Last week, Ghostcrawler made some comments about the odd state of prot paladin gearing, in particular our love of haste. In a blue post, he gave us the following tidbit:
Ghostcrawler said: For Protection, we do have plans to try and lower the value of haste relative to dodge and parry. We don’t want to make haste terrible for paladins, but we agree that it’s odd for it to be better than more traditional tank stats. It might require a nerf to Shield of the Righteous to do this, but our goal is not to nerf survivability overall. We just wanted to provide you some context if you see odd changes to tanking abilities.
And later elaborated on their reasoning:
Ghostcrawler said: The game just isn’t currently designed to support it. It creates potential problems such as:
– A Prot paladin competing with a Frost DK or Ret paladin over gear, meaning there isn’t enough DPS plate to go around.
– A Prot paladin considering a tier set with dodge and parry on it to be “garbage” because it doesn’t stack all haste.
– A Prot paladin looking at a Ret 2pc set bonus that she normally wouldn’t touch because now the stats aren’t that bad either.
In a world where tanking plate didn’t exist or every loot system used the personal LFR one or 100% efficient reforging then it might work.
We understand that having a lot of haste feels fun and visceral and is more dependable than dodge and parry. We’ll try to come up with a solution that keeps that in mind.
As you may have noticed, this sparked a lot of uproar amongst prot paladins. There are almost as many different opinions on the topic as there are players discussing it. Some love our new haste paradigm and want to keep it, others are somewhat ambivalent, and yet others are extremely anti-haste.
If you want my opinion (and if not, why would you be coming to my blog in the first place?), I think the haste paradigm is fun, but sort of silly. Several of Ghostcrawler’s potential problems are things I’ve seen in game. And I’d even add one to that list – having to re-spec to Ret after a boss kill to use coins, because my ideal tanking gear is actually on the Retribution loot table.
From a design standpoint, it’s sort of strange that haste is more attractive than traditional tanking stats. If “tank plate” is going to be a gear category, it should be attractive to all three of the plate tank specs. So far they’re 2/3, since Death Knights and Warriors are on-board. We’re the oddballs that steal DPS gear. Of course, they could eliminate tank plate entirely, but I don’t think they want to do that for homogenization’s sake, so let’s assume that’s off the table.
We here at Sacred Duty feel partially responsible for this conundrum. See, we’ve been ringing the “all tanks need similar haste scaling” gong for quite a while, because in Cataclysm we were the only tank that didn’t get extra resource generation from haste. Mel made the point in a post in December 2011, and it’s an argument we repeated in other venues as well all throughout beta. While I don’t know that we can take credit for getting Sanctity of Battle during beta, we were championing the cause before it was cool.
And then, in an ironic twist, warriors lost their haste scaling in beta, leaving them as the only tanks without some sort of haste-to-resource conversion. So.. sucks to be warriors?
In any event, it’s clear that our affinity for haste is causing problems.If we have 20% haste, it means we have 20% more resource generation and Shield of the Righteous uptime, which is big. It grants resource generation to other tanks (well, except warriors), but it’s not strong enough to make them favor it. Bears gain more rage from auto-attacks, but that’s only a portion of their resource generation, not all of it. A similar argument could be made for warriors back when they
were cool got rage from auto-attacks. Faster energy regeneration gives a monk more Jabs, but not more Keg Smashes, and the global cooldown limits how many Jabs you can fit in around Keg Smashes and Blackout Kicks. A Death Knight would seem to get the most linear scaling with haste, as 20% faster rune regeneration ought to directly lead to 20% more Death Strikes. It’s actually a bit of a mystery to me why they don’t value haste the way we do, but their mechanics have inherent synergy between avoidance and active mitigation (Blood Shield) that may tip the scales in favor of mastery and avoidance.
In my opinion, the problem with haste and protection paladins is twofold: haste is too strong, but more importantly avoidance is too weak. First, let’s look at haste and what can be done about it; then we’ll discuss the problems with avoidance.
