Can anyone give me some PLD tips?

[JackFross]

I understand what you're getting at here, but you're looking at a massive trade off for a tiny return. I was using a figure of speech when I was talking about "the math," but if we're going to get into the nitty-gritty of it, then I'll give you some rough napkin math to show you where I'm coming from.

Considering I literally just did, this should be fun. I'll bold your comments and reply beneath since it's easier than a series of quote blocks.

The debuff is 10% for a duration of of 20 seconds, applied immediately upon application. The application of RoH takes 5 seconds.
RoH takes 7 seconds to apply the STR debuff at 354 Skill Speed, because of RoH having an animation delay of ~2s and it being the third GCD of a combo. 2.5+2.5+2=7, not 5.

Going off of your example of a 40s encounter with roughly 1 auto attack every 2 seconds
Mobs in dungeons attack once per 3 seconds, on average. Sometimes the delay is longer.

that accounts for 20 auto attacks from a single mob over the course of the encounter
By actual calculations, 40s is only 13 auto attacks from that single mob - a drastically different figure than the one that you're using.


With a standard pull rotation of 1 Flash and 1 CoS, you're looking at a 2 second delay on your initial RoH, pushing it to 7 seconds. That accounts for 3 auto attacks that are un-mitigated on a single mob, followed by 17 auto-attacks that are mitigated. On the second mob, at an additional 5 seconds (12 in total), that works out to 15 mitigated attacks from the second mob. The third mob has the lowest return, at 17 -19 (if you throw in an extra Flash, for example) seconds. That works out to 11 mitigated auto attacks. Spread across all 3 targets, that's 42 mitigated physical attacks (again, not including direct attacks) out of a total of 60.
(It's actually just 39, not 60, and that's maximum, not expected. Stuns, movement, slow, etc cut down on how many autos you see.)

(which can vary dramatically depending on the size of the pull and the type of enemies within it).
Working with wrong numbers.
Let me reorganize this for you.
1 Flash -> RoH = 7+2.5 = 9.5s = 3 missed AA's, so you mitigate 10% of the following 6 AA's of this mob, since it only lasts 20s - not 30.5.
RoH on the second mob = 10+7 = 17s = 20s (3 left over) of RoH debuff = 6 more AA's with 10% down.
RoH on the third mob = 17.5 + 7 = 24.5 = 15.5s of RoH debuff = 5 AA's.
RoH on a fourth* mob = 25 + 7 = 32 = 8s of RoH debuff = 2 AA's.
And then the battle ends.

In the course of this battle, you mitigated a total of 6+6+5+2 = 19 AA's dealing 90% of their maximum, which amounts to a total of 1.9 AA's mitigated.
If you instead opened with 2-3 Flashes and used GB combos, the encounter would have ended 3s faster, netting a total of 3 AA's mitigated, a gain of nearly 50% mitigation.

The problem is that you're looking at quantity of mitigated attacks, rather than the quality with which mitigation is happening.

Unless the damage caused by those 3 attacks exceeds the total mitigated amount of the immediate effect caused by RoH upon application
Which it does, because 1.8 attacks is how much your RoH method mitigates, whereas 3 (a larger number than 1.8 by more than 50%) is how many AA's are mitigated by using GB instead and killing the mobs 3s or more faster.

then the benefit to increasing your dps output is clearly outweighed by the thousands of hit points you saved just by applying RoH
I addressed this point directly in my previous post. Thanks for completely ignoring me and arguing against points that I made myself with incorrect math.

I could get into further detail with GB and RA, but that's more napkin math than I'm willing to commit to, and it amounts to no more than peanuts in the bucket; so, to each their own with that one.
And I will reiterate, again, that the benefits of ending an encounter 3-4s faster by improving your dps by 300-400 points vastly outweigh the comparatively tiny amount of mitigation you get from spreading RoH debuff. I hate repeating myself and this is the fifth time I've had to make this point. You mitigate 10% of the damage dealt by the incoming physical attacks for its 20s duration, which is 6 autos. This amounts to 10% x 6 = 60% of 1 Auto Attack mitigated for every one cast of Rage of Halone - a combo which is dealing - quite literally - 60% as much damage as a Goring Blade in its place (which, again, I've shown will always in every encounter reach its full duration on the first mob you use it on, so long as your melee doesn't burn it down).

I think you can understand how someone would think that a statement like this is wildly misleading and flat out false, because it is.
No, you're still misunderstanding.
There are two different circumstances.

1. I intend to remain in Shield Oath for an extended period of time.
2. I intend to turn Sword Oath on as soon as possible.

