Concerning Grains: ToC25 Heroic (ToGC25)

And the lack thereof.

As I mentioned last post (wheee, a month ago), ToGC25 is very much not like Ulduar. And I mean that in a bad way; Hard-mode Northrend Beasts is punishing to a degree that makes it harder than the entirety of the normal instance, combined and tanked at once. And if you ignore the questionable logistics of that statement, you know it's otherwise true.

ToC25 is not hard in any way: the deaths we see on the fights are almost always due to disconnections (thanks Wintergrasp!) or people being so numbingly bored that they fail to notice that they're standing in fire with a rat gnawing on their head. And half the time, we just heal that person through it. In short, it is insultingly easy.

Then you get to ToGC25, and Gormok bites your tanks' nuts off. If he doesn't, Acidmaw and Dreadscale are sure to make up for it by killing your entire raid if they so much as twitch out of position.

Again talking to Aeman, he put my issue with the zone far better than I can: my issue with the zone is not its difficulty. It is ToGC25's startling lack of granularity in its difficulties.

For those playing the home game, by that I mean (er, Aeman means?) the degree to which a system - in this case ToC's difficulty curve - is divided into small parts or pieces (grains! tee hee!).

To illustrate, let's compare the difficulty curves of Ulduar and ToC using ever-so-awesome MS Paint graphics, shall we?

This is (roughly, so I don't have to graph 13 bosses) Ulduar:

Oh look, a nice curve, increasing as we go. It's designed the way games are supposed to be designed: increasing complexity and difficulty over time gives the player a chance to develop and adapt to the particular mechanics and themes of the content. It also allows them to partake of content early on, increasing their interest in finishing the content.

Yet here we have ToC+ToGC 25-man:

Note the spike: the closest I could get to representing the zone's irrepressible desire to gouge out your eyes with a grapefruit spoon. This curve is not, as I may humbly suggest we put it, "fun."

I love hard content. I do. Again, I don't mind having hardmodes that are simply beyond what our guild can do; I mind having zero content at all.

So what does this have to do with granularity - with gradual scaling of difficulty?

Well: Back in Ulduar, I knew I'd never see Algalon. It wasn't going to happen, and I was ok accepting that. The reason I could look at the fight and know it wasn't for me was because I could look at everything that stood between that fight and where I was and say "wow, that's a long way off, and this is already getting hard."

As players, our expectation is that the next challenge ahead of us has been designed to be doable in our current state, and that challenges down the road will require further growth - in this case, in gear, mechanics familiarity, and skill. In the MMORPG world, it is a natural extension of the idea of questing: you complete objectives in order, and are sent on to the next set of tasks and challenges. When we are sent on - often in a literal "go to zone B to see Person X" - we expect that we are judged to be sufficiently "ready."

This is how games and stories maintain continuity. And it's a reflection of reality, too: we lift little weights before big ones; we learn algebra before calculus; we learn to walk before we learn to run.

In raiding terms, ToC25-normal asks us to lift a 15 pound weight. Hardmode Beasts then asks us to bench 300 pounds. It is that disparity - and the implicit expectation of the designers that the one somehow leads naturally into the other - that so irks me.

Scaling Hardmodes

I was recently talking with Aeman, my fellow raid-leader with Ascended (and our fuzzy-bottom'd main tank), and he made an interesting point I found worth talking about a bit.

Namely, he mentioned to me offhand that there is a big difference between Ulduar hard-modes and CC25 hardmodes, and that as we work our way into CC25-Heroic, the trick is going to be getting people to really come to grips with that fact.

I couldn't agree more.

Well, actually, maybe I could - I've never really spelunked on down into that part of my mind to see how deep into absolute agreement I can go. Yes, that entire sentence was an excuse to reference spelunking.

Look at Ulduar hardmodes like those of XT or Ignis. As a guild, we had these reliably clear on 10-man maybe a few weeks after the zone came out - that's in t7.5 gear. In 25, they're perhaps a bit harder (given damage-scaling mechanics that we've talked about here before), but their overall difficulty comes out somewhere in the neighborhood of... Auriaya-normal?

