Dev Watercooler -- Critical Hits (And Misses)

‘Dev Watercooler’ is a blog series that provides an inside look into the thoughts and discussions happening within the World of Warcraft development team. In our first entry, Lead Systems Designer Greg "Ghostcrawler" Street laid down a few ground rules:
- No promises.
- Don’t read too much between the lines.
- No whining about the choice of topics we cover.
Critical History Lesson
In the original combat rules of World of Warcraft, melee classes could get 200% crits while casters could only get 150% crits. This was back when all the designers presumably played rogues instead of mages, which according to the forums is what we all play now (which makes our dungeon testing interesting, I gotta tell you.)
Over time, we added talents to allow various casters to get 200% crits as well. Warlocks “could” spend 5 points on the Ruin talent, for example, which you pretty much had to do to be a good warlock. As part of the Cataclysm talent tree evolutions we decided all DPS specs should be able to get 200% crits without investing talent points. There are still some inconsistencies though. Death knights can get 200% crits with both their melee and spell effects, while Assassination rogues get 200% crits with their physical attacks but only 150% crits with their poisons. Healers have always gotten 150% crits, both with their damage-dealing spells and with heals.
The overall design could be described as one that is simple to learn but complex to master. Or put another way, you know most of what you need to know if you’re told that crits do more damage. How much extra damage they do is one of those nuances that more experienced players learn over time and one of the things that makes classes feel different.
Or does it?
You could argue that we’re just keeping old rules that don’t really benefit the game. Is it very interesting that rogue poisons or Enhancement Lightning Bolts don’t have big crits? Does it make you feel different when you pick those classes or specs? Does it feel rewarding when you learn those subtle distinctions? I’d posit perhaps not. Homogenization is something we fight against all the time and one of the primary reasons that we don’t make class A’s ability work just like class B’s ability.
Homogenization -- A Dirty Word
If I can be snarky for a moment, players tend to beat the “homogenization!” drum too emphatically when they are losing something that is overpowered, and like to mock it as “flavor!” when we refuse to give them a cool ability that another class has.
Too much homogenization is a bad thing, no question. But do weird crit rules really fall into that category? There is a difference between being complex (which adds depth) and being complicated (which might just add confusion). We’d rather spend our “complexity points” on things that are really meaningful differences. Pick Assassination because you like daggers or poisons or maybe Rupture, not because you like small crits.
There are balance issues to consider too. Assassination rogues are never going to value crit as much as other characters are as long as some of their crits are smaller. We ran into the same issue with the damage-over-time-based specs when their dots couldn’t crit.
Healers Love Big Numbers Too
It can be an issue for healing as well. In Lich King, critical heals were virtually wasted because much of the time they were going to be overhealing. In Cataclysm, where healer mana matters more and even big heals can’t trivially top someone off, crits are more valuable. But they aren’t valuable enough. Getting 10% haste allows you to get a heal to a target 10% faster. Getting 10% crit allows you to heal a target 5% more. Is it any wonder that crit tends to get devalued for most healers? Resto shaman like it, but look at how many talents they have that make crits better for them. We’re strongly considering just letting all heals crit for double, just like most attacks. We don’t think this would have huge PvE consequences. Healers will heal for a little more, but even if they choose to start stacking crit, they’re going to do that at the expense of Haste, Mastery or Spirit. It could have bigger PvP consequences. Most PvP healers don’t have crit chances beyond say 10% or so, so they aren’t going to crit often.
We’ve been considering whether healing is too strong in PvP anyway. You may have noticed that we made the tooltips for Mortal Strike and equivalent debuffs intentionally vague for 4.1. As I write this, those debuffs are still at 10% healing, but we’re concerned that healing is too hard to counter and we might change that number. Changing it back to 50% would probably lead to the Mortal Strike debuff being mandatory for Arena comps again, but we never got much of a chance to see its effects at say 20%. A 20% Mortal Strike debuff could easily counter any excessive healing caused by 200% crits.
Changes Ahead?
Letting rogues and Enhancement shaman get 200% crits with non-physical damage would be a larger change, and not the kind of thing we would do mid-expansion. But it’s definitely something we’re considering for the future. That would only leave the damage spells cast by healers at the 150% crit range. We think we could make those full 200% crits as well. If we want to make sure the DPS specs still do a lot more damage, we have the knobs to do that. For example, we could buff passives such as Moonfury (the damage bonus for Balance druids) or Shadow Power (the damage bonus for Shadow priests) to make sure their spells still landed a lot harder than the healing specs did, even if the healers got big crits.
