The Light and How to Swing It: Secondary stat discussion for the discerning ret paladin

Since we're approaching what should be the final content patch before Mists of Pandaria, now would be a good time to discuss secondary stats. We saw in Wrath that things like haste and crit got out of control near the end of the expansion; a mage in my guild longs for the days when he could get his crit up past some absurd number and make fire "good again."
I don't have the heart to tell him that no amount of crit will ever make him good -- but I digress.
Let's take a look at those stats on our gear other than King Strength and how they work together to give us some big numbers.
The stats to watch
Hit and expertise Technically, our biggest DPS contributors after strength, hit and expertise are actually quite bland. Since both of them have "caps," they both stop being useful immediately after you reach their respective limits. If you really wanted, you could walk around with 25% hit, but you truly only need 8% (961 rating); any hit after that is wasted itemization that isn't going to help you hit the boss more.
In a similar vein, expertise stops being mostly useful at 26 (781 rating). Thanks to our Glyph of Seal of Truth, a prime glyph that every raiding ret paladin should have, we only actually need 16 (481 rating). I use the term "mostly" because there are different caps for if you're standing in front of the boss or behind, but since we are going to be staring at boss butt for almost every encounter, that is what we gear around.
After you pick off the hit and expertise pineapple on the pizza that is your new gear, what's left? Why, crit, haste, and mastery, of course!
Mastery Let's go over the easiest stat first, shall we? Mastery for the retribution paladin is a mechanic called Hand of Light. Essentially, each point of mastery will increase the damage of HoL by 2.1%, which originates from a percentage of your Crusader Strikes, Templar's Verdicts, and Divine Storms, after all buff and debuff calculations have been factored in (that is, Landslide, trinket procs, fight mechanics like Concentration, and so on).
Right now, this is generally the stat to stack, as it provides the greatest DPS return per point. There do exist scenarios where stacking mastery at the expense of other stats results in an overall DPS loss, but we'll get to those in a moment.
Critical strike Easy to understand but difficult to grasp, crit is on the heels of mastery in terms of DPS contribution. It affects every single attack we make, from auto-attacking while you're getting a pizza out of the oven to that mindblowing Consecration your target just hopscotched over. However, crit isn't as solid of a stat as mastery in that additional crit rating only increases your chance to score a critical hit on a target. While a 12% chance to crit is definitely better than a 10% chance to crit, we humble ret paladins know that when we leave decisions to the random number generator, we don't always get a desirable outcome.
Haste Oh, how the mighty have fallen. What was once our holy grail in Wrath is now nothing but a plastic cup with a leak in it. Recently, however, haste has been making a comeback. Censure, recently buffed by 40%, ticks every 3 seconds and is further modified by haste. The biggest contribution to haste's value is the fact that it buffs attack speed, meaning with an increasing amount of haste we get an increasing number of auto-attacks. These white melee hits not only do damage but lead to more Art of War procs and more Seal of Truth and Seals of Command damage. Even counting all of this, haste is still placing third in terms of DPS per point.
Of course, there is also the talent Sanctity of Battle. This is what made haste awesome during Wrath and continues to tease us to this very day. With current gear levels and talk of toning down ilevels and therefore stat budgets in the next expansion, we can probably safely say farewell to the concept of a 3-second Crusader Strike. Well, until Blizzard changes everything in Pandaria.
The myth of absolute stat weights
I really can't stress enough that absolute stat weightings are a myth. Most of what you see on sites like Elitist Jerks are generalizations meant to be applicable to the widest audience. With increasing item levels, however, we accumulate so much crit, haste, and mastery on our gear that they actually begin competing with each other for relative value.
Let's assume that you throw caution to the wind and try to stack your crit rating as high as it can possibly go. Your crit chance grows and grows while your haste and mastery remain at pitifully low levels. Somehow you reach 50% crit (not sure if this is even possible right now, but suspend that disbelief for a minute!), and you're looking to get even more gear with even more stats. If from this shiny new gear you were to get an extra 1% crit chance, that would logically put you at 51% crit. While it is technically a DPS increase to make this jump, since it comes at the expense of your other stats (that is, the pieces you are choosing are high in crit and relatively low in other stats), it is actually worth less than a 1% gain in crit chance at a much lower gear level.
How can we visualize this stat battle? With a versatile and very useful tool called SimulationCraft.
Without going to great lengths to try to describe how to use it, this program takes your character data and simulates your rotation, random procs and all, against a boss-level creature over a defined period of time. Then it does this again, and again, and again, up to 50,000 times in order to provide the most accurate results. There are options to do all sorts of stuff, but in the context of this column, let's look at the DPS scaling options.