In situation 1, my opener is:
FB>RB>[Shel+FoF]>GB>[CoS+SW]>FB>[Swipe]>SB>RA>FB>SB>RA>FB>RB>GB>FB>SB>[CoS]>RA>[FoF falls off].
And yes. I have used this opener to hold solid against evenly-geared DPS bursting 2-2.5k on Sephirot EX with and without a NIN giving me Shadewalker. It is safer to use RoH in the opener when undergeared, but even geared or better you shouldn't need it.
In general. Exceptions exist for situations where I want the Strength Down debuff, such as my opener on Ratfinx, which looks more like:
5s > [Shel>FoF]>FB>[CoS+Swipe]>RB>[SW+Ramp]>GB>[Awareness]>FB>SB>RoH>FB>RB>GB>[Concussion].
I use RoH here for the mitigation on the ensuing tank buster - not for a hate lead.

In situation 2, my opener is as I stated previously.

As for Dkr's, that's a bit trickier. I don't remember the exact values that run behind their opener, but if I remember correctly a fully buffed C&S opener generates more enmity than a bonus applied to a Power Slash combo. They get a lot of additional hate from their oGCD's, so it's not abnormal to be able to maintain hate without the use of Power Slash while in Grit. Someone did a full breakdown of how it worked months ago, but I honestly don't remember which thread it was in ...
I was looking into similar stuff a while ago. And yeah - if you're staying in Grit, you pull with a single Power Slash (since the ramp from DA+SE and DA+CaS isn't enough to hold against most openers safely enough for my taste) and then never touch it again for the duration of the encounter. If Grit is on the floor, you'll need Power Slashes to maintain hate.

The point I was making is that DRK has 0 Enmity generation in its optimal dps rotation. None of your off-gcd attacks even have added enmity to them. Everything is just benefiting from your Grit bonus. And even they can hold solid hate without needing hate combos so long as they stay in tank stance. Paladin, on the other hand, naturally has enmity built into its main DPS combo in the form of Savage Blade and is dropping a 3x enmity modded Circle of Scorn every 25s alongside a 30s-at-worst Shield Swipe for an additional burst. This couples with the numbers that show PLD can out-damage DRK in a tank stance v tank stance battle to say that it's ludicrous to assume you'd need a Rage of Halone to hold hate.


*Fourth Mob - When cycling RoH combos, you can RoH up to 3 mobs simultaneously. By the time your fourth RoH applies, your first has fallen off (just barely). This means that your "fourth mob" in this scenario could be the first mob and the calculations I gave would not change.

So, with all of this said, I don't disagree that cycling RoH is a fine way to handle things for a new player who's learning how PLD works. I just resent the implication that you would tout this method as being superior - or even comparable - to methods which can be shown to be better with hard mathematics in both dps and effective mitigation. The argument about healer DPS is absolutely ludicrous either way because, again, your RoH is really, honestly, not doing jack shit to how much damage you're taking from the mobs, as I've explained at length.

And.

Again.

I am not saying that it doesn't do anything. I'm saying it is negligible how much mitigation it gives you and the aggro generation is unnecessary, therefore you're just wasting your time doing it! I can tell you with 100% certainty that you using RoH is not going to affect how much I can comfortably DPS on SCH by any margin /at all/.

You know why that is? Because I can never cast a single heal spell in any 3-mob pull as Scholar and never need to do more than drop Regen on my tank as a WHM.

And that's the bottom line. You're doing things with good intentions and that's cool. The issue is that the effects are completely negligible to everyone who you want to think it might matter to.

[Februs]

Considering I literally just did, this should be fun. I'll bold your comments and reply beneath since it's easier than a series of quote blocks.

Now we're getting into some fun stuff lol ... Though, I'm spying some interesting discrepancies ... For the sake of space I'm following your format.

"RoH takes 7 seconds to apply the STR debuff at 354 Skill Speed, because of RoH having an animation delay of ~2s and it being the third GCD of a combo. 2.5+2.5+2=7, not 5."

Firstly, you're tacking on a lot of extra seconds here. It's true that RoH is delayed to the end of the ridiculously long attack animation, but that animation is not a full 2 seconds. You can time it quite easily, even if you're just using your GCD timer. At most you're looking at 1 additional second (unless you have some kind of wonky latency). Even with a Flash tacked on at the beginning, that's a 7.5-8 second combo. Further, that time is assuaged by the fact that the RoH method of mob tanking requires frequent target swaps, so an extra second on the end of the combo isn't breaking any deals here.

