The last class preview has been posted, this time for the Witch Doctor.
Out of the five Diablo III classes, witch doctors are receiving the most attention in patch 1.0.4. The goal for this patch, like for many of the other classes, was simple: identify the unpopular or hard-to-use skills, figure out what’s not working, and then make them better. In some cases, skills only needed slight tuning -- a little more damage here, or some increased durations there. In other cases, more significant changes were required. For the purpose of this preview, we'll focus on the bigger changes, which can be broken down into the following categories:
One of the core play styles for the witch doctor (and indeed the reason many people chose to play a witch doctor to begin with) is to have pets. Unfortunately, while witch doctor pets do pretty well in Normal difficulty, their survivability has been virtually non-existent in Nightmare, Hell, and Inferno. From our perspective, this isn't acceptable, so we're making some significant buffs to pets in 1.0.4. The goal of these buffs is to make pets not only more viable in those later difficulties, but also more enjoyable for players who prefer to base their builds around them.
From a design perspective, we want your pets to be durable enough so they can tank for you, but we don't want them to just be automatically immortal. The cooldown on summoning pets is there for a reason. Speaking more specifically, we'd like for there to be times when your pets have died, your cooldowns haven't refreshed yet, and you have that period of increased tension as you wait for the situation to stabilize again. On the other hand, we'd also like for there to be noticeable improvements for players who put thought and effort into their skill and gear selections to make their pets as strong as possible.
Trial and Error:
One of the first things we tried internally was to have Zombie Dogs scale their Life directly with their owner's Life (Zombie Dogs already inherit Armor and Resistance from their owner). This had mixed results. For example, if the player stacked a large amount of Life, Armor, and Resist, it was possible to have Zombie Dogs tank most of Act I and parts of Act 2 in Inferno. As much as it made sense to have Zombie Dogs scale directly with your gear, it actually inhibited a completely different playstyle: players who wanted their witch doctor to be more of a glass cannon, but still have their Zombie Dogs able to tank. And with that we went back to the drawing board.
The next time around we gave the Zombie Dogs a base amount of Life, and in addition to this base value, they would also receive 35% of their owner's Life. So, you had a Zombie Dog that could scale with your gear, but if you were built as a glass cannon you’d still have that base amount of Life to fall back on. To help with general survivability, we also gave Zombie Dogs some innate passive Life regeneration. This test was much more successful. The Zombie Dogs could survive through most parts of Act I and Act II of Inferno just fine and died only occasionally to really difficult encounters. In Act III and Act IV, however, they could take maybe one or two hits, but the outcome was always the same: dead Zombie dogs. We tried increasing the bonus to 100% of their owner's Life -- and even to 150% at one point, just to see what would happen -- but it was to no avail. The incoming damage just scaled up too high in those later Acts.
So, we made some more adjustments to their scaling, we gave them more passive regeneration, and we made pets resistant to even more AoE effects (such as Plagued, Frozen, and Mortar). The result was positive, but not perfect: Zombie Dogs could now tough out Acts I and II of Inferno, but they were still melting in Acts III and IV.
The Final Product:
Our final iteration was to give Zombie Dogs their own version of the wizard skill Force Armor, which limits the amount of damage a wizard can take in a single hit up to 35% of their maximum Life. Much like the rationale for reducing damage for AoE effects, pets take more damage from melee than players. Pets also don't back off when they’re low, make use of doorways, or avoid big attacks.
When translating "Force Armor" over to Zombie Dogs, we wanted to make sure they could still scale with the player's Life, Armor, and Resistances. So, rather than a flat 35%, the damage cap per hit is based on inherited Armor and Resistance values, and rather than scaling with the total Life, the mitigation amount is calculated on the base health of the Zombie Dogs, allowing additional Life to actually scale exceptionally well.
This might be a bit confusing, so let's set up an example using a level 60 witch doctor. Let’s say this witch doctor has 32,000 Life, 45% mitigation from Armor, and 30% mitigation from Resistances. (For clarity, this means that 55% of incoming damage gets past the player’s Armor, and 70% of the incoming damage gets past Resistances.)