Then Auriaya's own hardmode (which I hate with the fiery burning passion of a thousand suns) is roughly in the realm of Vezax or Yogg's normal mode. And from there, 25-hardmodes are just... silly. Yogg-0 and Algalon are as far beyond my ability (and interest) as nuclear chemistry is to a churchmouse.

Conversely, the very first CC25 hard mode - Northrend Beasts - requires a level of damage output and tank survivability on par with Yogg'Saron. That's straight to a difficulty approximately that of the hardest normal-mode boss this expansion. Owie.

On one hand, this seems fairly obvious. After all, every single CC hardmode is only unlocked AFTER the zone is clear on normal - Heroic Beasts is a further-in boss than Normal Anub.

But to raiders used to early hardmodes being like those of Flame Leviathan, XT, and Ignis, it's a punch in teeth. One of the things I have always noticed as a raidleader is that players respond fairly directly to the level of challenge they perceive. Something about knowing that you're looking at the last room in the zone kicks our trained gamer-brains over into "zomg final boss!" mode.

The tough thing is rebooting people into that peak-performance mode when they're looking at the very starting room of the zone again. No matter how many times you tell people "this fight is very hard, don't slack," there is a part of the brain going "it's just beasts..." (followed shortly by "ow ow ow why am I on fire? Why is a dirty wet rat biting my brain? Why am I dead?!")

Placeholders: may cause rash, severe ranting

So it's been a (long) while since I posted. So these things go. This is more my place to ramble to myself than anything else.

Today I want to talk about the Crusader's Colosseum.

Now, as my previous rants have indicated, I find the entire idea of the Argent Tournament a bit...well, pointless. But my conceptual objections aside, let's call the Crusader's Colosseum what it really is: a placeholder.

(*audience gasps in shock at my use of a dirty, dirty word in the MMO world*)

When Blizzard sent us WotLK (in a pretty, sparkly box), they told us it would have three raids: Naxx, Ulduar, and Icecrown. This claim generally amused us. Most of us old-timers pretty much knew they'd probably have to add at least one more raid just to have enough content to last till the next expansion.

Word on the street is that said fourth raid was originally planned to be Anub'arak's kingdom, Azjol'Nerub. But somewhere along the way, it got scrubbed - which is sad, as Azjol'nerub and Ahn'khet are two of the prettiest/spookiest zones Blizzard has ever made.

We can only guess at the reasons for said scrubbing, but the most likely is (in my opinion at least) that someone senior on the design team decided that 3 raids would be "enough," or that including AN as a raid would divert too much attention away from the Arthas/Titans storylines they were running in the two 78+ zones (Icecrown and Storm Peaks).

Honestly? I think it could have worked. But the snag the whole (surmised) plan seems to have hit is that they were simultaneously reducing the overall difficulty of their raids.

People blew through Naxxramas, OS, EoE, and even Ulduar (though to a lesser degree) like a tornado through a dust-bowl shantytown. With "WoW: Cataclysm" more than a year away, there was clearly some time that needed buying - via, of course, a 4th raid. At this point, I sincerely hope the previously over-ruled designers of the 4th raid got at least one dedicated staff meeting to dance around their lead designers chanting "toldja so toldja so" (that is, if they weren't already "let go" when their content got scrubbed).

Now they had a big problem: they couldn't just hustle down to Steve's office (let's call our hypothetical designer Steve) and ask him for copies of the development version of Azjol'Nerub. See, AN and AK had already been turned into 5-man leveling zones. Oops.

Stuck between a rock and a few million gamers, they needed content. As I see it, they sat down and brainstormed a few random boss ideas using existing mechanics and creature models, then shoehorned the mess into the Colosseum concept. Look, a raid!

Then, someone perceptive on the team notices that the entire zone is only about 2 small hops above being one big "go kill 8 Kobolds and bring me their nutsacks" quest - a little weak on the "memorability" and "players giving a shit" front. So they grab the Anub'arak model out of the old 25-man development raid and tack him onto the end. Throw in an Arthas cameo (remember, tying everything into the Arthas/Titans storylines is vital) and ship.

Thus, CC25 was born.

Now, let me make sure one thing is clear: I like CC as a raid. The fights may not be the best ever, but they're fun. There's no trash, the loot is good, and the hard modes are (yay!) actually pretty damned hard. This may be cardboard filler, but at least this particular cardboard (ahem) haz a flavor.