Secondly ... Why in the flying hell is your Skill Speed so low?? You're hardly even at half of what my Skill Speed is. I was going to compare our gear to figure out what was going wrong for you here, but when I clicked on your name I was notified that you don't even have a Pld at 60 ... O_o? I'm gonna give you the benefit of the doubt here and assume that you either have a max Pld on a different character, or something, but that still doesn't explain your low Skill Speed value ... Did you maybe confuse the values for Skill Speed and Spell Speed? Because I think my Spell Speed might be close to that number...

Mobs in dungeons attack once per 3 seconds, on average. Sometimes the delay is longer.

True ... ish.

This conversation has led to me paying attention to some very interesting things over my last few Extreme Roulette's. As you said, mobs attack with a frequency of one auto attack every three seconds, on average, and they seem to do so following the same timer as DoT's/Hot's (though, I still need to do some testing to see if it's exact). That seems to be the given rule of frequency. There are only a few exceptions that I've found so far.

The first is that include direct attacks such as GCD hits (which some mobs actually do have, I was surprised to notice), charge attacks, cleaves, etc, do not operate on the 3 second rule. So, the rate at which a mob hits you for a physical hit can, in fact, exceed 3 seconds. Same as with our attacks, mob attacks that actually have a noticeable charge bar interrupt their auto attack pattern; though, they do share the magic pixel effect where at the very beginning and very end of the charge their auto attack can resume if timed on the DoT. It can throw them a bit out of whack with the rest of the mobs.

The second exception I've found to the 3 second rule is magic based enemies. Some of them seem to have no auto attack at all, and the ones that do hit for much less dmg (those dead conjurers in Lost City Ex made me laugh). Because they also have much more frequent cast bars than their physical counter-parts, their auto-attack timers can get frequently interrupted.


It's actually just 39, not 60, and that's maximum, not expected. Stuns, movement, slow, etc cut down on how many autos you see.

That ... doesn't quite make sense, especially not as a "maximum."

Again, this varies wildly depending on the party and the size of the mob, but with what I would call "good" dps, I've done an ex roul run in which the first pull lasted just a little under 40s. In that time, with 4 mobs (all physical) on the same rate of attack I suffered exactly 44 auto attacks (and yes, I actually recorded it and counted, because why not?). This number actually does make sense, because the timing works out. Mathematically, it looks like this: 40s at 3 auto-attacks a second is roughly 13 auto attacks by one target. Across 4 targets, that's 53 auto-attacks. There is a discrepancy between 44 AAs and 53 AAs; however, that discrepancy is accounted for by the mobs dying at different rates. It's not like they all dropped dead at the exact same time, after all.

I repeated this test a few times now, and it works out similarly each time, varying slightly with the speed of the dps. With what I would say was "mediocre" or "slow," I suffered 51 auto-attacks in a full minute encounter. So there is some variance here. Whereas, 39 could definitely happen with exceptional dps, it is definitely not the "maximum."

Also, stuns definitely reduce that number; however, you should keep in mind that stuns can also be a double edged sword. If the Pld uses a stun (god forbid they exercise some utility, right?), they are also reducing their outgoing dps, as well as canceling any active combos. The funny thing is, a Pld's first stun effectively does exactly what you're recommending with speeding up the encounter. It eliminates 1 full attack from the mob's rotation, as well as delays their next attack until the next tick.

Working with wrong numbers.

It would seem that you are also working with the wrong numbers here. I've already pointed out that the animation time for RoH is not a full GCD (it's roughly 1 second), so tacking on an extra 2 seconds is inaccurate. You also seem to have a much slower GCD rate than I do ... Honestly, I'm a little baffled at how you could be that much slower than I am, because I shaved off several seconds from your estimations when I actually ran it on my Pld. That said, a half a second here and a full second here isn't what makes your numbers off. The problem is the estimated number of auto-attacks, as well as the rate of death spread across the mobs.

In some of the test runs I did (namely, the one's with exceptional dps), you're not wrong. Moving faster provided a little better mitigation, than applying the RoH Debuff. However, in some cases the additional speed didn't even come close to what RoH could mitigate, and what effected this value more than anything was the behavior of the dps team. Their overall speed and the spread their dps made a huge difference, whereas mine made only the smallest of adjustments and not always for the better. Some dps burned so quickly that there was absolutely no doubt that GB's potency was being rendered inefficient. Some were so slow that that I had time to apply multiple GB's, which meant RoH would have mitigated far more. Some Dps spread their Dps out (ranged, like Smn's are guilty of this), making for a more even death rate, causing a higher tick on both GB and RoH.

