[Prot] Seraphim VS Holy Shield Comparison

[Killingjudge]

MMO Champion Prot discussion forum: http://www.mmo-champion.com/threads/...WoD-Discussion

TL;DR -

Sacred Shield is the most efficient talent in its tier based on both numbers and the fact that it is an absorb shield instead of a heal, so it less likely to go to waste.

When a full rotation is possible and magic damage is not an issue, Seraphim is a great choice for mitigation.

I'm starting to lean towards Divine Purpose over Holy Avenger after some play testing. Free SotR attacks off the GCD (sometimes with upwards of 4 procs in a row) is incredibly complementary to both Seraphim and Holy Shield builds. Holy Avenger is obviously still a fine choice.

Using Seraphim correctly - A Rotation of Mitigation

Here is *a* video on the Seraphim rotation. It covers most of the basics of the rotation - https://www.youtube.com/watch?v=MAPl78fMsZc

Seraphim is a great skill that comes with both a large damage increase as well as large damage reduction effects.


Benefits of Seraphim? 15 of every 30 seconds we get:
1000 Bonus Armor
1000 Crit (Parry)
1000 Mastery (Increased Damage Reduction from SotR)
1000 Versatility


What does this add up to?

This is how the numbers ended up impacting the stats on my character unbuffed with an avg 655 ilvl

Stat	Before	After	Difference
Parry (from crit)	6.60%	15.69%	+9.09%
SotR Damage Reduction (From Mastery)	11%	18%	+7%
Damage Reduced by Armor	48.71%	55.09%	+6.38%
Damage Reduced by Versatility	2.09%	5.94%	+3.85%

The obvious drawback is we have to dump 5 Holy Power into the ability every 15 seconds and for the other 15 seconds we do not have the passive 15% block from Holy Shield to help with mitigation.

To solve this problem, we have to fill in the other 15 seconds with damage reduction greater than or equal to what 15% block would give us. To do this, we will need to utilize well timed Shield of the Righteous casts, Holy Power Management, and Glyph of Divine Protection.

Rotation of Defensive Skills
1: Sacred Shield (30 second buff)
2: Divine Protection (8 seconds of 20% reduced damage)
**Build to 5 holy power during these 8 seconds
3: Seraphim (15 seconds of *see above*)
4: SotR at 3 Holy Power for extra reduction
**should be able to build to 5 holy power before Sera runs off
5: SotR, any Holy Power gain, SotR (The first SotR cast should happen while Sera is still on to make use of the bonus mastery)
Repeat

The basic idea here is we can have an up time of 29 out of every 30 seconds where we are mitigating more than what we would passively

Comparing Numbers

The number comparison issue is that we need to add both the 15% block chance and the increased number of SotR casts when comparing Holy Shield against the Rotation of Mitigation.

To try to shorten this up, we are comparing buffs over a 30 second period of time. Because Divine Protection could be used on cooldown for either Holy Shield or Seraphim, we should not calculate it for either skill.

Seraphim:
15 Seconds of +9.09% parry ~ 4.55% Avoidance
15 Seconds of +7% Block Rating ~ 2.8% Damage Reduction
15 SEconds of +10.23% damage reduction (from Armor and Vers) ~ 5.12% Damage Reduction
3 SotR casts * 3 sec = 9 seconds of 18.09% damage reduction (from SotR) ~ 5.43% Damage Reduction

Holy Shield:
30 Seconds of 15% block (40% reduction) ~ 6% Damage Reduction
6 SotR casts * 3 sec = 18 seconds of 11% damage reduction (from SotR) ~ 6.6% Damage Reduction
8 SotR casts * 3 sec = 24 Seconds of 11% damage reduction (from SotR) ~ 8.8% Damage Reduction

***6 SotR casts was based on the bare minimum number of SotR you can cast without any procs from Avenger's Shield or Divine Purpose. After a few tests using Avenger's Shield procs, I averaged closer to 8 casts over 30 seconds.


Totals:

4.55% Avoidance and 13.35% Damage Reduction

VS

12.6% ~ 14.8% Damage Reduction


How will stat scaling impact these results?

With raid buffs or as our Crit stat gets higher, we are going to start seeing Diminishing Returns from parry rating, meaning that 4.55% avoidance is going to start getting smaller. None of the other stats should be impacted by Diminishing Returns.