My issue with the zone (again, "tournament" objections aside) is that it literally has zero point. It's 4 nameless bosses and 1 boss that was ALREADY FREAKING DEAD. There isn't even a Mr. Spock lookalike to pop out of the woodwork and explain the whole thing away with time travel and Tachyons. He's just back. /sigh.

What's showing through here is that Blizzard is essentially the modern equivalent of a paid-by-the-word novelist. We've been fed a heaping spoonful of "50 page orc battle." And to be honest... I almost appreciate that they aren't trying too hard to dress it up as anything other than that. Anything else would be like throwing a few raisins into my oatmeal and pretending it makes it into a delicious breakfast (I'm looking at you here, oh mother-of-mine).

To be honest, I'll generally take what Blizzard puts on my gaming table even knowing that they're doing it to keep me p(l)aying. Sometimes, they serve me this crap. I accept it because in the past it's often been followed by tasty tasty bacon (oh yes, we're riding this analogy all the way out). But at the same time, I still don't like oatmeal.

3.2.2... what?

Ok, so it's been a while since I updated (work has been... insane. Yeah, that's the word I want).

I've been tanking and raiding as 2H-deep frost, using Glyph of Disease to keep two diseases refreshed, and it's been going pretty well. My threat is looking fairly solid - at least compared to what it was - and my survivability has gone up with gear, experience, and some re-gemming.

Then 3.2.2 shows up.

For Frost, there are a couple of major changes in this patch.

First, Rune Strike will now hit with both weapons. This is, quite simply, freaking sweettastic. As one of our hardest hitting and lowest cost abilities as tanks, rune strike should be keybound to trigger on every ability you have if possible. Doubling it makes all frost DW builds just that much better.

The bigger change for frost tanks is that Blizz really messed around with Unbreakable Armor

Prior to this patch, UA took the same amount of damage off all incoming hits - independent of the damage of the hit. Glyphed, you were looking at 1.4-2k damage less per hit. It could even fully absorb some hits.

Post-patch, in Uld10/ToC5&10 gear and glyphed, UA adds on the order of 6-9% reduction from armor. 9% of a 10k hit is only 900 damage. Add in the reduction from +25% to +10% strength, and you have a further reduction in both added parry chance and in added threat generation.

So why am I not freaking out at this apparent nerf?

Well, one, they cut its cooldown in half. That means it's benefits are available a lot more often.

Second, because the reduction from armor is dependent on how hard the attack itself hits, this ability becomes better the harder a boss hits. Sure, against something that hits for 10k, it's not great. But against something that hits for 20k (as some bosses now do), you're talking 1800 reduction - and more if you have a priest or trinket proc that adds some extra armor on top.

The only time this becomes "bad" is when it pushes you over into armor cap - or when your armor is so high that diminished returns cut the benefit to 4-5%. Then, this change is terrible. So this is really a mixed change in terms of benefit/pain for us.

I see why they would make such a change - frost is focusing on dual wield, which is spiky. Thus, adding a reliable CD of regular mitigation helps smooth it out. I can also see it as a great way to ensure a pull's safety - 9% added mitigation right off the bat helps you while you're still getting Blade Barrier active, and the 10% str is a nice early threat boost.

It's a great idea of a change, but for truly geared tanks, it is all but a useless tool in comparison to similar abilities in the Blood in Unholy trees - abilities whose health and reduction benefits are not subject to diminished returns.

I have my doubts, so we'll see. I may be wandering over into the unholy tree in the near future, depending on what I see. Blood... well, I'll talk about why I hate that tree later ;)

Screw the math, Jim!

I don't know who Jim is, but he should really do it.

So today's post is a bit of a mental continuation from yesterday - or rather, it's an explanation of why I've decided myself in the Dual-Wield v. 2Hander shenanigans.

I got here by screwing the math and getting back to super-basics. Let's take a trip there, shall we? Yes, let's. Bring juice, I'm thirsty.