Another thing that made a substantial difference was the dodge rate caused by adequately spreading out Blinds using Flash. A speed burn rotation requires several Flashes at the beginning (2-3, as you said). Blind is resisted after 3. Therefore, throwing out all 3 blinds right off the bat reduce the total number of time enemies are blinded in a fight, which reduces your dodge rate. Applying Blind in between RoH, however, spreads the blinds out to their full duration, increasing the overall dodge rate throughout the fight. I was surprised at how much of a difference that made over the total dmg taken. Another thing it made a difference to was the application of GB. Using 3 Flash at the start provides for a much slower application time of GB than RoH, so the one circumstance in which you pointed out GB being less effective ("so long as the melee doesn't burn it down") was not so uncommon. Some of that can be avoided by being choosy on which target actually gets the GB, though, so that one didn't provide a big difference unless I made a mistake on application.

No, you're still misunderstanding. There are two different circumstances.

No, I'm not misunderstanding. I get what you're saying now, but what you're saying now is not what you said when you first posted. You wording was very explicit. There is a big difference between "I never ever" use something, to "here's two circumstances, wherein one of them you are actually using it." It's not like I'm psychic lol. Not a big deal though. All cleared up, and I don't disagree with either rotation per se. It just depends on the dps. Some of the Dps I've run with would most certainly cause issues with situation 1, but if it works for your dps team then all the power to you.

The point I was making is that DRK has 0 Enmity generation in its optimal dps rotation. None of your off-gcd attacks even have added enmity to them.

You actually pointed out the issue right here. It's not that your other Drk abilities don't have or generate enmity. There is no physical attack in the entire game that generates 0 enmity. They just don't have an enmity bonus. It's good to keep in mind that every attack has flat enmity multiplier working in the background, even if the combo itself doesn't generate a bonus. So long as the attack is generating high dps output, you can maintain a lead on hate. The "problem" is that no matter how high a Drk's dps is, they don't have the potency and speed to out dps a true dps character; so, a lead is still necessary and you have to be very careful to make sure you're bursting properly .... which is a royal pain on Drk, because they have a billion frigin' oGCD's to manage.

That said, Drk should have no problems when compared to Pld, even with their oGCD enmity boosts. Shield Swipe's dmg potency was nerfed pretty hard, so the amount of enmity generated is actually quite small. The same goes for CoS. Even with the multiplier, the dps output of the attack is so tiny that it doesn't compare to a buffed Carve and Spit. These moves get an enmity boost, but their baseline enmity is so low that it levels them out. It's similar to how Flash falls a bit behind to overpower. It has a similar enmity modifier, sure, but without the baseline damage tacked onto it, it doesn't hold a candle to the amount of enmity generated by War's or Drk's.

In the end, I never intended to imply that one method was superior to the other. I was simply pointing out that their is a trade-off made between the two. Your original post made a few vague claims (and errors) that made it sound as though pushing for that extra dps was a landslide better, when, in fact, its not that black and white. Quite frankly, neither of them are the better method. In all of the tests I did, the amount of dmg mitigated was equal, or near equal, varying slightly on the circumstances of the fight. In some cases, extra Dps was definitely the way to go. In others, the extra mitigation of combined debuffs had far more mileage. Clear times weren't any different between the two either, so there wasn't a clear benefit in time efficiency. Honestly it really falls down to the party composition, dps competency, and what the person is comfortable with.

[JackFross]

Now we're getting into some fun stuff lol ... Though, I'm spying some interesting discrepancies ... For the sake of space I'm following your format.

Lots of good points, but most of your argument is ignoring something really important and all gets blown out of the water by a single thing. Not gonna argue re: RoH, since I was just assuming it's as long as Chaos Thrust - both seem to take forever to apply the thing. It's not about animation lock so much as it is about when it actually lands.

Secondly ... Why in the flying hell is your Skill Speed so low??
Because you were using a 2.5 GCD in your calculations, so I did, as well? I just clarified that I was using base sks as my point of reference. I never said it's how much sks I have on my Paladin (seen here). I have both characters linked to this account... Not sure if it's possible to see both in my profile or whatever.

This conversation has led to me paying attention to some very interesting things over my last few Extreme Roulette's.
Never looked to deeply into it all. That's pretty neat stuff!

That ... doesn't quite make sense, especially not as a "maximum."
And here's where you just go completely off the rails arguing against something I never said. We were specifically talking about a 40s encounter against 3 mobs when I mentioned 39 autos. 13 per mob at 3 mobs is 39. You were calculating (incorrectly) 20 per mob at 3 mobs for 60 autos. I never once extrapolated beyond this to more enemies. So you then go on to say how wrong I am when you apply a comment about 3 mobs to an enemy group of 4 or larger. What is that intended to prove? That the comment I made is still completely correct?

