Tuesday, April 13, 2010

Hasted DoTs

So in Cataclysm Haste is going to affect DoTs. This, developers say, is necessary to avoid having Haste be much better for some class than others. And they are completely right about that. Having Haste not affect half the damage of some classes means that it is a wildly swingy stat. Stats will always have some variability in usefulness for different classes, but having them relatively close together is useful. It is how Haste will affect DoTs that is the problem.

Right now Haste can affect a handful of key DoTs in the game through talents or glyphs. Haste simply shortens the total duration of the DoT. So your DoT used to tick 5 times in 15 seconds and you have 25% haste, now it ticks 5 times in 12 seconds. At some point there was a blue post that said that haste was sort of a boon but also in a way a drawback on DoTs because you ended up casting them more. I sort of shrugged that off at the time - after all, I can't get myself upset every time someone says something that is wrong on the internet - but apparently they had a plan to put their wrong-headed thinking into action.

Suppose you have a 15 second DoT (3 second period) and a 1.5 second nuke (I choose these numbers so everything rounds well, obviously). You cast your DoT, then you cast 9 nukes, then you cast your DoT again. Your DoT ticks for 2k and your nuke hits for 5k. So in 15 seconds you do 5 DoT ticks and 9 nuke hits for 55k damage.

Now you get 25% haste. If this shortens the length of the DoT then you cast the DoT every 12 seconds. How many nukes do you cast in the intervening time? Well, they now take 1.2 seconds. With the GCD to cast the DoT you cast you have 10.8 seconds left. 10.8/1.2 = 9. So you cast nine nukes, the same number as you did without haste. So you do your 55k damage in 12 seconds instead of 15. That means you do 55k/12 = 4.58k dps instead of 55k/15 = 3.66k dps. So 25% haste gives you a 4.58/3.66 - 1 = 25% damage boost. Yes, 25% haste, 25% more damage.

They are saying that they want DoTs to tick more often but have the same duration when you get haste. In that case, the 15 second DoT would still last 15 seconds but would tick every 2.4 seconds, getting an extra tick in there. They are also saying that reapply a DoT will extend the duration without resetting the tick timer, which means you probably want to reapply before it falls off1.

So now you cast your DoT and then you'll want to cast it again when the last tick happens before the 15 seconds elapses. So you cast the DoT, wait for the 1.2 second GCD, cast 11 nukes in 13.2 seconds leaving you 14.4 seconds into the DoT duration. You don't have time to cast another nuke so you reapply the DoT. Now you get 6 DoT ticks and 11 nukes in 14.4 seconds. That's 67k/14.4 sec = 4.66k dps. Under this system a 25% haste buff gives a 4.66/3.66 - 1 = 26.8% damage boost.

Of course that is a very simple example. I ran a simulation for shadow priests using my own current average damage numbers for each spell to get the correct ratios between them. With the current system 25% haste gives just over a 25% damage boost while 50% gives just under 50% damage boost over a 5 minute run (difference is accounted for by clipping). With the new system, 25% haste would give a 28% damage boost and 50% haste would give a 54% damage boost.

It's not really that big a difference, but on principle it is crazy. DoT classes long suffered from getting little to no benefit from Haste rating. Now they will get a better benefit than others. Why go with this system instead of with the system that makes the benefit identical? I assume its because they really think that casting DoTs more often puts them in a disadvantage.

Guess what, Arcane mages with more haste cast Arcane Blast more often. That's what haste does.



1. In the example given, choosing not to refresh the DoT would give 6 DoT ticks and 12 nukes every 15.6 second for 72k/15.6 = 4.61k dps, a very slight dps loss over the alternative and refreshing before the DoT falls off. This is because if you refresh before the DoT falls off you get a tick every 2.4 seconds, whereas if you let it fall off you "waste" the last 0.6 seconds of each cast.

Thursday, April 8, 2010

Fast, Powerful Heals

You should be able to use your efficient spell virtually forever (or at least much longer than the average encounter length). That efficient heal may be fine for earlier dungeons or raids. Eventually though you're going to hit points where someone is going to die in between casts of the efficient heal. In those cases you should switch to an inefficient but fast or big heal. Yet, if you focus on the inefficient heals too much, you will run out of mana.

As players gain more and more regen in later tiers of content, they'll need to rely on their inefficient heals more, but will also have the stats to be able to do so.

Not every healer has enough spells to currently make the decisions between efficiency and other choices. In fact, we'll probably have to adjust the toolbox of everyone to some extent.

