I’m not especially proficient when it comes to math, and this barely qualifies for a napkin label, but if anyone could help validate some findings, I’d surely appreciate it.

I carried out some testing of Holy Prism versus Light’s Hammer. The main points of the conclusions were as follows:

When you use yourself as the prism, you hit up to five mobs for 20.840 (this was my average non-crit attack over the course of a more than 1000 attacks) holy damage each. If five mobs are present, you thus do 104.200 Holy Damage whenever you use Holy Prism.

With Holy Prism’s 20 second cooldown, you can use it three times for every one time you can cast Light’s Hammer. So, to compare, Holy Prism would on average do 312.600 Holy Damage in the same time frame within which Light’s Hammer can be used once.

In a similar test, my Light’s Hammer attacks hit for 6960 per mob, per tick. That translates into 278.400 Holy Damage on average when 5 mobs are in the arcs of Light’s Hammer for its full duration.

The obvious advantage of Light’s Hammer is, of course, that it’s able to heal a similar number of players while doing that damage, where Holy Prism, in this scenario, would only heal yourself. However, in terms of damage output, you’d need a substantial number of mobs (6) to stay inside the Light’s Hammer arcs – and stay alive, of course – for the full duration of the spell before Light’s Hammer outdoes Holy Prism.

Did I miss something? I was comtemplating whether or not spell delay due to priorities would be a factor, but I - and this is purely anecdotal - rarely, if ever, find that I can't use Holy Prism on cooldown.

Not as far as I can tell. In terms of AoE DPS in isolation (i.e. only considering the two spells), Holy Prism's damage wins out over Light's Hammer except in a few specific situations (adds that live between ~10-19 seconds, such that the comparison is only 1 HPr to 1 LH).

In practice, Holy Prism is just harder to fit into the rotation. If you're running a tight rotation, you really shouldn't have much room for Holy Prism (unless you're not using those spare GCDs for Sacred Shield maintenance, which is a fair point if you're only concerned with DPS). It's still a raw DPS gain over LH, but not by very much (<400 DPS, <200 HPS) because it requires you to push back other fillers (Cons and HW). In my own experience, I barely ever have time to cast Light's Hammer once per minute, let alone finding 3 GCDs for HPr.

That said, if you're aiming for raw DPS, you're not going to pick HPr anyway, because ES is better still (to the tune of ~1500 DPS).

I use Light's Hammer because it's more versatile in raid encounters, in my experience. It's not the best DPS spell, nor is it actually the best raw HPS spell (hostile-cast HPr should still win that contest, I believe), but it's a mini-tranquility. So I generally use it to help recover from raid-wide damage events. There are enough "group up and heal through massive AoE" moments in MSV and HoF that we've been planning out LH rotations the same way we do with Devotion Aura (though to be fair, having 2 protection and 2 holy paladins in a 10- or 25-man raid is not necessarily ordinary).

Just to add to Theck's comments on Light's Hammer, since its healing and dps are relatively concentrated time-wise compared to Holy Prism making it more "cooldown"-like than "rotational" if that makes sense. As a result, its much better for bursty situations. Force and Verve on Vizier is fantastic for Lights Hammer since their durations are roughly similar and with good placement you might get half or more of the raid within its area of effect.

Additionally, and IMO what really makes it shine, is that it can do *both* its full healing and full damage. ES with high Vengeance hits like a truck or heals a ton, but it won't do both. HP (which, ahem, I wouldn't use on Garalon. Body is too big) can be a nice addition to your rotation but it will either do modest single target healing and decent AoE damage or modest AoE healing and decent single target damage, but it can't do both. LH, however, is AoE for both healing and DPS at all tims.

On, say, Wind Lord, its amazing. We group up the boss and the adds and AoE them, with the raid spread around the mob mosh-pit. LH can hit all of the non-CC'ed mobs, damaging them all, while also healing both tanks, all the melee and anyone else who is standing in close.

Edit - Also, incorporated a bunch of typo fixes and additions/omissions that people have sent me in the last week or two and I've been slow about putting in. Thanks for sending them!

Edit2 - They've updated the real PPM list and added a few more enchants. River's Song is now 4 rppm, Dancing Steel is 2 rppm, and Colossus is 6 rppm. So, with zero haste and 100% on-target time (and Hit and Exp capped), you should get roughly 40% uptime from Dancing Steel, 45% uptime on River's Song and 6 Colossus procs per minute.

