Elitist Jerks
Register
Blogs
Forums


Go Back   Elitist Jerks » Class Mechanics » Paladins

Reply
 
LinkBack Thread Tools
Old 10/04/12, 4:38 PM   #61
bromli
Von Kaiser
 
Dwarf Paladin
 
Korgath
Originally Posted by Exemplar View Post
Pre-raid dungeon gear gets you to in the 40k range. 0.2% of 40k is 80 DPS. T14 Heroic gear can push you to around 80k. 0.2% of 80k is 160 DPS. An extra crit or two above average produces the same DPS. One or two less and it eats any potential gain. This is to what I refer when I talk about how easy things can hide in standard RNG variance.

Additionally, there is nothing bad or wrong about hitting 5 HP. While it's entirely possible that using TV at 4 HP may cause a few less conflicts and thus grant a couple empty GCD, it's not guaranteed and it doesn't much matter. Let's say you earn 120 HP over a fight. Whether you use them at 4 HP or 5, you still can only get a total of 40 TV/Inq. 5HP just strongly (strongly!) discourages you from using another HP generator. If you impose a 4HP rule, all you're doing is hitting the wall earlier (an arbitrary wall, at that).

Not sure what other 'many things' we do for such a marginal increase. I personally recommend at least 2-3% DPS increase before anyone tries to apply something non-standard. Even at that scale it can unfortunately be all too easily hidden within RNG variance from fight to fight. Haste vs Str gems is probably even less of a gain (at ratings in OP it's 46 DPS a gem), yet folks obsess. Since folks want that detail, I'll notate the OP with your priority as a possibility.

I'm pretty sure Balhale will agree simulations are not perfect. My spreadsheet and SimCraft are not an actual fight. Just because it looks better on paper does not mean it is actually superior - there's a margin of error even if our math is 100% accurate. 1% on paper could be 3% in a fight... or it could be negative 4%.

P.S. Ignoring cooldowns, execute, and procs is throwing the baby out with the bathwater.
First, more is more. While .2% is certainly within a range that sims could be incorrect, it is also certainly within the range of things we do for a dps edge. Looking at a single gem, using a red gem over an orange/yellow gem has even less of a gain, and also has a cost. (It's just video game money, but many people are willing to inspect someone else's gear and point out the error of that one socketing choice.) I've spent a lot more time following rogue theorycraft over the years, and those guys would cut off a finger to gain 3 dps. There is no cost to choosing which of 2 buttons to hit, at least not one that has been demonstrated, with at least some indication to the contrary. I'm not here to tell everyone they're "wrong", but i do find this priority simple to execute, and still fail to see any downside.

Second, standard doesn't necessarily mean better. It just means that's what is currently accepted. If we aren't looking for gains in every nook and cranny, what are we doing exactly?

Third, i would argue that there is a cost to building to 5 holy power. Namely, what happens when dealing with mechanics.

- If you are at 5 HoPo when off-boss movement begins, all generators cast while moving are wasting HoPo. During epicenter on Feng the Accursed, for example, you can reliably generate 1-2 HoPo every time.

- If you need to use a hand spell, divine shield, etc, being at 5 HoPo will make you much more likely to waste HoPo afterwards. At times, even 3 or 4 generators end up clashing. After having used a utility spell followed by TV5, it's quite typical to have at least 2 off cooldown. While sometimes you can plan around this, often you cannot. Being so GCD locked at times, I think this is easy to see.

- When a boss parries your TV, a similar effect occurs. Worst-case scenario is probably the case where you end up wasting an AoW proc. Let's say the boss turns to cast while you are casting TV5, and Exo would be next, or AoW procs during that GCD. The boss parries you, so you re-cast TV5, and during this GCD AoW procs again. If you had been at 4 HoPo, the response would be to use Exo, then TV5, and you would have Exo up again from AoW.

Boundless Conviction is our leeway for dealing with mechanics, and even at times the proccy nature of our rotation. Having this leeway available is a good thing.