~~

As I see it, tanks have two aspects to our job:

1) Poke the dragon in the eye a lot so it stays really pissed off at you
2) Don't die as a result of pissing off the dragon

Or, put in less weird terms, keep threat and don't die. Just about everything else is a part of or secondary to these tasks - much in the way that for a healer, everything is a part of or secondary to the primary goal of "Don't let the peoples fall down."

The thing about both aspects of tanking is that they go hand-in-hand: if you don't keep agro, it doesn't matter if you're alive; if you aren't alive, you very much do NOT have agro. Except maybe from your raidleader.

So when I look at dual-wielding, I see two things:

One, it's an obvious threat boost. Goal 1, better resolved.

Two, it gains a slightly higher level of general avoidance at the cost of frequent and unpredictable spikes of larger damage. So what we have to ask ourselves is: what is the status of goal 2?

Having sat on the other side of the healer-tank battery for a few hundred raids myself, I can say from my own experience that the hardest tank to heal is a tank taking random spikes of damage. It makes it impossible to anticipate when it is safe to heal others, or how much cushion there is - i.e. at what point the tank can take a survive another hit or two.

So while dual-wield may provide higher statistical avoidance in terms of raw stats, its parry count makes it very prone to bad luck. Mitigating bad luck is part and parcel of a tank's job.

When I think it over in my head, I simply cannot find any way that dual-wield, in its current (3.2 WLK) form, is viable in 25 man raids. End of story. I will be re-speccing on my lunch break today.

Also, this has been a nice example of "lessons learned in moving from 5-mans to raids."

Parry Clarifications (math involved)

So, it seems a bit o' clarification is in order following my last posting.

Aeman (my partner-in-crime for Raidleading and Ascended's Main Tank/Bear Tank extraordinaire) has noted that my description of "Parry Haste" and "Parry Gib" was a bit unclear. So, all credit to him for the following clarification:

Definitions and Mechanics

The base mechanic here is that when a boss parries you, the boss' next attack is hasted by about 40% of its remaining swing time (there is some disagreement). So if a boss has a swing timer of 2.8 seconds - very slow for a boss, I'll admit - their parry would haste them by up to 1.12 seconds, making the second hit occur 1.68 seconds after the first.

And yes, I had to use a calculator for that.

The Double-Parry "Gib"

One thing that makes this "scary" is that this mechanic can stack, up until the point at which this "would result in a reduction to less than 20% of the swing time remaining." (credit: WoWWiki).

Against our example boss, this would reduce his swing from 2.8 swing to 1.68 seconds, and then to 1.008 seconds. However, the odds of this happening are quite low indeed.

In my parse of last night's run, I was at something like a 3% chance to be parried by bosses (due to expertise). In, say, 2.8 seconds, I will typically do around 4 melee attacks and 1-2 specials. The rough odds of 2 or more of those being parried are essentially:

1-p(0 parries)-p(1 parry) =
1- [(6!/0!6!) * (0.03)^0 * (0.97)^6] - [(6!/1!5!) * (0.03)^1 * (0.97)^5] =
1- [1 * 1 * 0.832] - [6 * 0.03 * 0.859] =
0.014, or about 1.4%

(Many thanks to Bernoullius for helping me remember how the hell the binomial equation for probability works. High School math was waaaay too long ago ><)

1.4% is already an incredibly low chance - and remember, this risk is all a function of the number if swings you put in per each of the boss' swing. Against a boss with a 1.5 second attack timer, where you're likely to get in about 2 melee and 1 special attack, the odds of a double parry fall to a stunning 0.2%. It's just not happening.

Single-Parry Spike Damage

What is far more likely than a double-parry - and is in fact the main risk of dual-wield - is that with a higher attack speed, we are going to see a lot more parries per-fight. More attacks simply means more total parries (eg 3% of 100 attacks is clearly greater than 3% of 30 attacks). Period. No tricky probability math needed.

For example, not counting specials, in a 3 minute fight, a 3.6 speed weapon will strike the enemy 50 times - 1-2 of which will be parries. Compare this to dual-wielded 1.5 speed weapons, which will total a whopping 240 strikes in that same time - ~7 of which will be parries.

This is clearly a lot more "damage spike" events - a half dozen or more in a relatively short (3 min) fight - compared to maybe one or two.