Also of note, Exemplar's spreadsheet at Elitist Jerks is another great resource for determining stat weights and gear upgrades, but it isn't as flashy as something like SimCraft. There are a few minor quibbles between the models that each resource uses, quibbles that produce discrepancies such as haste being valued over crit for Exemplar's spreadsheet at a certain gear level, while SimCraft says crit is stronger than haste. If you're ever unsure about which stat would boost your DPS more and the community at large isn't sure either, stick with what has worked in the past until enough evidence comes forth that making the switch would be worth it.
Change: The bottom line
I'll be the first to admit I know what you're thinking: "Those are some fancy-schmancy graphs you got there, Dan, but what is the freakin' point?!" If you took nothing from any of that, remember that stat weights change, and what is good with a certain set of gear could be totally different with some new gear you get off of Onyxia's thrice-resurrected carcass.
Filed under: Paladin, (Paladin) The Light and How to Swing It






Reader Comments (Page 1 of 1)
apharrington Nov 9th 2011 8:03PM
Who's the skirt? Yes I mean the mom. Heyooo!!
Whig Nov 9th 2011 10:00PM
What's the value of mastery as opposed to strength? For instance is 10 strength worth 100 mastery? I know strength is by far the better stat but by how much?
I have a gear choice that requires me to choose between 15 strength and 110 mastery. Seems like strength but it made me wonder.
So is it possible to make some kind of rough equation, or are there variables I don't understand?
Apple Nov 10th 2011 12:32AM
Stat weights are preposterously complicated. If you wanted to get an analytical solution for the ideal enchant/gem/reforge setup, you'd be dealing with a six-degree-of-freedom stochastic differential equation. In simple English, a mathematical clusterf#(%. This is why the best theorycrafting programs literally play your rotation like the columnist described. Such a simulation will get a different result for different gear sets, which is why it's so hard to give an easy, numerical answer.
That being said, I'd ballpark the value of Strength as a little less than twice as valuable as mastery, crit or haste.
Elzam Nov 10th 2011 12:52AM
This is the kind of question that the blog aims to avoid: the relative value of stats changes too much with gear permutations. For example, I'm actually seeing my crit inch ever closer to the value of Mastery.
One thing that should be mentioned that I didn't see in the blog, however, is that currently Exemplar's spreadsheet favors Crit and Redcape's favors Haste. They've worked together and can't quite find a solution to why this is, so they've generally agreed to disagree. Because of this, I rely heavily on simcraft for personalized results.
Melodee Nov 10th 2011 7:21AM
@Elzam: Slightly mistaken (only slightly!). Exemplar's spreadsheet favours Haste and SimC favours crit, and Redcape isn't active anymore. They decided to agree to disagree as yeah, there was nothing they could find to explain this, and it was generally concluded that our helter-skelter RNG would have a bigger impact on your DPS than the secondary stat you chose. Interestingly, there was also a conclusion that the haste value on one was equal to the crit value on the other, and vice versa.
I have no idea how that's mathsy, I have to admit. I was certain, although my brain is a small, illogical place, that there would be a "right answer" at some point just because there had to be.
From WoL parses/ret paladins in high end progression guilds, the answer for "best" doesn't seem much clearer - some reforge away from crit before haste, some reforge away from haste before crit, some (this was discussed on EJ) maintain a 2:1:1 ratio of mastery to the other two stats. There's even some debate about haste being better at lower latency.
Sometimes I miss simple times /o\
Dan Desmond Nov 10th 2011 7:36AM
Elzam, I actually did mention that here:
"...quibbles that produce discrepancies such as haste being valued over crit for Exemplar's spreadsheet at a certain gear level, while SimCraft says crit is stronger than haste."
Whig, Apple is right. As I stated, stat weights are quite variable and there aren't simple equations - there are complex equations that each sim is based off of, and we can see in the disagreement between Exemplar and Sjogren (one of the SimCraft devs) that they don't even have the same equation.
That said, generally so far we have seen that 1 point of strength has been worth anywhere between 1.4 and 1.8 points of mastery. If you're choosing between 15 strength and 110 mastery, the 110 mastery should win no matter what.
Zaet Nov 10th 2011 7:59AM
Anyidea if for the final tier level of gear will stacking haste will be viable?
in the way that we can get CS to a CD so low that we casn do: CS-judgment-CS-something else-CS-etc.
gewalt Nov 10th 2011 9:22AM
Just strap on the highest ilvl gear you can get, (ignore any items with strength deficit exp/mastery gear with stam bonuses usually have a 40-50 str deficit to flag them as specifically designed for tanking) and reforge to your hit/exp caps.