I didn't see mention of this after the first post, but what do you guys think about "seal twisting"? That is, getting 5 stacks of Censure on the target and switching to Seal of Insight until it needs to be renewed. This of course is very threat-dependent and would require monitoring your threat as well as knowing to switch back before Censure drops.

It's not a game breaking amount of healing, but I always feel it's good to not stress out healers if possible. I only have been doing this in heroic 5 mans, but I'd appreciate any insight as I plan to start raid tanking after a few LFR runs. If it's a stupid idea, I should learn and stop doing it now.

But the buff durations are very different.

DS is 2 RPPM, 12 second duration
RS is 4 RPPM, 7 second duration

So over the long term, DS is 24s/60s uptime of 1650 strength, or ~660 strength
RS is 28s/60s uptime of 1650 dodge rating, or ~770 dodge rating

With the state weights on askmrrobot (str at 0.5, dodge at 0.25), dancing steel still comes out strongly ahead with DS at 330, RS at 192.5.

Windsong is 2RPPM (random procs) that's 8s/60s haste, 8s/60s mastery and 8s/60s crit, 1500 rating each. At the haste > mastery weights of 1.0 haste, 0.9 mastery, that works out to 380 points. So windsong might still be better actually. I have no idea how to weight colossus though.

I was filling in the spreadsheet that's included in the OP yesterday. On opening (as xlsx) it produced an error resulting in a message saying 2 formulas had to be edited/deleted (one in each sheet). So not sure if anything is wrong with the conversion to xlsx or if the file broke.

Then after filling in the stats from my armory the numbers seem a little off, especially the weighting of haste. I'm unsure if this had to do with the formulas being edited, my difference in stats (haste is almost 4k, mastery 770) or if I'm just missing something.
Any info around that?

Besides that, as far as lvl90 talents go, Light's hammer really seems the way to go as tank. specific situations could put other talents ahead, but in general the fact that you'll probably have at least 1 friendly and 1 enemy in it for full duration (you and the boss) makes it more useful than ES already (more total output).
I only use ES if I specifically need to do more damage, or more healing, which is rare.
HP will definitely find it's use in several aoe-heal OR aoe-DPS situations but I found it very hard to rotate it properly.
On top of that, it's output is more smooth over a fight and less packed together like LH is, and thus (in my opinion) less valuable.

A last concern I have with my current tanking performance is the absolute MASSIVE amount of parry on available gear. My ratio is completely off and I see no way I can balance it out without sacrificing a bunch of haste/mastery in favor of dodge, which I don't see being a better situation at all.
With 4 items in my gearset having dodge, and 8 with parry, I found it hard to get under a ratio of 5.5, far from the optimal 3.5.
What are thoughts on this? Should I balance these out in the expense of haste/mastery or ignore it momentarily and do my best reforging away parry and picking up dodge pieces?

The spreadsheet problems are likely due to how Google Docs handles some things differently than Excel does. Since the document works fine on Google in the first place though, there's really no need to port it over to Excel. Running it in its intended environment should fix any issues.

The dodge:parry ratio is really only applicable for those who have chosen the avoidance path. Since you're in the control/smoothing gear setup, ignore it entirely. Now that we can reforge into two stats that aren't dodge or parry there's practically no reason to worry about the ratio in reforging, gemming, or gearing as a whole.

What bothered me about the spreadsheet is that I have opened and worked with it before as xlsx format and it didn't give any errors at all. I'll try testing it as a google document tomorrow though, thanks for the tip.

And about the gear, I've read that wrathblood had underrated the value of dodge and parry after hit/expertise cap due to the damage already being smoothed so much. I'm just unsure how this maths out and if there it's possible to reach such high level of haste/mastery that dodge would become that much more effective. Much how DPS stats for other classes/specs synergize over different tier, prioritizing different stats after a long time of stacking up another.

You did however made me realize there is no dodge:parry ratio for us indeed and that we need to look at all our 4 secondaries together. The ratio only dictates whether more dodge is better than more parry (or vice versa) in "snapshot moments" of our gearing.

Thanks for the clarification. From looking at the spell (Seal of Insight - Spell - World of Warcraft), it's not obvious that it couldn't crit (which seems odd, tbh).

I had originally sent this information as a PM since I wasn't yet sure if it was good enough to post, but now that I've completed it I believe it is:

I recommend everyone read Damage Smoothing Follow-Up | Sacred Duty thoroughly. Note that while the exp/avoidance set does perform decently well at longer attacks, it isn't spectacular.