Another issue is the % uptime of Damage Reduction from SotR casts. Because one works on 9 out of 30 seconds and the other on 18 to 24 out of 30 seconds, mastery will scale better with Holy Shield as well.

i.e. if my mastery were a base 20% and jumped to 27% with Seraphim the numbers look different:

Seraphim:
15 Seconds of +9.09% parry ~ 4.55% Avoidance
15 Seconds of +7% Block Rating ~ 2.8% Damage Reduction
15 Seconds of +10.23% damage reduction (from Armor and Vers) ~ 5.12% Damage Reduction
3 SotR casts * 3 sec = 9 seconds of 27% damage reduction (from SotR) ~8.1% Damage Reduction

Holy Shield:
30 Seconds of 15% block (40% reduction) ~ 6% Damage Reduction
6 SotR casts * 3 sec = 18 seconds of 20% damage reduction (from SotR) ~ 12% Damage Reduction
8 SotR casts * 3 sec = 24 Seconds of 20% damage reduction (from SotR) ~ 16% Damage Reduction

Totals (with 20% based DR from mastery)

Seraphim: 4.55% avoidance, 16.02% Damage Reduction

VS

Holy Shield: 18% ~ 22% Damage Reduction.


Conclusion

In conclusion as our gear gets better and our Crit and Mastery Ratings rise, Holy Shield starts even the playing field. With my current gear however, I can determine that the increased Parry from Seraphim outweighs the Damage Reduction from Holy Shield. I will be doing further research on where the Crit/Mastery plateaus are that will start to make Holy Shield the better talent.



I'm looking for some help with numbers. Specifically in comparing Damage Reduction vs Avoidance. What % Damage Reduction = 1% avoidance? Also, does anyone have a calculation for the Diminishing Returns on Parry rating?

[Killingjudge]

Shameless bump. Really looking for some help with these DR numbers.

[sexybeast]

It may be worth tweeting Theck with a link to the thread so he can let you know?

Really nice write up btw, I thought I'd already said that but obviously didn't press send

[Killingjudge]

Thank you very much and thanks for the advice!!

[Earthbound]

I only play Ret, but I was thinking of making Prot my offspec for heroics...blocking spells just seems like a really good thing. Why isn't that talked about more?

[Deleted]

Your entire post relies on the assumption that we care about TDR (total damage reduction). We do not. It is an outdated and effectively useless metric. We care only about damage smoothness, not total damage taken. Your work, while admirable in the sense that a lot of effort has been put into it, tells us relatively little of actual use.

[Jester Joe]

I only play Ret, but I was thinking of making Prot my offspec for heroics...blocking spells just seems like a really good thing. Why isn't that talked about more?

The only thing I can think of off the top of my head as someone who just started raid tanking again is that most the time a boss uses a spell, it's aimed to someone other than the tank anyway. I mean, of course there's some bosses that are different, but still. Also last time I tanked was end of Cata, so I might just be wrong.

[Killingjudge]

I only play Ret, but I was thinking of making Prot my offspec for heroics...blocking spells just seems like a really good thing. Why isn't that talked about more?

At the top I quickly mentioned that HS gains a whole other advantage when spells are involved. Most bosses in Highmaul deal most of their damage through physical attacks or mechanics spells that can't be blocked, so other than really brackenspore, we are mostly worried about physical damage.

[Deleted]

you may want to consider set bonus's as well btw.

[Killingjudge]

you may want to consider set bonus's as well btw.

Once they become available, they will be considered. This is actually just an exert of a Highmaul guide I've been making.

[Deleted]

At the top I quickly mentioned that HS gains a whole other advantage when spells are involved. Most bosses in Highmaul deal most of their damage through physical attacks or mechanics spells that can't be blocked, so other than really brackenspore, we are mostly worried about physical damage.

Believe me, my friend - on Mythic Brackenspore, we are also mostly worried about physical damage.

[celinamuna]

dem 200k melee swings sure do tickle.

[Emrus]

Sorry to nitpick. But, you forgot to add that Seraphim also gives 1000 multistrike. It might not help be the best stat for mitigation, but the 1k multistrike still gives the equivalent of 18.2% extra selfhealing and 9.1% external healing while it is up.

[sexybeast]

dem 200k melee swings sure do tickle.

Yeah... This! Lol :/

[Deleted]

Yeah... This! Lol :/

I'm actually genuinely unsure what the hell they were thinking when they were testing Mythic Brackenspore.

"Hey, guys - I have this crazy idea for a boss! What if we make him hit for 200k all the time, and then also add a 5-second unavoidable debuff that makes it impossible to heal people, during which he still melees! There's NO way that could backfire, right?"

I mean, they promised us tank damage would be less severe than in MoP to compensate for our reduced self-sustainability, but I'M PRETTY FUCKING SURE HEROIC GARROSH NEVER MELEE'D ME FOR 650k...