[...] Most healers spam one heal now. Druids for all of their spells can do very well with just Rejuv and Wild Growth. Disc priests spam PW:S and Holy priests spam CoH on cooldown. We want to carve off niches for other spells. Flash Heal can be awesome if it runs you out of mana to hit nothing but Flash Heal.

[...] Casting a couple of fast / expensive heals to save someone's life is exactly what you should be doing. Casting nothing but the fast / expensive heal because, hey why not, leads to pretty repetitive gameplay.

[...] We will have to change numbers for Cataclysm. It's probably a safe assumption that we'll adjust every number in the game in order to account for changes such as larger health pools, lower mana regen and all of the talent tree changes.


This sounds great... kind of. What I mean is that it sounds great provided that they realize that in many cases the fast, powerful spells we spam these days are *also* the most, or among the most, mana efficient spells in our arsenals. I've gone over PW:S before, but when they pull out the examples of Rejuvenation and Wild Growth it gives me a facial tick. I think they know - I really hope they know - that druids use these spells almost exclusively because they are better than other spells in nearly every conceivable way, not because they have too much mana.

Levelling Up

If you've played WoW for five years you know that while the game has held onto a lot of very important thematic elements the difference in game play between now and the original release is staggering. Even just over the last year since the release of Wrath of the Lich King the game has undergone some really significant revisions. Of course revisions are always made with the endgame in mind, but Blizzard has not skimped on those revisions that are targeted towards the leveling and questing experience.

When I leveled my first character, a druid, the best way I could find to beat enemies was Moonfire and auto-attack with my staff. Cat form, all abilities included, seemed to do less damage than just swinging whatever stick you had. Attack spells were costly compared to my mana regen and didn't really hit that hard (it was almost impossible to get any spell power gear). Moonfire and swing beat an enemy in about 20-30 seconds, and that was the best I could do.

It's a real testament to how bad the leveling process was for druids that when I later switched to a warrior I leveled as protection and was impressed by how easy it was. And these were the days when Battle Shout cost 20 and parries and dodges generated no rage (*way* before devastate).

Blizzard at some point decided that you should be able to run around and beat monsters, complete quests and go up levels pretty much no matter what class you pick and no matter what talent tree you go up. There are still a few trees that are obviously sub-par, but not as sub-par as the leveling options that many classes used to have.

But despite the efforts that have been made, a lot of classes need a serious overhaul for leveling, and it has a lot to do with talented spells and new abilities. By "new" I don't necessarily mean all that new, I just mean abilities that have been introduced since Classic. Talented abilities and these new abilities are generally not available at low level, and that means that abilities that you are supposed to have to make your character do things are abilities you might not get until level 40, 50, 60, 70 or 75.

Go make a paladin. So you cast a spell that last for 30 minutes then you auto-attack. Okay, no big deal, its just level one. Go train your level one ability, it's an aura you put on when you auto-attack. Well, still level one. What does level four hold? Judgement, a spell you actually cast, and you can cast it once every 10 seconds! So generally you cast it when you engage and enemy then you auto-attack until the enemy is dead. You also get blessing of might which makes you do more damage when you auto-attack. Level ten? Put up your buffs and auto-attack, though you can stun someone every minute if you like, but it doesn't do any damage. All through to level 20 you get a better selection of auras and blessings to have on yourself when you auto-attack. At level 20 you finally get another spell you can actually cast to defeat your enemies. It has an 8 second cooldown. It also costs so much mana that given the fact you have no mana regen, you will have to drink every three monsters if you actually try to use it, and it probably doesn't do a lot more damage then you normally do anyway. Probably best to auto-attack. At level 40 you can finally put a point into a talent that gives you mana back when you judge, so you have mana to use a spell other than judgement, though it's really level 42 before you can max it out and use it effectively. At level 45 you've put two points into another talent that will let you use Exorcism to hit your enemies, although only on an on-crit proc with your bad crit chance, and only ever 15 seconds even if you are lucky. By this point you've got a +15% bonus and a +10% bonus to the damage of Crusader Strike, an ability you don't even have yet. At level 50 you finally get Crusader Strike, and you are finally actually playing a retribution paladin. You hit buttons when you fight your enemies instead of just auto-attacking.

Blizzard wasn't happy that all elemental shamans did was cast lightning bolt so they gave them another spell to mix things up a bit. It's called lava burst and you get it at level 75. What about shamans who want to level up as elemental? They can cast lightning bolt until they are 75. Arcane mages basically aren't even playing until they get Arcane Blast at level 64.

Strangely enough, for hunters they actually got it right. They give hunters their mana regen aspect at level 20, despite the fact that it was added to the game in the expansion. They also lowered the cost of shots so hunters can afford to do more than auto-attack until they get steady shot at level 50.

