Retribution Paladin Rework in Review – AoE Trade-Offs, Scaling, and Patch 10.2

In terms of playstyle, Crusading Strikes has become the dominant build in the last few months and is much more popular despite the large amount of downtime it causes. Personally, I enjoy playing around timed Holy Power gains, so either Crusading Strikes or builds that hardcast Crusader Strike are both styles that I think are fun to play with. It does feel a little strange to play with Blade of Justice generating 1 Holy Power, but rotationally the spec is still great and miles better than feeling forced to weave in low impact buttons like Consecration.

Overall, the rework did a great job at addressing the issues keeping Retribution from seeing play at higher levels and improving how fun it is to press your buttons, and the increased popularity of the spec is a testament to that. Although most of the rest of this article will be fairly critical of how the tuning has been left, the design itself is still great and I can’t emphasize enough how much more fun it is to play. The problem we’ve run into now is that there’s been almost no adjustments done since that release to address any of the issues that have become apparent more recently.

Talents

There’s a lot to like about the Ret talent tree – it has a ton of choice nodes with well designed, interesting abilities, and you have essentially no nodes that are purely designed to roadblock you from taking the talents that you want. You can just about take any talent you want from any position in the tree. It’s a massively improved tree thanks to the rework. The biggest issue with it now is that since that rework, there’s been essentially zero work done to address any tuning issues left since that point.

There’s a substantial portion of the tree that is just never taken under any situation in PvE, mostly because they’re just too undertuned. Some examples:

Not to mention talents like Justicar’s Vengeance and Consecrated Ground that are clearly designed for uses other than raiding or Mythic+.

I would love to have alternative builds focused around buffing Divine Hammer with Sanctify or making Judgment much stronger with Judgment of Justice and Judge, Jury and Executioner, but these talents are so undertuned that it’s just not feasible to experiment with them.

The talent tree as it is has ended up pretty stagnant, with a single very obvious best way to build it for single target, and little room to experiment with different builds past the Crusading Strikes / Templar Strikes dichotomy. For Mythic+ you have more leeway to modify it, but it ends up being the same 27 or 28/30 talents with the last two or three being personal preference choices rather than genuinely new builds. The talents you don’t take are just too much worse than the talents you always take to compete.

AoE Issues

At the moment, you have to have a really really good reason to justify taking AoE talents in raiding situations. On almost every boss, even bosses with substantial cleave like Forgotten Experiments or Echo of Neltharion or Sarkareth, the vast majority of players will use a full single target build even though there should be value in being able to do cleave damage on all of these fights. The reason for this is twofold; firstly, putting a point or two into cleave talents barely increases the overall damage you do. Secondly, the amount of single target damage you give up to gain any cleave damage is never worth the gain outside of basically Mythic+ situations.

On these types of raid encounters, the talents that you can realistically take are Blessed Champion and Inquisitor’s Ire or Tempest of the Lightbringer, instead of Divine Wrath and Vanguard of Justice or Penitence. Any AoE or cleave gain here is small, in the region of a couple of percent. There’s just not enough of an incentive for players to want to run these talents even when there should be an opportunity to use them, since the potential gains are so small and the single target loss is usually much more substantial.

Here’s the single target damage loss you currently take from speccing into various AoE talents, all of which are used in AoE oriented builds such as for Mythic+:

In total, Ret loses around 22% of its single target damage in order to spec into AoE talents, which is about 30k DPS in current gear.

Anticipating comments that “These numbers don’t add up to 22%!”: Percentages can’t just be added together, and some of these talent swaps end up overlapping with the effective value lost, e.g. just swapping Final Reckoning and Execution Sentence also massively reduces the value that Executioner’s Will provides, even though you still have it talented.

Losing single target to gain AoE is not in itself a bad thing; every spec has to do it to some extent, and it at least theoretically should create more situations where you use a variety of different talents. The problem is that to the best of my knowledge, in 10.1 Ret loses more single target to take AoE than any other spec in the game. Running simulations on each spec’s t30 SimC profile in both regular single target talents and the most popular Mythic+ build to approximate AoE oriented talents, I ended up with the following results:

Here’s the data used to make this graph if you want to take a closer look at any of it. It’s likely that for many specs listed there are builds that lose a much lower percentage to still do reasonable AoE damage; this graph’s purpose is to illustrate how bad of a situation Ret is in relatively rather than try to make fully representative comparisons between all specs. Even if a spec is listed at say, 12%, and a build exists that is actually only an 8% single target loss, it is still much better off either way than Ret that will guaranteed lose more than 20% of its single target damage. There may also be some specs for which Mythic+ talents aren’t a perfect approximation of what they’d use for an AoE oriented build in other content, but they will usually strike a fairly good balance between taking all reasonable AoE talents while not dropping any major single target gains where possible.

In any case, it’s clear that Ret is a massive outlier for how much it loses to gain AoE. It’s the only spec that loses more than 20% of its total damage, and even of the specs above 15%, most of the others are receiving massive rebalancing in 10.2, especially Rogues and Havoc that are receiving major reworks. Ret loses twice the damage that most specs lose on average. The DPS loss that Ret suffers for swapping a single talent like Execution Sentence is more than what some specs lose for their entire AoE build. I could go on, but the point is clear – Ret’s talents need tuning so that the loss is much less significant, and a build other than pure single target is viable for raiding.

