Warriors and a Barbarians Fury

#1 - July 19, 2012, 2:59 a.m.
Blizzard Post
So basically, instead of tieing warrior rage to white hits and stats, what about having attacks that generate rage and attacks that spend rage. This way the player doesn't have to stand there and wait until something happens, he/she can make stuff happen when he/she is ready. Rage isn't something you manage, like energy or mana, its something you build up until you're ready to unleash its power onto your enemies with a swing of your axe, a thrust of a sword or a block with the shield.

Crit equals harder hits, haste allows more DOT ticks and faster autoattacks (sounds weird when you're pushing buttons as much as possible, but I'm open to suggestions for haste).

(Matthew Rossi suggested it on WoW Insider: http://wow.joystiq.com/2012/05/19/the-care-and-feeding-of-warriors-the-fury-of-diablo-3/)
Forum Avatar
Game Designer
#4 - July 19, 2012, 3:10 p.m.
Blizzard Post
Rage isn't something you manage, like energy or mana, its something you build up until you're ready to unleash its power onto your enemies...


So less like energy or mana and more like... combo points? :)

I'm not just being snarky. I think it's a legitimate point. If you have yellow attacks that produce X rage and then a finisher that consumes Y rage, then you have a system that works fundamentally a lot like combo points or Holy Power. Ultimately we know that system works, but I'm not sure that means that every class should use it. It's a little deflating when you switch characters only to discover that the resource mechanic feels almost the same.

I actually think we're pretty close on the rage design for Mists and that what some of you are reacting to isn't the mechanic as a whole but the current numbers on beta. By moving some rage generation to yellow attacks we have given warriors and bears more control over building up their resource rather than it being totally passive. By reducing some of the exponential scaling of the resource model with gear, we think we have broken the cycle of warriors who are weak at launch and overpowered by an expansion's end. Recently, we felt like rage generation was too high in starting gear, which wasn't giving warriors too much room to grow. We lowered it, but that ended up feeling too punishing at low level, so we're trying to strike a balance.

These aren't the kinds of decisions that can easily be mathed out, because so much of it comes down to the *feel* of combat, which we can only get a handle on by trying different designs and playing them ourselves, and asking other players (like you guys) to try them as well. Blizzard is famous, perhaps infamous, for design iteration. You sort of have to stand back and let us do our jobs here. We need to be able to adjust things in beta without the community freaking out every time a number goes down (and I'm not just talking to warriors here). I totally understand that beta design can feel like a roller coaster, which is one of the reasons we try to limit design churn after launch. It's totally fine -- and in fact we welcome it -- to provide feedback along the way. Helpful feedback is the sort of thing several warriors did when they said they felt like rage generation from white attacks was too low and rage generation during Enrage was too high. It's a little disappointing to read "rage generation feels low now, therefore the entire system must be scrapped." :(

I could wax philosophical about the similarities and differences between warrior rage and barbarian fury; indeed the strengths and weaknesses of various resource systems is something the WoW designers discuss frequently with the Diablo designers. Just remember that the combat situations are very different. D3 is about very short fights against huge groups of enemies when you are often alone or in a small group. Even if you have a tank, you don't really have a healer. Except for the boss fights, the fights are unscripted and unpredictable. I'm jealous of the combat situations the D3 designers can put you in, because there is room for a lot of different AE attacks and survival buttons and movement abilities. On the other hand, because the fights are short, there isn't a lot of opportunity for the rotations and situational abilities that the WoW classes have and I've always enjoyed the emphasis on the synergy of large groups of players in WoW (whether it's a raid or a BG or just a guild hanging out in chat). It's actually a little refreshing both as a player and a designer that they do feel so different (to me anyway).
Forum Avatar
Game Designer
#29 - July 21, 2012, midnight
Blizzard Post
I think there's a certain amount of rage/sec that make warriors feel good, lets call it X. Too much, and the mechanic stops working as a limitation. Too little, and you feel starved and it's no fun. The problem is that X isn't different to a maxed out warrior at the end of the expansion or a beging warrior in greens, but by tying it to gear progression you force either the too much end of the scale on the endexpo warrior or the starved feel on the questing warrior.


I agree with that, but here's the thing: we don't want warriors to scale linearly. What I mean by that is that as you increase gear, you'll be hitting harder and critting more often. Both of those are fun, don't get me wrong, but after you get over the thrill of BIG NUMBERS you start to see behind the curtain a little. Fundamentally, there isn't a huge difference between hitting a bad guy for 3000 or 30,000, especially when the bad guy's health has probably increased by ten as well.

What feels cooler is if you actually feel different with great gear. Instead of just hitting harder, you are hitting different buttons. For warriors, that means having enough rage to use Heroic Strike frequently instead of rarely. But that means we need to start warriors in quest greens at the "rarely" side of the dial, so that when you're in epics in the final tier, you're now at "frequently."

Historically, what has happened is that warriors start at "never" and then grow into "always." Rage scaling with gear was too good. It needs to scale, and a little bit of exponential scaling is ideal, but the end of the curve can't rocket up towards the sun.