Haste Makes Was… wait, I’ve already used that pun
First, let’s clear up some confusion. Haste is not an “amazing” stat for protection. A lot of people seem to have come to this conclusion based on my results. But if you look at my posts, it’s not really that impressive. It’s a terrible stat for total damage reduction (TDR), and has been since beta. And it’s not even that great of a smoothing stat all by itself. That early simulation showed it strictly behind mastery in all categories.
It wasn’t until we discovered how strong hit- and expertise-capping was that haste really came into its own. And even then, if you read through any of the follow-up posts that analyzed the topic in more detail, the control/haste gear set didn’t show a significant survivability advantage over the control/mastery or control/avoidance sets.
Hit and expertise are really amazing stats for survivability. But haste is not. It just happens to be about as good as the more traditional tanking options after you’ve already maxed out hit and expertise.
Haste also has an ace up its sleeve that mastery, dodge, and parry do not: DPS. Sanctity of Battle makes haste a passable survivability stat, but simultaneously makes it a beastly DPS stat. Because our entire rotation scales with haste, 10% extra haste is 10% more DPS. In fact, once Vengeance is considered, haste is generally a better DPS stat than strength.
It’s worth noting that we’re the only tanks that get such a significant DPS bonus from haste. Death Knights, Warriors, Bears, and Monks are all limited by the GCD, so even with 20% higher resource generation, they won’t get a full 20% increase in DPS. At best they reduce the number of empty GCDs they have, or manage to replace a few more fillers with resource spenders. It’s only thanks to Sanctity of Battle that paladins aren’t subject to the same 1.5-second-GCD constraint.
And that’s really the problem with haste: it becomes a “king stat” for smoothing once hit and expertise caps are reached. We’re given a choice between three stats that all give roughly equal amounts of survivability, but one is also a significant DPS boost. We don’t lose anything by choosing to focus on haste instead of mastery or avoidance; that extra DPS doesn’t cost us anything. There’s no trade-off involved.
If haste is going to give us a huge DPS boost, it can’t also be our favorite survival stat. We have to lose something in the survivability department by choosing haste over mastery or avoidance, which means haste ought to be less effective than mastery or avoidance at spike mitigation. It’s not enough for it to be dead last for TDR – in fact, I suspect that over-reliance on TDR metrics for tank balance is what put us in this situation in the first place.
So how can Blizzard fix the “haste problem?” The most obvious solution is to nerf or eliminate Sanctity of Battle. If they took that ability away, it would take haste completely off of our radar. Curiously enough, in that situation we would still have better haste scaling than warriors because we’d get more Seal of Insight procs and Sacred Shield ticks at higher haste values. But it would be a small enough effect that we would ignore it for gearing.
Removing Sanctity of Battle completely has some downsides, though. It would re-create the problem we had in Cataclysm, which was that we scaled incredibly poorly with encounter-specific raid buffs like Essence of the Red. Partially nerfing the effect instead, such as reducing cooldowns and GCDs by 50% of your haste instead of 100%, would allow for the benefit of haste to be tuned more finely.
But either of those solutions risks making the haste gear that players have spent time collecting and valor points upgrading irrelevant. And the valor upgrade issue is a particular problem, as it’s a finite resource. If haste suddenly becomes significantly worse, players that stacked haste may feel like the victims of a particularly cruel bait-and-switch.
Since Blizzard has already suggested that they may try to fix the haste problem by nerfing Shield of the Righteous, let’s consider how they might do that.
- They could nerf the duration of SotR (say, from 3s to 2s). This would indirectly nerf all of our active mitigation stats: hit, expertise, mastery, and haste, while leaving avoidance more or less untouched. I don’t think this is likely though, as it would also encourage the “boss swing timer addon” issues that were complained about in beta.
- They could let the duration of SotR be reduced by Sanctity of Battle, such that at 10% haste the duration was divided by 1.1. This would keep the DPS benefit of Sanctity of Battle and keep the “fun” part of the hasted gameplay, but eliminate the survivability benefit. It would also make timing SotR to boss swings a pain in the ass, and probably isn’t all that simple to implement from a technical perspective either. And in practice, I don’t think we’d stack haste if it gave no survivability benefit, so this solution would feel sort of backwards. Removing Sanctity of Battle would be a simpler and more logical way to get the same effect.