Real talk though, don't take Holy Shield for Brackenspore. It makes the breaths ever-so-slightly less sketchy but leaves you complete fucked whenever the big add has 5+ stacks.

[Deleted]

So, being a bit of a critic here.
Your entire post falls short because you ignore a good few factors


-The assumption that TDR is actually in any way a relevant measurement
-Ignore DPS increase
-Ignore multi-strike gain from Seraphim
-You do not account of the difference in uptime and cooldown and resource cost, considering how they fit into different fights and playstyles. When comparing apples and oranges you need to first consider if you are making apple pie or orange juice. Making conclusions based on patchwerk simulations is a terrible thing to do for tanks.

"Hey, guys - I have this crazy idea for a boss! What if we make him hit for 200k all the time, and then also add a 5-second unavoidable debuff that makes it impossible to heal people! There's NO way that could backfire, right?"

Ka'Roz

You have to consider though that most people while doing Garrosh were comparably a lot better geared than people at mythic bracken are atm. Right now you are probably around 15 ilvls below Mythic item level, while on Garrosh you were most likely 4+ item levels above it atleast, considering your kill was in Feb I would assume you were almos 10-12 item levels above heroic gear. I am pretty sure me and Fraggoji were the only two protadins that did H:Garrosh below heroic item level gear on average, atleast for the kills.

So if you are comparing bracken with garrosh, you probably would have to think about how hard bracken would reck you at 690-695~ item level, which would be pretty manageable.

[Deleted]

So, being a bit of a critic here.
Your entire post falls short because you ignore a good few factors


-The assumption that TDR is actually in any way a relevant measurement
-Ignore DPS increase
-Ignore multi-strike gain from Seraphim
-You do not account of the difference in uptime and cooldown and resource cost, considering how they fit into different fights and playstyles. When comparing apples and oranges you need to first consider if you are making apple pie or orange juice. Making conclusions based on patchwerk simulations is a terrible thing to do for tanks.





Ka'Roz

You have to consider though that most people while doing Garrosh were comparably a lot better geared than people at mythic bracken are atm. Right now you are probably around 15 ilvls below Mythic item level, while on Garrosh you were most likely 4+ item levels above it atleast, considering your kill was in Feb I would assume you were almos 10-12 item levels above heroic gear. I am pretty sure me and Fraggoji were the only two protadins that did H:Garrosh below heroic item level gear on average, atleast for the kills.

So if you are comparing bracken with garrosh, you probably would have to think about how hard bracken would reck you at 690-695~ item level, which would be pretty manageable.

As a 10-man raider at the time with very little seniority and obscenely bad luck with overall drops I was barely 570 when we killed Heroic Garrosh - 571 to be exact, if I am not entirely incorrect, which accounting for instantly available item upgrades is roughly HC gear level. On top of that, Heroic Garrosh was clearly intended to be the hardest, most tightly tuned boss in the instance, while Brackenspore is supposed to be perfectly killable in entry-level Mythic gear.

In fact, a more accurate comparison in terms of damage would be Sha of Pride or perhaps Iron Juggernaut - both fights where with proper CD management you were never in any sort of danger even at 550ish iLvl so long as you weren't trying to be cute and you knew how to play your class.

The fact of the matter is that Blizzard has once again completely failed to deliver on the promises they made in terms of tuning and quite frankly have no clue how to challenge the current generation of tanks and healers without simply grossly overtuning damage intake.

[Theck]

Conclusion

In conclusion as our gear gets better and our Crit and Mastery Ratings rise, Holy Shield starts even the playing field. With my current gear however, I can determine that the increased Parry from Seraphim outweighs the Damage Reduction from Holy Shield. I will be doing further research on where the Crit/Mastery plateaus are that will start to make Holy Shield the better talent.

This is more or less what we've observed in sims. In Highmaul gear, Seraphim starts out with a fair lead over Holy Shield in terms of survivability. Once you get into Blackrock Foundry gear, the combination of set bonuses and having more stats available helps push Holy Shield ahead. Note that this is all as measured using a useful survival metric (TMI) rather than DTPS.

It isn't all that surprising either. Gaining 1000 of several beneficial stats is a huge gain when you only have 1000 or 2000 of those stats to begin with. As you start out with more of those stats, gaining 1000 is a less significant increase. That's one downside to the fixed-stat-amount model of Seraphim, though the devs can obviously adjust for that by buffing it as necessary (e.g. buff it back to 1500 of each stat for the T18 patch, etc.).

I'm looking for some help with numbers. Specifically in comparing Damage Reduction vs Avoidance. What % Damage Reduction = 1% avoidance?