Every other decision you can possibly make about ret gearing is completely irrelevant compared to spending 2-3 minutes practicing rotation on a training dummy.
Ret is in the position that blizzard has been trying to get every other spec to.
gewalt Nov 10th 2011 9:23AM
Really now comment system... this was NOT supposed to be a reply to zaet.
Brad Joseph Nov 10th 2011 11:14AM
Should Crit and Mastery almost be equal then? My wife plays a ret Pally, and i try to help. I feel like her mastery is in the 18's and crit is in the 8's...or some difference like that. What should those numbers be around for someone with item level 368? I always have a hard time figuring it out. Her dps is typically all over from 10k to 19k depending on the buffs it seems.
terph Nov 10th 2011 1:02PM
Well, that's an interesting question because it also depends on her trinkets and the situation. I had to increase the gap between mastery and crit recently because of a weird issue.
I got a bunch of new gear all in one run including a crit-boosting trinket(Vessel of Acceleration). Stand-alone, my crit rating actually got rather close to my mastery rating. Still more mastery, but not the huge gap that I had before.
The problem was that I also had the Apparatus of Khaz'goroth trinket, which will boost your highest secondary. You definitely want the trinket to pick mastery, but with VoA boosting my crit, I would very temporarily have more crit that mastery. And since I was stacking my cooldowns, the AoK trinket was boosting crit. And it was not good. So I had to juggle my reforging a bit to make mastery have a bigger lead over crit.
cursedmonk87 Nov 10th 2011 3:26PM
I do miss those haste-stacking days, Pulling top dps on LK above better geared locks and mages. I have to admit it's been hard to kick my speed habit; it's just logical: More haste means more Crusader Strikes, means more Holy Power, means sooner finishers, right? Well I just started raiding Firelands (after a summer WoW-hiatus) and I just couldn't get my numbers up with it. I've been leaning more towards Mastery in re-gearing after Hit/Expertise, Crit still just seems like an offering to the RNG gods, who are very fickle indeed. Our rotation is small, and it's slow, I love the Exorcism proc, but other than that I don't find Crit to be paying off over the base damage increase you get from mastery. Divine Purpose is still an RNG love/hate, but at least I don't have to worry about making that 15% proc any better or worse gearing-wise.
Off-topic to stats, I would still like at least one more attack to fill my rotation, maybe a bleed or other DoT that's a little less RNG dependent, or an AoE finisher like Divine Storm was at the Cata start. But I'm getting a little bored with Jdg, CS, HW, ...CS, Jdg, CS, BIG SMASH!!!, Proc, proc, proc, repeat. I actually quite enjoy the Holy Power system (*ducks*) but I tend to waste a lot of procs just because I don't want to lose them after a pull. I mean Rogues have what, 5-7 finishers/point burns, can't we have a third? Don't just buff our numbers, give us more damage utility!
/end rant. Continue on-topic discussion.
cursedmonk87 Nov 10th 2011 6:40PM
Whoops I should actually take the time to read the Class Q&A transcripts before going on a rant. In 5.0 (with a grain of deeprock salt), Divine Storm is going to be a finisher again, and ret will get Hammer of the Righteous added to our rotation for another AoE and a charge of HP. And of course Judgement will give us some HP too. Okay now I'm happy-ish.
Go about your business, Knights.
Wikid Dec 20th 2011 10:46AM
ive done a lot of research on the subject of crit over haste, and i ascribe to the school of haste. i know a lot of ret pallies that favor crit, and their damage is typically all over the spectrum. some fights they do squat while some fights their damage goes through the roof. MY ret pally, however, favors haste above crit, and i have a much higher "baseline" damage output, meaning i wont do less than "x" but can do much more with some lucky procs. As any retadin knows, you hit one wrong button and dont get lucky with a proc, your choices are to either wait for a cooldown or fire off a timed-cast Exo, which is a mana drainer. With my current gear (mostly 378's, a few 391 valor pieces, and the 2 piece tier set from raid finder which are 384) my CS cooldown is 3.5 seconds with the buff from judgement, so even if i flub i only have to wait .5 seconds before i can CS again. The new 2 piece tier bonus is the kicker, as it makes your judgements give you 1 holy power. If i do cs, judge, cs, bam 3 holy power, only half a second downtime and im smashing that TV after 3 gcd's. With the gear i have, there is enough crit on it that i can reach hit and expertise cap easily by only reforging away from crit, and then whatever crit is left over goes into mastery. If i wanted to to keep all that crit, after refoging everything with haste i would still have to cut into either crit or mastery just to reach hit/expertise cap, even with 10 free expertise from the glyph