- They could nerf the magnitude (mitigation percentage) of SotR. There are a few ways they could go about that nerf.
- The least likely is to nerf its mastery scaling – that would hurt all four active mitigation stats, but it would hurt mastery much worse than the others, and probably wouldn’t do anything to reverse the gearing paradigm. It would just make us like mastery even less.
- The more likely way is to reduce the base mitigation of SotR. For example, dropping the base mitigation from 30% to 20%, but leaving the mastery scaling untouched. That would reduce the smoothing effects of hit, expertise, and haste, but would leave mastery more or less unchanged apart from second-order synergistic effects.
- And of course, they could do both (nerf base mitigation and mastery scaling). Again, I don’t think that’s likely, as it would be doing a lot of damage to mastery for no apparent reason.
- They could buff the duration of SotR but reduce its magnitude. For example, if the buff duration was increased to 4.5 seconds, but the mitigation reduced significantly (i.e. 50% or more) to compensate. The mitigation reduction could be in either the base amount or the mastery scaling, though I think base is still more likely. This would mean we’d have much higher uptime on SotR in general, but the effect would be much smaller.I don’t think this is very likely either, just because it may end up having the opposite of the intended effect. In theory, having high uptime favors mastery over haste, because you’re getting an extra swing of mitigation out of that mastery. However, if we can build up enough haste to keep SotR up 90+% of the time, that’s probably more advantageous than increasing the mitigation value. It’s the SotR equivalent to block capping, and I’m pretty sure they’ll want to avoid that situation based on prior experience.
Obviously that’s not an exhaustive list, but it covers the most likely ways they could go about nerfing SotR. A more major overhaul of the system could completely change the paradigm too, but that seems pretty unlikely in the middle of an expansion.
I think the bigger problem facing the developers isn’t what to do with haste, it’s what to do with avoidance. As we’ve seen, haste isn’t a great stat, it’s just better than the alternatives. And the problem is that the alternatives, in this case dodge and parry, are far too weak to compete. Unfortunately, it’s difficult to pinpoint a single reason that explains why dodge and parry perform so poorly. It’s a fairly complicated problem, which also makes it hard to fix.
First of all, when you start out with low avoidance, adding more avoidance doesn’t do much for you. It’s obviously great for TDR, but going from 20% avoidance to 25% avoidance isn’t going to have an appreciable effect on the smoothness of your damage intake. And the diminishing returns equations intentionally keep avoidance from becoming too attractive once you have a lot of it. So right from the outset, you’re discouraged from even attempting to go down the avoidance path.
Damage models also have a pretty large effect on avoidance’s value. The game has embraced the triage model of damage and healing, such that we expect to be able to survive 3-4 hits from most bosses, even on heroic modes. That makes stochastic effects like dodge and parry strong for total damage reduction (TDR), but it still doesn’t help them in the smoothness department. They’re still not controllable, so you can’t, for example, use them to guarantee that you won’t take 5 hits in a row.
If the game were balanced such that healers risked running out of mana healing a tank, then TDR would be more attractive and stats like dodge and parry would have more room to shine. But in my experience, that ‘s not really a risk in progression raiding, and hasn’t been since the first few weeks of Cataclysm. Healers definitely run out of mana from excessive raid healing, but rarely do from a simple tank healing assignment. And more often than not, cleaning up execution on the fights does more for mana efficiency than gearing will. A healer can’t heal you much differently at 25% avoidance than at 20% avoidance, because it’s hard to notice that difference on the fly. The spikes they’ll be reacting to are going to be very similar, just slightly less frequent.
And in a tanking system that revolves so heavily around active mitigation, stats that contribute to the active mechanic tend to be far more interesting than passive stats. That’s one reason we focus on hit, expertise and haste. The synergy between those three stats and mastery serve to make all of them better, while dodge and parry have no such interaction. In fact, dodge and parry get weaker as you stack more active mitigation stats, because the average hit size being avoided goes down.