There are actually easy solutions to all of the problems I've given here. In the case of mages, shamans and various other classes that get now-pivotal abilities late in life, just make more ranks of the ability and give it out sooner. Sure, you introduced in it an expansion, but that doesn't mean it has to be forever above the pre-expansion max level. For paladins the fix I often talk about with a paladin wielding friend of mine is to give paladins crusader strike at level 20, and to make it cost nothing. Is that somehow going to cause problems? Are holy paladins going to be broken because they can swing their terrible damage to spell power trade-off fast one handers for 200 by wasting a GCD? Having the kay abilities of a tree embodied by it's talents is fine, but when you hit level 10 and get your first talent point, you shouldn't look at a tree and say, "Well, half of this does nothing until I put the 51st point in." And no one should be auto-attacking their way to level 40, even if they get nice buffs to do it with.

Tuesday, April 6, 2010

Titan's Grip

Titan's Grip really bothers me. Warrior's dual-wielding two-handed weapons is just plain great, so the concept is fantastic. The fact that a 51 point talent has a flat 10% penalty to damage, however, is not so great. For a new player reading the talent, they probably blink a couple of times and try to figure out if it means what it says. After all, it takes a lot of math to figure out how much better it is to use two-handed weapons instead of one-handed weapons when you dual-wield, and it isn't obvious that doing so is a greater than 10% damage boost1. Admittedly, how much extra damage you would get from dual-wielding two-handers would be just as mysterious if you took away the 10% damage penalty, but having a talent that does unknown extra damage is just way less confusing than having a talent that does unknown-10 extra damage.

Of course they have a pretty good justification for why Titan's Grip can't be simple and elegant and exclude any penalty. It would just be too much of a damage boost for a single point. When talent points are too good, the developers rightly point out, it makes things worse. Particularly for talents that are deep in talent trees. Basically if they balanced Fury Warriors correctly then you'd end up with every other talent in the tree being significantly worse than average to make up for Titan's Grip. Level 59 warriors would balk at being Fury because they can't get the talent that makes the tree work.

Okay, fine.

Mind Flay, Shadowform, Vampiric Touch, Starfall, Dual Wield. That is a list of talents that I am sure are better than Titan's Grip without a 10% penalty. Three of them are from the same tree! Running a weekly Sartharion kill last week a balance druid in the group won the damage meters and 30% of his damage came from Starfall. 30% for an instant cast he hit every minute. Maybe we can think of Mind Flay and Vampiric Touch as different because they are extra abilities, and so they are more interesting than a large flat damage boost, but Shadowform is probably the most ridiculous single point damage boost ever conceived. A straight 15% damage boost plus enabling both your haste and your crit to have more than just a marginal effect on your damage. I would be extremely surprised if I was wrong about Dual Wield for enhancement shaman. Developers have directly stated that they have no intention of supporting two-handed weapon enhancement shaman as a viable spec, and acknowledge that it is not at all viable as it is.

But none of these talents have a counterbalance. Dual Wield doesn't come with a damage penalty because otherwise it would be too good.

I don't think that Crusader Strike, Stormstrike, Divine Storm or Bloodthirst are better points than Titan's Grip with no penalty, but they are in the running. When you get a point that basically lets you make an instant attack when you would otherwise be standing doing nothing it's a pretty big deal.

Yes, talent points that are too good cause something of a problem. But there are lots of talent points that are too good, Titan's Grip has been a constant obsession of some developer. Someone out there really, really doesn't want Titan's Grip to be too overpowered, and that obsession over Titan's Grip leads to an ugly, clunky fix, while so many other overpowered talents are left to do their thing unimpeded. In fact, shadowform just gets better and better. Every few patches they think of another 10-15% global damage boost to tack onto it for free.

There are other solutions to overpowered talents. One is to use the prerequisite system that they have built into their talent trees that currently serves no apparent purpose. By making an intentionally weak talent a prerequisite to a known overpowered talent, you balance out the point spent on it. The 46-50 points in Fury could be something useful that doesn't increase or only barely increases your damage. People would say, "Why do I have to take this lousy talent to get Titan's Grip?" and then developers could freely admit, "Because Titan's Grip is so good it needed to cost six points." If shadowform cost 11 points then it would make a lot more sense.



1. I mean that if you hit level 10 the first time you played the game and looked at your top tier talents, it would be fair for you to think "Is that even good?", not that there is any lack of consensus among those that know about the game that the answer is, "Yes, that is very good."