If we made all those changes, then any crit in the game would be at 200%. It would be a very simple rule, and I’d argue any loss of class distinction is more than made up for by the positive balance ramifications. As always with this blog series, this is just speculation. You’re more likely to see 200% healing crits sooner, but even that isn’t something we’ve fully embraced yet. It’s just the kind of thing we discuss when hanging out at the bar... er, I mean watercooler.
Greg “Ghostcrawler” Street is the lead systems designer for World of Warcraft. He crits on a 19 or 20.

Dragonmaw
Crits suffer from the same issue that all RNGs(random number generators) do - they're not really that random. Even with all of the normalization processes running under the hood these days I still see examples of horrible streak behavior all the time. One of the places this is extremely obvious to me lately is the ret paladin's talent Divine Purpose, which provides a 15% proc to make a few of my core abilities activate as if I had full Holy Power. Many times I go a fight (or 3) without a single proc and then suddenly in one fight I get 6 in a row. The end result is I end up sitting around waiting for Divine Purpose or any of my other procs to activate a great deal or on the other hand, frantically trying to use up all of the procs without totally shafting my rotation.
Crits are another victim of streak behavior and this can be really detrimental depending on your spec. Sure, rogues aren't as bothered by crits since a lot of their damage is poison which doesn't get as much benefit but what about holy paladins for example which require near constant crits to keep one of their buffs up. I can't list all of the examples of these types of abilities/talents but I'm sure there are others. Some trinkets are affected by this too. This is more of an issue of their frequency than their potency though.
As for the power of criticals, making them too strong adds more chance for powerful bursts which have an effect on both pve and pvp. Did I lose to that other player because he was better than me or because he happened to get a giant pile of critical damage that made the game think 2 of him were hitting me? Bringing all crits down to 150% would still make them powerful but would make them much less decisive wins through luck.
If you were looking for ways to keep classes from becoming homogenized, folding higher critical damage into specific abilities or talents would a way to go. The guaranteed crit of elemental shamans on lava burst comes to mind. This provides them controllable spike damage that is regulated by the cool down of the spell itself. If you wanted it to be an even more powerful ability you could add that extra 50% critical damage into that spell alone.
tl;dr version - Criticals should all be lowered to 150% to make them powerful but not overtly so. My game play experience should be based on my skill, not my luck. Crits are icing on the delicious WoW cake.
Nesingwary
Class skill that increases crit chance for a few seconds during a buff, i.e. deepfreeze, are skill based. It gives you a moment when you can sieze and opportunity when your opponent has made a mistake. Sort of like boxing, you kidney shot him and switch and throw a fat upper cut when he is dazed by a .5sec of sharp pain
Twisting Nether
Dragonmaw
Azuremyst
I don't want talents like them i just want elemental pvp to be balance with other pvp caster specs. Maybe something you SHOULD consider is giving elemental (and even resto) shamans a better 4 peice for pvp gear. 1.5 sec off the CD on grounding totem, is ok in 3's but with the 25 sec CD in 4.1 it will be useless.
Anvilmar
Steamwheedle Cartel
Drenden
Hellscream
Tanaris
Detheroc
Earthen Ring
Malorne
Magtheridon
Moon Guard
Eldre'Thalas
Bronzebeard
Azjol-Nerub
Blackrock
Anvilmar
Maiev
Blackrock
Cenarius
Ravenholdt
Cenarius
The Underbog
Tortheldrin
- No more then 9 unique abilities for each one of the 3 skill tree (specs.)
- An ability can be used only when you have points in certain skill tree.
- Skill trees could be combined to the certain degree but the number of abilities must never exeed 9.
- The use of one of an initial abilities would result in choice between the next several optional abilities.Once the chosen ability used, according to certain game situation, the option of using the next set of possible abilities would show up on screen.
- Once the ability is used, it would not show up in any set of optional abilities untill the cool down of that ability would expire.
By trying to imagine countless possible rotations of abilities' combinations and understanding the simple but yet powerful mechanic behind the mantioned above scheme, one could realize all the beneficial aspects of it.
Saurfang
Norgannon
Nesingwary
The Forgotten Coast
Frostwolf
Silvermoon
Jaedenar
Nesingwary
Stonemaul
Nagrand
Stonemaul
The Forgotten Coast
Nagrand