All of your discussion here backs up the fact that [time]x[mobs]/3 = Maximum amount of auto-attacks you'll take in any given encounter. None of your findings dispute this. So all you did is corroborate my point that 39 is the maximum amount of autos you'll see in a 3-mob encounter that lasts 40s.

Like - this discredits quite literally half of your post, from you simply misunderstanding and misrepresenting the very accurate point I made regarding how many hits you'll take in an encounter.

It would seem that you are also working with the wrong numbers here.
I'll just start with RoH not being "roughly 1 second" anymore than it is "roughly 2 seconds" since it's more like ~1.5-1.6s with my 2.38 gcd. Rounding down to 1 is arguably a worse estimation than me rounding it up to 2. Neither is good, though, is the bottom line here, so let's not play the "I'm right, you're wrong" card. Neither of us is right on this count in our calculations.

And I will repeat, again, that Goring Blade is stronger than Rage of Halone the instant the Goring Blade attack does damage. It doesn't need a single DoT tick to be more damage. If you can then hit another Riot Blade before that Goring Blade mob dies, you've gotten more potency than you would have from a Royal Authority. We're currently (at a 2.4s gcd) sitting 12-17s into the encounter. Spoiler alert! That mob isn't dead yet. Let me say this again and hopefully get it through to you this time.

Goring Blade does not need 24 seconds of uptime in order to be your strongest attack by a wide margin on any given mob.

Another thing that made a substantial difference was the dodge rate caused by adequately spreading out Blinds using Flash.
This is fair, but it's worthwhile to note that the blind effect seen here can be done using Goring Blade as well. You can Flash once to get them to stick and then Flash again after the first Goring Blade combo, like you're doing with Rage of Halone. This may be shocking to you... but if you have 3 mobs, Rage of Halone doesn't do much more to hold all three of them to you than Goring Blade does. One of them with RoH still ate a 0 bonus enmity hit, and you can still somehow get to Flash without losing hate.

When coupled with Sheltron>Shield Swipe on the Fast Blade target and Circle of Scorn, stuff is stuck to you pretty well (for ~10s) with just the 1 leading Flash and a Goring Blade combo. You have until the BLM drops Flare or the SMN drops Deathflare to make sure things are on you solid - before then, the enmity generated by DPS is low. At sweet levels of Spell Speed, Fire III is still ~2.7s cast, and each of the following 3 Fire II's is also ~2.7s to cast, and then the Flare takes upwards of 3, putting us easily above 10s before the hammer drops, which is more than enough time for flash > combo > flash. And Dreadwyrm Trance requires 3 stacks of Aetherflow to be spent and then a solid 10s of dot ticks before Deathflare drops. Plenty of time.

No, I'm not misunderstanding.
... It's exactly the same thing. I took the exact wording I used and broke it down for you, because you misinterpreted the clear context I had given in my first post. Go back and read it.

You actually pointed out the issue right here.
Whoooosh. That's the sound of my point flying over your head. Let me break this down in steps for you.

In Grit:
Dark Knight uses no moves with additional enmity mods and can maintain solid hate throughout an encounter without ever needing to touch Power Slash if they remain in Grit.

In Shield Oath:
Paladin spends more than 20% of their GCDs on a skill which generates additional enmity (Savage Blade) while also pressing two off-gcd attacks on 25s and at-worst-30s intervals, each of which both add to overall dps output and generate additional enmity.

The argument:
A Paladin in Shield Oath should never - under any circumstances outside of tank swaps with a Warrior - need to use Rage of Halone for purposes of hate generation. They're packing so much passive threat in their main DPS combo that it's ludicrous for you to make a statement like you did:

You wrote, and I'm quoting here "with Shield Oath active, you should never ever need to hit RoH to maintain hate." Which, at first glance, indicates that you never use RoH at all, even in your opener (hence the words "never ever"). I think you can understand how someone would think that a statement like this is wildly misleading and flat out false, because it is.

Actual, hard comparison of enmity/second of off-gcd attacks from PLD and DRK:
DRK:
Low Blow: 100/25 = 4
Dark Passenger: 150/30 = 5
Plunge: 200/30 = 6.7
Carve and Spit: 450/60 = 7.5
Salted Earth: 75x7/45 = 11.7

Total: 34.8 at worst

PLD:
CoS: 450/25 = 18
Shield Swipe: 525/30 = 17.5
Shield Swipe: 525/15 = 35

Total: 35.5 at worst

Now, in fairness, I ignored Reprisal, which WILL proc, but is a +3.5 enmity/second gain, at best. And that isn't even beginning to consider the fact that Paladin's enmity per gcd is VASTLY superior to Dark Knight's. Grit and Shield Oath provide the same passive boost (x2.7) so let's look at rotational enmity, now.