These comparisons are all done with 10.1.7 numbers, but there will be some differences in 10.2. So far we’ve had the penalty for Blade of Vengeance removed for your primary target, which reduces it from a ~4.5% loss to merely 2%. This is an improvement, but will not significantly change Ret’s overall problem – with only this change, it will still remain the spec with the most single target lost in the game when using an AoE build. In addition, the current tier set makes Hammer of Wrath a usable button for AoE, which in turn keeps Vengeful Wrath and Vanguard’s Momentum usable on AoE. Without it, Hammer of Wrath will drop to bottom priority for AoE and its associated talents will no longer be valuable either, widening the gap between single target and AoE focused builds as some will opt to drop Vanguard’s Momentum and swap Vengeful Wrath for Divine Purpose.

On top of the talent issues, Ret has ended up in a similar position to 9.0 where Divine Storm isn’t worth casting on 2 targets without running all of the talents that buff its damage, leaving you with just Wake of Ashes, Empyrean Legacy, and Consecrated Blade as cleave damage, along with a couple of other small proc effects. A weak Divine Storm also means that Ret’s damage on 2 or 3 targets is especially weak, even when you take all talents that you have access to which improve AoE damage, which will become glaringly obvious on encounters like the Council of Dreams in the upcoming Amirdrassil raid.

The result of all of this is that Ret has ended up being a spec that does poor single target damage in Mythic+, and basically only has the option to do single target damage in raids. There’s virtually no talent diversity based on the encounter since you only have one viable way to build your tree, even when there are extra targets you could cleave. This drawback hasn’t been a huge issue for damage in Aberrus since it’s an extremely single target oriented raid, but it’s unlikely that future content will mask this weakness to the same extent going forward.

Scaling

It’s difficult to try to discuss scaling since it’s a pretty nebulous term that players sometimes use to describe a few different phenomena. From my experience, when players talk about scaling they most commonly are talking about how much a spec gains from an increase in stats or item level relative to other specs, so that’s the general concept that’s being talked about here.

Despite no major nerfs or other losses such as a much weaker tier set, Ret has gone from comfortably one of the top DPS specs at the end of Vault of the Incarnates to middling to lower end now, nearing the end of Aberrus. This is a pretty good indication that a similar thing is likely to happen in Amirdrassil, especially since the difference in item level between tiers is going up from 26 to 39. I believe the main issue is that there are just a lot of poor interactions with secondaries for Ret. For example, Critical Strike is less effective than it would be otherwise since Hammer of Wrath, one of your hardest hitting abilities, always crits, as well as Avenging Wrath providing a large amount of it as a regular cooldown. Mastery is now pretty bad since it was nerfed by almost 40% during the rework, to the point that it’s just a worse version of Versatility since it only affects Holy damage. The stats that Ret ends up using are defined more by being the “least bad option” than actually having significant positive interactions with any part of our damage toolkit.

On the other hand, the next tier set for Ret is certainly going to be better than the current one even despite the recent nerfs to it, and Ret will also be able to use the new Legendary weapon, so there are certainly some new systems that could benefit Ret more. As long as active tuning is done it will probably work out fine, but it’s definitely concerning to go into a tier knowing what happened prior, especially when Ret is one of the lowest performing specs at high percentiles…

Low Skill Ceiling, Low Skill Floor

Ret is one of the easiest specs in the game to pick up and play at a “not totally awful” level. The basics of the rotation are very straightforward, and even if you end up mismanaging your resources, your damage is not significantly negatively affected. It’s not necessarily a bad thing to have specs in WoW that are friendlier for players who want something they don’t have to focus on as much to get decent output from them, or for newer players to have something easier to learn, and given the popularity of Ret it makes sense to have it be one of them.

However, past the point of learning the basics, there’s very little to you can do to maximize damage. Managing your resources better and optimizing your cooldown windows doesn’t actually result in any significant damage gain. A cursory look at top logs for many fights will show many players managing to do very high damage despite wasting massive amounts of Holy Power or completely misaligning their cooldowns, something which would have had a real negative impact on your damage at any other point in the game. While it’s good to have a spec be accessible, you also need to keep the incentive for improvement at it for everyone else.

The impact of this can be pretty clearly seen on log statistics, as an example. At low or middle percentiles, Ret is somewhere between average to very strong, depending on the specific difficulty or encounter you look at. At higher percentiles though, from around 90% or above, Ret suddenly drops to one of the bottom performing specs. Essentially, a 20th percentile Ret parse is a lot more damage than a 20th percentile parse for most other specs, but a 99th percentile parse is much worse. I don’t think this is a problem that can be addressed easily or quickly, but I hope it’s considered when any changes are made next expansion.

Final Note

This post probably seems pretty negative, but it’s more relevant to write about things that still need some improvement rather than things that are already great. Again, the gameplay is great, and the defensive and utility options you have are all widely applicable and make Ret valuable in a ton of different types of content even when its damage is just average. Being in a situation where the main barrier to competitive strength is tuning problems, instead of fundamental questions about the value your spec can provide is a much better position to be in than Ret has been for a long time, but it doesn’t mean those tuning problems don’t exist either. I’m hoping to see some more active tuning done to help the areas mentioned here to match the fantastic work done in 10.0.7.