As far as cooldowns, execute, and procs go, I readily admit that my rotation mapping cannot deal with very much RNG. That said:

- Using Sanctified Wrath, you never use judgment. The sequence goes like this:

HoW > CS > HoW > TV > HoW > Exo > HoW > TV > HoW > CS > HoW > TV

If Exo is up, you replace CS with it. If AoW procs with 2+ HoPo, you do:

TV > HoW > CS > (AOW procs here) HoW > Exo > HoW > TV5 > HoW > TV

- Using Holy Avenger, you alternate generator > finisher > generator > finisher, simply using the highest priority generator available. The question of judgment vs. TV4 is irrelevant here.

- Divine Purpose i cannot efficiently deal with in this way, obviously. My argument here would be that the free TV you are getting would have essentially the same effect as using a utility ability. That GCD has to come from somewhere, and the lower your HoPo, the less likely a clash will occur afterwards.

- Art of War: I have actually done some rotation mapping attempting to see what kind of effect AoW has on the 2 priority lists. Very painstaking, as what I did was write out the rotations for each priority, with AoW proccing at different times, and compare the results on a case-by-case basis. This is the kind of thing i doubt anyone wants to dig through. Additionally, it's still primitive in that I never accounted for multiple AoW procs in a short time frame, or DP procs at all. There are cases where the standard priority comes out ahead, but overall it is behind. In every case, the gains are small to completely nonexistent. I'm guessing this is where most of the ~.2% gain is coming from, actually. AoW seems on the surface to be another factor like utility spells, in that it consumes a GCD, making a clash more likely later.

- Execute range: The only reason my suggestion is viable at all is that we have more HoPo generators available than we did in WotLK. Every time you drop judgment behind something else in priority, there is a chance that you will be able to fill every GCD with something better. I can't see any reason why adding another, superior generator to the list would do anything but make judgment even less attractive, as the chance you can fill every GCD with a non-judgment generator only goes up.

It is not my intent to step on any toes. I greatly appreciate the effort you guys have put into the sims, and this thread. It's just that I saw an opportunity for a gain, one whose quality of life implications may well surpass the minor patchwerk gains.

United States Offline
Reply With Quote
Old 10/04/12, 7:47 PM   #62
averykey
Glass Joe
 
averykey's Avatar
 
Tauren Paladin
 
Magtheridon
So is 4hptv > J > 3hptv a dps increase?

I remember seeing a few well thought out post in the last thread, and I see he just re-posted in this thread.

United States Offline
Reply With Quote
Old 10/05/12, 1:27 AM   #63
wellsdaddy
Glass Joe
 
Human Paladin
 
Terenas
Read though the OP and the few new pages here and hadnt seen any feed back on the Elemental Force enchant so I figured I would post my personal results. I'm not sure how thoroughly its been compared/tested (as I myself just ignored it initially and opted for Windsong) but my results through SimCraft actually favor Elemental Force in pre-raid gear. Fully raid buffed and at ilvls from 460-465 so far, its coming out ~ 400 dps higher. I realize sims dont tell the entire tale of the tape, but we are using them to give us relative points of comparisons. Again, this was using the latest version of SimC, I havent dabbled into Exemplars spreadsheet just yet. I consider myself relatively proficient with the sims/spreadsheets by now but I'd like to see if others could support/debunk my personal findings for the time being. Cheers,

United States Offline
Reply With Quote
Old 10/05/12, 9:00 AM   #64
Pulout
Glass Joe
 
Blood Elf Paladin
 
Zul'Jin
Gara'jal question

Spiritual Innervation is a buff that is stacked based on healing in the spirit realm in that fight. My question is would it be more beneficial to us to just self heal as much as possible in the spirit realm to stack the buff? Also, if that's the case, what glyphs would we use for that encounter? Would battle healer be viable or a complete waste of time (assuming we only use seal of insight in the spirit realm and truth up top)? I'm just trying to figure out the best way to comfortably avoid the enrage.

Offline
Reply With Quote
Old 10/07/12, 7:17 PM   #65
Fierss
Von Kaiser
 
Draenei Paladin
 
Whisperwind
I just picked up Execution Sentence for the first time (I liked the healing of the other two spells), and have noticed some issues with its tooltip and its behavior with haste.