DRK:
Scourge>DA+SE>Delirium>DA+SE>Delirium
[ 500+(150+250+400)x2+(150+250+280)x2 ] / 13 = 266.15 enmity potency per gcd.

PLD:
GB>RA>RA
[ 150+230+590+(150+200x3.5+350)x2 ] / 9 = 374.44 enmity potency per gcd.

That extra 3.5 from Reprisal isn't gonna matter when comparing these.

This is not a vague claim - it's a fact of how much enmity the two Jobs generate when using their maximum DPS combos in tank stance. DRK holds solid without using Power Slash when in Grit with a drastically lower output of enmity, so it stands to reason that Paladin can do the same, considering the massive amount of additional hate that they can shit out.

I was simply pointing out that their is a trade-off made between the two.
The major flaw with the testing you did was that you were testing with DF, not with a solid group doing the same thing from pull to pull. The DPS of the party fluctuates each time you hop back into the dungeon, causing there to be no constant in your testing. A test without a baseline - a test with no Control Group - is not a scientific test that should be taken seriously. You're not testing anything other than the gross disparity between different party comps and different player skill levels. Saying "I didn't notice any notable difference in encounter length when I changed my rotation" doesn't imply anything when you're changing any of the following variables between those two tests:

1. Which Mobs are being fought.
2. How many Mobs are being fought.
3. The Jobs of the DPS players in the instance.
4. The players who are playing the DPS players.
5. The person playing the healer, and what Job they are.
6. What cooldowns you use to mitigate damage in that pull.
7. What buffs you use for offensive potency in that pull, like Fight or Flight.

And a small sample size in this game is incredibly misleading due to the drastic variance that you can get simply from a DPS inflicting critical hits with AoE attacks. You'd need dozens of pulls to replicate to the point that you would see the calculable differences between these two strategies in terms of encounter time.

At the end of the day, it's the same argument people give for using Potions to increase DPS output - it's another buff that does absolutely increase how much damage you do. The problem is that variance in damage and critical hit rate can make you do more damage without a Potion than you do on the next pull with one. Potions don't increase your flat damage output on a given pull - they increase the average damage output that you'll do over all pulls.

Same exact thing applies here. Sure, you didn't notice a difference when swapping with variables all over the place, but there's no way of knowing whether or not using Goring Blade combos actually did NOT accelerate the pull. The numbers say, flat out, that it did. It's not really a thing that's up for debate. I will argue forever that the "benefit" of strength debuffs on a single target in a multi-target encounter is negligible in comparison to how much damage you leave on the table when you do such a combo. As a point of comparison, my Paladin currently:

Shield Oath on a dummy:
FB>SB>RoH:
580 + 780 + 1000 = 2360 damage

FB>RB>GB:
580 + 930 + 970 = 2480
With 200 damage DoT ticks.

The amount of damage you sacrifice by using RoH over GB is massive and only increases the longer the mobs stay alive.


I won't argue your last set of points re: DPS composition and comfort level, since that's completely fair, but I still disagree that a top-skilled player who is using Paladin in dungeons should ever be hitting RoH on trash pulls, unless they messed up their rotation and lost hate on something. That's all I'm arguing here. It's not a judgment call - it's not something that is just as beneficial either way. One method is clearly superior. I have yet to see any discussion from you or anyone else which has remotely convinced me otherwise.



PS. On the topic of Shield Swipe - it was absolutely a buff to Paladin, but a nerf to the skill. It increases PLD's overall damage output, while decreasing the utility of the skill in comparison to how it used to be. It was a DPS loss post-3.0 to use Shield Swipe more than once between Goring Blade applications - and even that once was pushing it.

[Februs]

Lots of good points, but...

Fair enough. I think we can put the issue to rest now. It's been a pretty solid discussion. Just for the sake of some follow up:

(seen here). I have both characters linked to this account... Not sure if it's possible to see both in my profile or whatever.

That makes WAAAAAY more sense. I was legitimately baffled at that one.

I originally used 2.5 seconds as my starter because I typically don't like to assume that everyone has their skp up to snuff. They might primary a different tank, after all. I had to adjust my value when I started doing the actual field tests, but when you posted one that was almost half of mine I assumed it was yours and got very confused.

