufficient to overcome this. At low itemization levels (like, preKara) shiv/envenom does okay, but it scales very poorly such that by the time you're talking late T5 and beyond, it's obviously inferior.

Hence, until someone demonstrates theorycraft that says otherwise, I'm not inclined to spend to much time worrying about it.

 

Citrus

Actually I think it's the opposite. If we look at the WWS data for the Anetheron kill.
White damage was constant in both, 64% of overall damage.
Shiv did 18% (96,000).
Sinister strike did 25% (115,000).
Envenom did 10% and 52,500 damage during the fight.
Rupture did 5% and 24,700 damage during the fight.

Now as you can see, Shiv only did 20,000 less than Sinister Strike. But Envenom did 28k more damage than Rupture.

Looking at the individual damage of Envenom and Rupture. Rupture average ticks for 280, which means on average a full duration Rupture was doing 2240 damage. Envenom non-crits average 2178 and crits average 4386.

I'm going to check this on some other boss fights in BT like Gorefiend and see how it goes also.

 

• Aldriana

Sure, in raw numbers. But lets look at percentages, which are more illustrative:

Shive + Evenom did 28%.
Rupture + SS did 30%

So, while the raw amount of damage might be higher for Shiv/Envenom, the total amount of damage done was higher as well; hence, assuming white damage is relatively constant between the two builds (which should be a safe assumption), the dps contribution of CPGs and finishers is lower with the shiv/envenom build, meaning it's overall dps contribution is lower, hence the overall DPS is lower.

I'd also like to note that I don't consider WWS of fights shorter than half an hour as indicative of anything; there's just to much variation, particularly in a 5 min fight like Anetheron. Samples need to be large enough so as to be statistically significant, or supported by theory, before they're particularly interesting.

Regardless, it's not really germane to this discussion anyway, since this sheet could not easily be adapted to do shiv cycles, so even if I did think it was a competitive cycle (and I don't) I'd make a separate shiv sheet rather than modifying this one.

Edit: as an example of what I mean by variance, it should be noted that, for instance, the SS trial only had a 25% crit rate on SS (versus ~35% crit for autoattacks in both trials), and Envenom had a 50% crit rate in that test. The fact that rupture was ticking for 10% more damage in the Shiv/Evis test makes suspect unequal buffs (or unequal proc uptime) as well. Any results that might be gleaned from this are thus buried in the statistical noise.