That's not an easy question to answer. First, let's make the (bad) assumption that we just care about DTPS. If you start in a hypothetical case where you have 0% damage reduction and 0% avoidance, increasing your avoidance by 1% means you take 1% less damage, as does increasing your damage reduction by 1%. In that case they'd be perfectly equal.

However, now let's say you already have 25% avoidance. Adding 1% avoidance isn't 1% damage reduction anymore. It's 1% in an absolute sense, but not in a relative sense. At 25% avoidance, you're taking 75% of the possible incoming damage. At 26% avoidance, you're taking 74%. You're now taking 74/75 = 98.67% of the damage you were taking before, which is a reduction of 1.33%. You could do the same sort of calculation with flat damage reduction effects (assuming they're additive, like adding more armor or avoidance is). So the relationship between the two really depends on how much of each you have. And of course, there are DR curves on avoidance to consider as well.

Which brings us the the larger point - DTPS isn't necessarily the metric we care about as tanks. In fact, for most of the history of WoW, it hasn't been. If you look back at TBC- or WotLK-era theorycrafting threads, nobody cared about DTPS. What mattered was Effective Health (EH) - having enough EH to survive the boss's burst damage. The EH metric was a simple (but fairly effective) method of estimating how much burst damage you could take in a small window of time, assuming the worst case scenario (i.e. no avoids, since you couldn't guarantee them).

In a lot of ways, adopting the concept of EH was like Prometheus bringing fire to the theorycrafting community. We got a lot better at actually measuring what a tank cared about, but we also were still able to spectacularly burn ourselves with it. It was simple, but also very limited in what it could estimate because it intentionally ignored a lot of factors (avoidance, block if not block-capped, cooldowns, etc.).

I'd like to think that, by analogy, the Theck-Meloree Index (TMI) is me handing you a flamethrower. It's designed to measure exactly what a tank cares about - a burst window of ~6 seconds - in a way that gives you continuous variation. As such, it's a much more comprehensive tool than EH since it accounts for everything, including avoidance, block, etc. If you can sim it, TMI can measure it. Want to know how changing your cooldown patterns reduces your spike susceptibility? TMI can tell you that. Want to know whether you should start pooling holy power 3 or 5 seconds before Seraphim comes up? It can determine that too.

Of course, like a real flamethrower, you can also burn yourself horrifically if you're not careful!

But in any event, the point is that in a real survival metric, avoidance tends to get devalued somewhat. In EH, its value is strictly zero; TMI is quite a bit more forgiving than that, since it has more granularity. But avoidance does end up less valuable than an equivalent amount of flat damage reduction (as measured by equalizing DTPS), simply because there's always that chance you take four attacks straight to the face because you got unlucky.

Also, does anyone have a calculation for the Diminishing Returns on Parry rating?

Everything you need should be here:
http://www.sacredduty.net/2014/08/06...eturns-in-wod/

[Deleted]

As a 10-man raider at the time with very little seniority and obscenely bad luck with overall drops I was barely 570 when we killed Heroic Garrosh - 571 to be exact, if I am not entirely incorrect, which accounting for instantly available item upgrades is roughly HC gear level. On top of that, Heroic Garrosh was clearly intended to be the hardest, most tightly tuned boss in the instance, while Brackenspore is supposed to be perfectly killable in entry-level Mythic gear.

In fact, a more accurate comparison in terms of damage would be Sha of Pride or perhaps Iron Juggernaut - both fights where with proper CD management you were never in any sort of danger even at 550ish iLvl so long as you weren't trying to be cute and you knew how to play your class.

The fact of the matter is that Blizzard has once again completely failed to deliver on the promises they made in terms of tuning and quite frankly have no clue how to challenge the current generation of tanks and healers without simply grossly overtuning damage intake.

I do not think you can make that exact comparison as the entry bosses in SoO was tuned to be loot pinatas. That is only true for Kargath, and to some extent twins in HM. You also have to think of the damage orientation of the bosses. Bracken is mainly a physical boss for tanks while a fight like Jugger, the mines and fire was the primary damage.

A fair comparison would probably be doing Malkorok or Thok in 550~ item level, in which case I would say Bracken is a bit harder than Malk and a bit easier than Thok.
Malk and Thok was tuned for about 550~ item level in the same fashion Bracken is probably tuned at around 660~.

[Kheirn]

Something I've been struggling with which some of you might be able to answer, is how to pool HP for Seraphim. If I'm at 5 HP, CS and J are coming off cooldown, but it's still ~1.5-2 seconds until Seraphim is off cd, should I just wait for Seraphim to come up and wait with CS and J or should I use up 3 HP for a SotR even if that delays Seraphim and reduces uptime? How important is it to use it the second it comes off cd?