So dodge and parry are beset from all sides. Theyr’e great at TDR, but we don’t care much about TDR as tanks. They can’t compete with active mitigation stats for smoothing incoming damage because they’re stochastic and unpredictable. By gearing for active mitigation and smoother damage, we actually make dodge and parry less attractive. And perhaps the final nail in the coffin, they’re boring. Players don’t get excited about dodge and parry because they’re passive, set-it-and-forget-it stats. The function they perform happens in the background, with no input from the player, and lacks a feedback mechanism to make the player care about them. They’re reminiscent of the era where a tank’s survivability was primarily determined outside of the raid instance, while they were optimizing their gear. And that’s not the environment we live in anymore.
To paraphrase Mel: “Dodge and parry have been bad and irrelevant for three full expansions now, it’s not news that they’re garbage. They just used to be garbage floating to the top of a sea of garbage.” And I think he’s right. We haven’t been excited about dodge or parry since at least the Burning Crusade, if then. Dodge/parry have consistently been poor stats since the start of Wrath. We just never had a very strong alternative to compare them to until active mitigation made hit, expertise, and haste “good.”
I think any real “fix” avoidance will have to address not only their numerical inferiority, but their psychological inferiority as well. Dodge and parry are just too passive to be much fun. They don’t have to become active mitigation stats, but they need to interface with our active mitigation in a way that makes them interesting and exciting. Some small effect that makes us care about avoiding an attack would make an incredible difference in how dodge and parry are perceived.
Warrior’s Revenge is a good example of this. Avoidance isn’t just a passive background mitigation effect for warriors because avoiding lets them do something more often. That’s the sort of effect that will turn avoidance from a dump stat into something we look forward to, or even prioritize.
It’s not hard to come up with mechanics that would do just that. Here are a few examples:
- Change Grand Crusader’s trigger from CS/HotR to avoidance. Thus, avoiding an attack would have a chance to refresh the cooldown of Avenger’s Shield and grant holy power. That might be too similar to Revenge, but it covers all the bases: gives us something active to do and ties into our active mitigation through resource generation.
- Chance to generage 1 holy power every time you avoid an attack, with a 3-5 second internal cooldown to limit income during AoE tanking (the same could apply to the previous idea, of course). This is similar to an early beta version of Glyph of Blessed Life or Honor among Thieves. Passive resource generation that scales with avoidance.
- Avoiding an attack could grant a buff that increases the duration of the players next SotR by 0.5 seconds, perhaps stacking up to ~3. Or a 1-second increase that doesn’t stack. The stack limit would naturally keep AoE tanking from becoming too trivial.
- Avoiding an attack could increase the mitigation of your next SotR by ~5-10%, with a similar stacking mechanism.
- Avoiding an attack could reduce the cost of the next SotR by 1 holy power. This puts a natural limit on the resource generation effect, and has the potential to mix up the rotation a bit since we’d be able to cast SotR at 1 or 2 holy power on occasion.
I could go on, but you get the idea. Something – anything – that makes avoidance matter to us is an improvement. Especially if it does so in a way that requires attention on the part of the player. Otherwise, avoidance is doomed to remain the low-priority, boring stat that it has been for the last 5 years.
The TLDR version of this post is that the problem we’re facing isn’t just a case of haste being too strong. It’s good, but it’s not really all that strong compared to the defensive alternatives. Hit and expertise are strong, but haste, mastery, and avoidance are all fighting for a distant third place trophy. Haste just has better fringe benefits than the others.
The real problem is that avoidance is too weak, too passive, and too boring. And to add insult to injury, it’s anti-synergistic with active mitigation stats. So if we choose gear to make our active mitigation better, it makes avoidance even less attractive to us.
Reducing the value of haste so that there’s a legitimate trade-off between DPS and survivability is a good, and in my opinion necessary, first step. But the root of the problem really lies in the lackluster state of avoidance. If “tank gear” is going to continue to be a thing, then something really needs to be done to make it attractive to paladin tanks.
And the way to do that is to make it both numerically worthwhile and fun. What will it take to make avoidance fun? We need to get rid of the anti-synergy and add some real synergy between avoidance and active mitigation. Nobody likes their gear to work at cross purposes. We need it to be interactive, so that a player gets clear feedback that makes an avoided attack exciting and enjoyable. We need it to empower or enable us in a way that impacts our rotation.
And for avoidance’s sake, we need it yesterday.