Never looked to deeply into it all. That's pretty neat stuff!

Right? It actually made me interested in doing a full breakdown of how it works. Unfortunately, that would take quite a bit of time, and I don't think I could accumulate all of the data without resorting to .... lets just say "frowned upon" methods.

And here's where you just go completely off the rails arguing against something I never said. We were specifically talking about a 40s encounter against 3 mobs when I mentioned 39 autos.

You misunderstanding a little here. The reason I deviated from your 40s/3 mob example is because the examples I was using above were not calculated theory. They were actual pulls. I did a set of Ex roul runs, recorded the footage, slowed it down, and manually counted each auto attack as they hit, the timing they hit with, and what number of them would have been mitigated based on the application of RoH. Then I tallied the total number of dmg. Sadly, every damn Ex roul I got during this period was Lost City Ex (I swear the game goes in spurts. Last week I only got Anti-tower >.> ). The first pull of that dungeon is 4 (including the rhino beetle), so I went with that.

It was a relatively small sample. I only included about 6 runs split evenly (3 RoH and 3 GB/RA rotations) in my actual sample. In the end, though, the results for total dmg sustained were not vastly different regardless of which method I used. The values were near enough to equal that it just didn't matter, so long as I held enmity. What I did find, though, was that the behavior of the Dps made a much larger difference. This also ties into your next point about Goring Blade.

Goring Blade is stronger than Rage of Halone the instant the Goring Blade attack does damage.

Firstly, I never intended to imply that Goring Blade wasn't stronger than RoH in potency. That's not even a question. The question was the total value of Dps being drawn from the move over Royal Authority, assuming that the DoT gets clipped by the Mob's untimely death. In most cases, you are correct. In my test runs I found that GB was the preferable choice; however, in the cases of extreme dps it may not be. It just depends on how narrow your enmity lead is, and if the additional Flash required to keep it would count as a loss to the overall damage provided.

This is fair, but it's worthwhile to note that the blind effect seen here can be done using Goring Blade as well.

Again, though. we're talking additional flashes here and with a lot of variance on Dps. With the average Dps team found in DF, sure. You can spread those flashes out and chances are your enmity will hold pretty firmly. With exceptional Dps? Different story. Without establishing a lead on hate, you run the risk of losing it. To get a lead on hate using a Dps rotation, you have to burn your own dodge rate by clipping the effect of blind, casting it early. It's also worth noting that another contributing factor was that the mitigation of RoH stacks with your other CD's and passive defenses. It's not just working by itself. If you parry or block, for example, you're basically reducing 20% of your incoming dmg, and the leftover damage is nerfed by the 10% provided by RoH. It all added up.

Whoooosh. That's the sound of my point flying over your head.

Not at all. I get what you were saying on single target dps, and I don't dispute it. I use a similar method when boss tanking, and, in most cases, I don't have any trouble maintaining hate on bosses when using a Dps rotation.

It was mob tanking, where I was questioning if there wasn't a difference to be found, and I was merely pointing out that the actual damage dealt does play a role. A lot of people seem to ignore the fact that the damage output of a character can vary their enmity. This is one of the reasons why War's could strip enmity off of Pld's and Drk's while in the OT position before the latest set of patches. Fell Cleave bursts don't have an enmity bonus, but they did out-pace the baseline enmity provided by the standard rotations of Pld's and Drk's. The same can be said when comparing Drk and Pld. Over a sustained period of time, one clearly does more dps than the other. Pld gets more enmity boosts per turns of the GCD, but their overall lower dps evens things out. I don't think this is a big secret, either. It's pretty well common knowledge that Pld's walk an enmity tight-wire, and it's one of the reasons why all 3 tanks have to stay as geared as possible when running with competent or exceptional dps.

The major flaw with the testing you did was that you were testing with DF

This was actually the point. Real World Data. If you stack the deck, you get stacked results. Given that most Ex Roul runs done by most tanks are done in DF, stacked results weren't the most favourable outcome. Instead, what I was looking for was an average. I could have grabbed 3 of my friends and done these test, but that would not have been an accurate reflection on the common experience in DF. As it happens, the first 2 out of your listed controls were maintained, because, for some strange reason, DF was broken and only ever put me into Lost City ... which was annoying, buy ir kept the pull sizes and mob types identical across my runs. My buffs and defenses were also near identical (as close as they could be), because I was always tanking and I ran the dungeon the same way each time. I was completely constant in each set of runs (RoH runs vs Dps runs). The only thing that varied was the dps and healer, and I stand by that those two things should vary, because the majority of people do not run with a fixed party. The fact is that DF does, on a day to day basis, see a dramatic variance in the type of Dps provided during a run. The only way to account for that random element is to attempt an average across it, not skew the data by fixing it to a single point.