When you cast Inquisition, the tooltip of ES shows a 4.4% damage increase rather than the expected 30% increase, while our other holy damage spells show the proper 30% increase. I spent 2 minutes on a target dummy and believe the spell is actually doing the proper 30% more damage, though I didn't do a thorough check.

Also, I don't think Execution Sentence is factoring in haste properly (read: at all). At 14.03% haste (5963 rating, including Windsong's haste proc), I believe I should be past the first breakpoint for an extra tick, but I get 9 ticks and the explosion at the 10 second mark, same as when I go naked and run 0% haste.

Offline
Reply With Quote
Old 10/07/12, 10:40 PM   #66
Balhale
Piston Honda
 
Tauren Paladin
 
Bladefist
That is correct, Execution Sentence does not benefit from haste and hasn't for awhile and isn't modeled in SimC as doing so.

None of the 90 talents benefit from haste at all.

United States Offline
Reply With Quote
Old 10/07/12, 11:49 PM   #67
Ronark
Don Flamenco
 
Ronark's Avatar
 
Blood Elf Paladin
 
Hyjal
Originally Posted by Balhale View Post
That is correct, Execution Sentence does not benefit from haste and hasn't for awhile and isn't modeled in SimC as doing so.

None of the 90 talents benefit from haste at all.
This was intentional according to a Blue post response to myself asking why nothing scaled with Holy's Mastery, Haste, and why the healing from LH/ES/HP scaled with inquisition.

United States Offline
Reply With Quote
Old 10/08/12, 7:36 AM   #68
aylen86
Piston Honda
 
aylen86's Avatar
 
Human Paladin
 
Malygos (EU)
Originally Posted by wellsdaddy View Post
Read though the OP and the few new pages here and hadnt seen any feed back on the Elemental Force enchant so I figured I would post my personal results. I'm not sure how thoroughly its been compared/tested (as I myself just ignored it initially and opted for Windsong) but my results through SimCraft actually favor Elemental Force in pre-raid gear. Fully raid buffed and at ilvls from 460-465 so far, its coming out ~ 400 dps higher. I realize sims dont tell the entire tale of the tape, but we are using them to give us relative points of comparisons. Again, this was using the latest version of SimC, I havent dabbled into Exemplars spreadsheet just yet. I consider myself relatively proficient with the sims/spreadsheets by now but I'd like to see if others could support/debunk my personal findings for the time being. Cheers,
As Windsong is currently bugged (procs occur far too less - confirmed by a blue poster here), Elemental Force seems our enchant of choice at the moment.

Probably a good time to test the Elemental Force enchant and gather some data about its behavior.

Last edited by aylen86 : 10/08/12 at 7:43 AM.

Online
Reply With Quote
Old 10/08/12, 7:52 AM   #69
Fordragon
Von Kaiser
 
Fordragon's Avatar
 
Blood Elf Paladin
 
Stormrage (EU)
Originally Posted by aylen86 View Post
As Windsong is currently bugged (procs occur far too less - confirmed by a blue poster here), Elemental Force seems our enchant of choice at the moment.

Probably a good time to test the Elemental Force enchant and gather some data about its behavior.

From what I see it's only bugged for casters.I still get a lot of proccs as ret.

Greece Offline
Reply With Quote
Old 10/08/12, 8:40 AM   #70
Exemplar
Bald Bull
 
Human Paladin
 
Scarlet Crusade
Regarding Windsong and Elemental Force:
Before we can decide which is better, we need an important piece of data. Namely what is eligible for procs. We know the proc rates per Blizzard directly, but what can cause those procs has not been nailed down.

Example from the past:
In the beginning of Wrath, Avalanche was a crazy good enchant. This is because Ret had a chance to proc on: autoattack, seal proc, Censure (re-)application (thus three chances per auto!) amongst other things.