If you really want to test whether avoidance starts gaining ground then I might recommend playing with the spreadsheet if you get it working. Note the avoidance weights compared to the haste/mastery weights (and make sure you're hit and expertise capped), then add 100 to haste or mastery (and try 100 on both of them) and see whether avoidance's value changes any. If it goes up then there is potential for avoidance to start getting better at a certain gear level. If it stays the same or goes down though, there's no chance whatsoever that avoidance can be as good as haste or mastery.

One of the things to keep in mind with this is we have two very synergistic stats in haste and mastery. Not only do you have to look at whether avoidance's value might go up, you also have to look at whether haste or mastery go up too, and whether they go up by more than avoidance. If they go up by more than avoidance and routinely do so as you add more stats then chances are avoidance is something to, well, avoid in favor of haste and mastery assuming excellent play.

As for the ratio blurb: It is not that we don't have the ratio. We'll always have the ratio where adding more dodge or parry will minimize the diminishing returns on them. It is whether we even need to consider it is the question since other stats have more value. If it weren't for the fact that we can reforge into two stats instead of one to get good gains then the ratio might be thought about more due to the nature of reforging.

Huh. I just crunched some numbers. I may be eating some crow. At 10-man damage levels, Colossus appears to beat everything pretty handily in TDR right now. I won't include any haste because it should impact all their proc rates similarly.

At 6 procs/minute, that's 48k/minute in shields or 800 dtps. I crunched my numbers a little differently from Theck, but the general result was in the same ballpark. The 200 Haste and 200 Mastery from Windsong requires 213K dtps (blockable physical dps, that is) which is vastly beyond anything even 10-man heroics will throw at you.

Obviously you wouldn't be trotting Windsong into heroic Will, though, right? You'd bring Dancing Steel or Riversong. Lets try out Riversong because it seems like it'd be the best at TDR. 45% uptime on 1500 Dodge is an average of 675 Dodge.

As Theck demonstrated in his follow-up to Damage Smoothing, the more Hit and Exp you stack (and Haste, but that's another topic) the better Mastery gets. By the time you've capped Hit and Exp, Mastery becomes better at TDR than Parry or Dodge. So, 675 Dodge isn't as good as 675 Mastery for reducing damage. While the exact % will vary from gear-set to gear-set, with the set I'm using its roughly 75% as good.

Riversong is a good bit closer, but the breakeven merely drops to 118K DTPS (dodgeable damage, that is). Which is still multiple times higher than the nastiest 10H MSV fights.

Part of the issue is tank swaps. If you're only tanking Feng half the time, then your overall melee DTS won't be as high as your "effective" melee DTS when you're actually on him. The problem with this is that Colossus is winning by more than a factor of 2. Its winning by a factor of 5 or more. If you could Dodge or Block Stone Guard's Rend Flesh (you can't, this is just hypothetical), it would be by far the closest, but it still only gets you up to 40-50k DTS, 1/2 to 1/3 your break-even.

Colossus is actually better on tank swaps (a lot of them on this tier) because Dodge becomes valueless when you're no longer taking melee swings. But there is almost always some sort of AoE going on to be absorbed so Colossus ends up retaining some percentage of its value.

So, that handles TDR. The question then becomes which handles burst better. Neither is controllable so that's out. Really, we're looking for which will be up the most often and take the biggest bite out of that damage. But what constitutes a "burst period"? This will vary from fight to fight, but I would say its 4-5 seconds. River Song is up a flat 45% of the time, so it will be up, roughly speaking, for 45% of your burst period.

A simplistic way of looking at it is that the length of your burst period as a % of Colossus's proc frequency is the likelihood of you getting a proc during the burst. Colossus procs roughly every 10 seconds, so with a 4-5 second burst period, you have a 40-50% chance of getting a proc during that period or roughly comparable to River Song. So, Colossus has better TDR and comparable burst reduction to River Song. Yikes. I didn't see that coming.

I'm not ready to anoint Colossus the new BIS enchant until I've thought things through a bit more, plus I'm partial to Dancing Steel over River Song anyway. Colossus vs Dancing Steel is an interesting tradeoff, very similar to Haste vs Mastery in a way with one bringing more TDR and the other bringing more DPS. But, if this holds up, the new enchant hierarchy would go something like:

Dancing Steel = Colossus > River Song > Windsong > Weapon Chain