The only thing that I would say was a downside in these tests was that the sample size was, admittedly, small. 6 runs is not enough to establish a reliable average; though, it was more than enough to get an idea of what to expect, should a larger sample be pooled.

In the end, like I said, it comes down to comfort and party composition. Given how utterly minimal the difference is between the two methods, I can't say that I tilt in favour of one way or the other. I, myself, vary my own method accordingly, and I can't say that I stick to one religiously. It just depends on what you think is needed for the random circumstances that DF throws you into ... unless, of course, you are running a pre-made. In which case, all bets are off, since that's a totally different animal altogether.

[JackFross]

Fair enough. I think we can put the issue to rest now. It's been a pretty solid discussion. Just for the sake of some follow up:

I just have a few like single-sentence remarks to add... And one doozy re: your scientific process.

You misunderstanding a little here. The reason I deviated from your 40s/3 mob example is because the examples I was using above were not calculated theory. They were actual pulls.
Yeah, I get that, but then you were comparing real world data with different scenarios to calculated data with a specific scenario. The 39 hits was specifically the max you'd see in a 40s 3 mob pull. Rightly so, with +1 mob there (for 4 total), the max would be 52 - which wasn't exceeded by your 4-mob trials. I was just upset/offended by the way you worded things, making it seem like my calculations were incorrect, when your findings corroborated those results.

You can spread those flashes out and chances are your enmity will hold pretty firmly. With exceptional Dps? Different story.
Definitely fair to some extent, but missing the main point. Even in the RoH scenario, one mob eats just a Fast Blade. And in the case of more than 3 mobs, you're gonna need an additional Flash either way.

This is one of the reasons why War's could strip enmity off of Pld's and Drk's while in the OT position before the latest set of patches. Fell Cleave bursts don't have an enmity bonus, but they did out-pace the baseline enmity provided by the standard rotations of Pld's and Drk's.
The only time this was (and still is) an issue was when both tanks were out of tank stance. And the reason has very very little to do with Fell Cleave - and much more to do with using Butcher's Block under Berserk. It's not about DPS - it's about using enmity moves when you're not in the Main Tank position.

This was actually the point. Real World Data. If you stack the deck, you get stacked results.
That's not at all what I was implying. The differences afforded by swapping from RoH to GB and back are small in terms of time saved (which, yes, means it's not super amazing wow better) and, as such, you're not gonna be able to notice them when you have one group of DPS doing 2k combined and another where the SMN is pulling 2.5 on his own. That's the issue with testing in DF.

It's how science works. If you want to test 1 variable (how much time is shaved off a fight by using GB instead of RoH?) you need to keep all other variables which affect this one constant between attempts. If you have variable kill times based on DPS being different and doing different things you cannot draw any meaningful conclusions about your impact on the encounter. It's not "stacked results" - it's "viable results which actually mean something in this particular context."

The only way to account for that random element is to attempt an average across it, not skew the data by fixing it to a single point.
You do not need to account for the variance in raid DPS when calculating or figuring out the impact of your DPS on the clear time of an encounter. You need to get a control group of DPS who will do X DPS in Y encounter and then run that encounter Z times using RoH and using GB. Compare the kill times where the only variable is the one you aim to measure and come away with an actual conclusion that means something.

Having variances all over the place in party composition, player skill, and raid dps and then saying "I didn't notice a reasonable difference between pulls where I used GB and where I used RoH" is completely meaningless! Of COURSE you couldn't notice - of COURSE the DPS being better makes a larger difference. That's not the point. It's not the important factor that I was discussing when comparing the two strategies with respect to mitigated damage. You can't get a reasonable approximation of how much additional damage is being mitigated by GB in comparison to RoH without all other variables which affect this difference remaining constant.

Examples of other variables you may not have tracked:
1. Did anyone cast Virus, Disable, Rend Mind, Dismantle on the adds?
2. Did you ever get an AST throwing Bole at you?
3. Did you ever get an AST using Balance, Arrow, etc?
4. Did you have SMN/SCH using Shadowflare?
5. Did a BLM drop Lethargy? (probably not, but still) Maybe Apocatastasis?

And then of course the variance in healer DPS and DPS DPS.


tl;dr: The math says - flat out - Goring Blade combos will amount to more mitigation in the long run by trimming 3+ seconds off encounters simply by using it rather than not. None of your testing remotely disproves this. x_x