Originally Posted by Blizzard
Dancing Steel: 1PPM on melee attacks that land.
Elemental Force: 5PPM on melee damage, or non-periodic spell damage/healing, with a 0.1-second cooldown.
Windsong: 1PPM on melee damage, or non-periodic spell damage/healing, with a 1-second cooldown.
Knowing what qualifies as melee attacks (does Judgement count? HoW? Before you blindly throw them away, know that in the past both have sometimes procced mechanics that specifically stated melee attacks) is important. Knowing whether seal procs and Censure application count (just because it says spell damage/healing doesn't automatically mean non-damaging Censure application cannot proc - again, sometimes it has).

If you want to assist but don't know how to determine these things, just run a nice long sequence against training dummies, then post WoL or similar. Both a full-ability-usage log (or multiple) as well as an auto-attack only (with SoT active) would be useful. Short runs may not iron out the RNG and be adequate to calculate proc rates (which we shouldn't require since Blizz told us), but they can be used to find whether or not a specific ability was eligible to proc. If you plan to try, please stick with it for 15+ minutes straight in a log.

Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

Offline
Reply With Quote
Old 10/08/12, 10:14 AM   #71
Exemplar
Bald Bull
 
Human Paladin
 
Scarlet Crusade
Please excuse the double-post. Unrelated to my last post:
I decided to implement the option to add TV4 before your last ability (Judgement, unless someone decides on some weird priority setup) in my spreadsheet. This permits Inq > TV5 > HoW > Exo > CS > TV4 > J > TV3 > SS. Spreadsheet also has some minor tweaks/bugfixes as well as gear updates (location of some items, JP gear at 458, and similar).

Findings:
On a 450 second fight there is zero DPS difference between TV4 and TV5. I ran this both in my present gear as well as a rough Heroic T14 set (rough in that it's not some claim at BiS). The quantity of every attack does not change. The quantity within/without any given CD did not even change (I didn't expect that). You're just juggling around some TV5, TV4, and TV3. No better conflict resolution seems to be happening.

It's entirely possible that using some non-standard spell, such as a Hand, would cause less damage loss via TV4 than it does normally. Since I don't model that, I cannot determine. If you want to add TV4 into your priority, by all means, go for it. It won't hurt anything, but neither does it seem to help.

I ran 3 SimC at default with my gear and took the averaged DPS. I then added a line for "actions+=/templars_verdict,if=holy_power=4" before Judgement and ran 3 SimC with my gear, then took the average.

Patchwerk:
TalentTV4?DPS
SWNo57483
SWYes57431
HANo54324
HAYes54285
DPNo56589
DPYes56564

Hmmm, maybe Patchwerk's a bad example. Movement should exacerbate HP5 issues, right? So let's swap to SimC's HelterSkelter mode.

Helter Skelter:
TalentTV4?DPS
SWNo53145
SWYes53082
HANo50047
HAYes49998
DPNo51840
DPYes51804

Yeah, the HP4 averages were a bit lower in every run - maybe one additional conflict rather than any less. It's a tiny variance and possibly within the usual variance of SimC runs, so I'd say they're effectively the same.

Bottom Line: HP4 is indistinguishable from the standard suggested priority.

Latest spreadsheet: here

Rock: "We're sub-standard DPS. Nerf Paper, Scissors are fine."
Paper: "OMG, WTF, Scissors!"
Scissors: "Rock is OP and Paper are QQers. We need PvP buffs."

Offline
Reply With Quote
Old 10/08/12, 1:48 PM   #72
Grognard
Von Kaiser
 
Dwarf Paladin
 
Sen'jin
Originally Posted by Exemplar View Post
...Knowing what qualifies as melee attacks (does Judgement count? ...
Judgment is a full-fledged melee attack that is used at range, and is capable of proccing "on melee attack" effects; I tested this on dummies after 5.0.4 to see if Judging alone would produce a Gurth tentacle. Stood far enough away so that auto-attacks would not interfere with the test, and successfully popped tentacles using nothing but Judgment - I believe I even did the test without a Seal to ensure that it could not possibly be a further "bug" in Seal mechanics.