Once you factor in some passive Life regeneration and healing from health globes, Zombie Dogs can do reasonably well in Inferno. Players who decide to go with a glass cannon build while using Zombie Dogs will have pets that can tank for short periods of time. Meanwhile, players who build with some survivability and choose pet-oriented passives like Fierce Loyalty, Zombie Handler, and Jungle Fortitude will find their pets to be extremely durable and very capable of handling all Acts of Inferno.
As it stands now, without Vision Quest, many builds feel like you never have quite enough Mana.
Don't get me wrong: feeling like you always want more Mana can be a good thing, otherwise the resource system isn’t really doing its job. Even so, there are two major issues with Vision Quest that we want to address. The first is that it can feel very "feast or famine" when you're using it; sometimes you have near limitless Mana and at other times you're starved for resources. Second, and perhaps more importantly, it forces you to keep four skills on cooldown in order to be useful. This can be frustrating for a witch doctor who wants to use a cooldown skill strategically, but ends up casting the spell early for the Mana regeneration benefits.
Let’s use Big Bad Voodoo as an example. Big Bad Voodoo might be ready to go, but you need it on cooldown for Vision Quest to stay active. So, you cast the skill with only a handful of enemies on the screen. Then, no more than 20 seconds later, you come across a nasty Elite pack. While this would be a great moment to drop a Big Bad Voodoo to help you kill everything in sight (and ultimately avoid being killed yourself)….the skill is, of course, on cooldown. This can be a very aggravating experience! This isn't a dilemma we want players to face on a regular basis, so Vision Quest is getting redesigned for 1.0.4.
We're keeping the focus of the skill on Mana regeneration, but we're going to shift the way you get that regeneration away from needlessly spamming cooldowns to attacking and doing damage. The first thing we're doing is increasing the baseline Mana regeneration of all witch doctors from 20 Mana per second to 45 Mana per second. Not only does this help to alleviate the "feast or famine" effect, it also acts as a big buff to witch doctors who choose to skip Vision Quest. As for Vision Quest itself, it will increase Mana regeneration by 30% for 5 seconds after dealing damage with Firebomb, Corpse Spiders, Poison Dart, or Plague of Toads. One of the fun things about this set up is that you can combine it with a Spider Queen (Corpse Spider rune) or a Pyrogeist (Firebomb rune) and they’ll keep Vision Quest active for you the entire time they’re out.
Of course, Vision Quest going down to 30% can seem scary. Base Mana regeneration is increased, and the new mechanics actually allow for Vision Quest to have a very high uptime, but is it enough?
As we continue internal testing, one of our checks to determine how well Vision Quest is performing is to see if a level 60 player can still summon hordes of stampeding Zombie Bears. While we can't accommodate every skill and build combination out there, the goal for Vision Quest is that a player who has chosen the right passives and gear will still be able to summon waves of stampeding bears for at least a few seconds. The new Vision Quest is a lot less "feast or famine" than before, which means some players won't be able to spam Zombie Bears for quite as long, but the tradeoff is you’ll have a more consistent stream of Mana coming in, and (more importantly) you'll have your cooldown-controlled skills available to use strategically for maximum effect. A more reliable Mana stream, being able to use your cooldowns, and having the option to use other active and passive skills seems like a better design for the class as a whole for the long term.
In case you're wondering, we’re not touching Splinters or Zombie Bears this patch. While these are the two most popular witch doctor skills right now, it’s probably not just because people love the sound of Splinters or the look of Zombie Bears (though both of those are pretty cool). Instead, their popularity is likely due to how attractive these skills are, both in terms of damage output and overall feel. To help compensate and open up more build options, we’re buffing a lot of other skills to make them as appealing as Splinters and Zombie Bears.
Speaking of how a skill feels, the reason players avoid many of the lesser used witch doctor skills have more to do with the skill feeling "slow." For example, Firebomb, Plague of Toads, and Corpse Spiders all have animation timing issues which are being improved for 1.0.4. In general, all of these skills will cast faster, which will make the class feel snappier and more responsive. We're also doing a straight damage increase on many skills including (but not limited to) Acid Cloud, Firebats, Firebomb, and Spirit Barrage.
That wraps up the witch doctor, and all of our class previews! We hope you're excited about these changes and look forward to hearing your feedback.
Be sure to check out our other class previews for patch1.0.4:
Wyatt Cheng is a Senior Technical Game Designer for Diablo III. He doesn't always fold on a K-10 offsuit.