Offline
Reply With Quote
Old 10/08/12, 2:42 PM   #73
bromli
Von Kaiser
 
Dwarf Paladin
 
Korgath
Thanks for the attention to this. I've been trying to figure out some ways without spreadsheets to plug some of the holes in my analysis, and it's really hard without sims.

For those who REALLY want to squeeze out tiny bits by micro-managing, there is another factor to consider here. Namely, the remaining cooldown on CS. At any given time, there are 3 possible situations regarding CS. Either it is available, it was just used, or it will be available next.

1) If CS is available, it would be used over judgment anyway, so this case is ignored.

2) If CS was just used, then (ignoring all other abilities) you could do J > TV5 > CS without delaying anything. It does not surprise me at all that choosing TV4 > J > CS would be a loss. On face it seems it would be, as you delayed J without gaining anything. Only procs and Exo/HoW would change this, and while those factors could push it slightly in the direction of TV4 rotation, procs have fairly unpredictable ramifications on specific situations, and I was having trouble setting up the experiment, so to speak. Certainly nothing I've done shows anything conclusive in favor of TV4 here.

3) If CS is about to be available, then you are choosing between J > TV5 > CS (CS is delayed here), and TV4 > CS > J (here, J is delayed twice). This was the original situation i was thinking about, and my analysis seems to indicate that TV4 is a small gain here. However, this is certainly complicating the rotation noticeably, and for gains that are arguable at best.

After the initial feedback that simply tweaking the priorities showed a small gain, I decided this was worth exploring in depth. Since there is now evidence that it is in fact a loss, I'm going to leave it alone. Even if there is a gain in case 3, it is very small, and the losses in case 2 would seem to outweigh the possible gains in terms of a unified priority list. In any case, the idea that there is zero opportunity cost in terms of complexity has gone out the window. Minding the remaining CD on CS is something I don't want to worry about during progression. Perhaps more skilled rets than I would find this to be simple, but at the very least there is a cost.

Additionally, glyph of double jeopardy on target-swap fights would make J slightly more attractive, which i hadn't considered. This pushes standard priority even further ahead, and leaves TV4 rotation in a place where it's dependent on situation, subject to the remaining CS cooldown condition, and with very small gain even under ideal conditions.

Thanks again for showing this some love. Happy facerolling, everyone.

United States Offline
Reply With Quote
Old 10/08/12, 4:03 PM   #74
Balhale
Piston Honda
 
Tauren Paladin
 
Bladefist
Did some Windsong testing. It appears that Seal of Truth and Censure application do not proc it. Autoattacking a 90 dummy for fifteen minutes with only a Mograine's Immaculate Might equipped resulted in 15 procs on 252 melee swings for a proc rate of 0.0595 which matches the expected 0.06 for a 3.6 speed weapon and 1PPM.

I'll test Elemental Force and what attacks can proc it/Windsong later, but I'd expect that it will also not proc on Seal of Truth and Censure.

Last edited by Balhale : 10/08/12 at 4:08 PM.

United States Offline
Reply With Quote
Old 10/08/12, 4:39 PM   #75
Charybdis
Don Flamenco
 
Blood Elf Paladin
 
Area 52
Originally Posted by Grognard View Post
Judgment is a full-fledged melee attack that is used at range, and is capable of proccing "on melee attack" effects; I tested this on dummies after 5.0.4 to see if Judging alone would produce a Gurth tentacle. Stood far enough away so that auto-attacks would not interfere with the test, and successfully popped tentacles using nothing but Judgment - I believe I even did the test without a Seal to ensure that it could not possibly be a further "bug" in Seal mechanics.
As past experience should have shown, different abilities can proc different effects differently. Even the same abilities can proc different effects differently. It is entirely possible that Blizzard has screwed up on some of the coding for different abilities, especially as they undergo revisions. Thus, every enchant needs to be tested with every ability (including some glyphs even) to ensure we know exactly what does and doesn't work.

Last edited by Charybdis : 10/08/12 at 5:07 PM.

Offline
Reply With Quote
Reply

Go Back   Elitist Jerks » Class Mechanics » Paladins

Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
The Retribution Concordance - 4.0.1 Exemplar Paladins 482 12/06/10 3:31 PM