So while the odds of a double-parry are incredibly low, we ARE clearly increasing the chances of a single parry from the boss at any given time.

Now let's be clear: this is NOT a "gib." After all, it is all but impossible for a tank to completely remove a boss' parry chance (it's something like 52ish expertise to do that). Every tank needs to be able to survive a parry-hasted attack from a boss as a simple fact of being a tank: they WILL happen.

What it is instead fair to say is that this makes DW considerably more prone to repeated "spikey" damage throughout a fight.

Clear(er) Conclusions?

It pains me to say it, but while the Parry-Gib as usually thought of (double-parries or multiple parried attacks in a row) may not really be a concern, dual-wield is definitely prone to taking a lot more spikes of damage than with a single weapon.

Now, if you can survive this sort of damage, it's not really as much of a concern. But Blizzard has steadily been reducing our health and our armor as tanks - and if you're where I am (Uld10 and H:ToC5 gear, each of these parry-hasted attacks hurts quite a bit if it hits.

So right now I'm torn. Having taken a big spike off Runemaster Molgeim last night (25% damage buff + a parry hasted attack = dead me), I'm a bit concerned by the number of parries I can mathematically take in a fight. I definitely got feedback from healers that I was spikey on damage taken - but it was ALSO from Molgeim, who is already one of the kings of spiking damage onto tanks.

I'm pretty much our first DK tank in a long time, and I don't want to be carrying around the "squishy" or "needs extra watching" labels with our healers, either for myself or for the class.

It's clear to me that dual-wield makes for higher overall avoidance, as I said in my last post. 4% raw avoidance from runes on the weapons, in addition to tanking stats on them, makes for that. But I'm definitely feeling the risk of short-term, high damage events right now.

So what I think it all comes down to for me right now is whether or not the dual-wield is actually necessary for my threat generation. I need to do some more testing to this end - but it certainly seems that in the long run, I'll be moving away from dual-wielding.

Dual Wield: Trade-off, Trade in?

Ever since Blizzard announced Death Knights - and that they would a) tank, and b) not use shields - the prospect of dual-wield tanking has been a hotly bandied-about topic on the various DK forums. Desiring to keep my shoes clean of muck and my eyes clean of pointless vitriol, I have completely and utterly avoided these discussions.

2hander tanking has worked just fine for me since I rolled my cute little gnome, and so I've stuck with it. But with the addition of Threat of Thassarian in the frost tree as of Patch 3.2, and suddenly finding myself in possession of a pair of Peacekeeper's Blades, I've been re-examining the possibilities.

So let's talk advantages, then disadvantages - then watch me derive conclusions for the benefit of all! (There will be cake and punch afterwards in the cafeteria).

ADVANTAGES

First, the talent. Threat of Thassarian (for those too lazy to click links) is a 3-point talent providing a 33/66/100% chance that Blood Strike, Plague Strike, Obliterate, and Frost Strike will strike a second time using your offhand weapon.

Simply put, this is the talent that has made Frost-spec DPS viable. To cut out all the math, it is a significant overall increase to the damage caused by any strike it works with, and has the added benefit that it doubles the chance for Rime to proc off of Obliterate (given it being chance-per strike and double the total strikes).

Tanking-wise, this does two things. One, it increases your total damage, increasing your threat generation. Two, it also increases Rime procs, increasing your AOE threat potential by a good deal.

In some fast testing last night, I found that given weapons of approximately the same quality (iLVL 219) and appropriate specs for each option, dual-wield provided about roughly a 300-350 DPS increase while tanking - an increase that would be amplified in threat terms.

And that's given that tanking weapons are relatively fast, which is less than ideal for dual-wield (the reason for that is another topic for another time).

Second, dual-wield allows you to rune two blades with the 2% parry chance rune, giving you a raw 4% bonus to your avoidance. And as tanking blades tend to carry far more tasty stats like Parry rating, Dodge rating, and expertise than most 2-handed weapons, you'll get an avoidance bonus there. For a spell-damage boss, you could even re-rune with dual 2% spell reduction runes (sure to make your healers happy).

DISADVANTAGES

What kept most people (but not all) from being dual-wielding tanks was the dreaded spectre of the "parry gib." For those (wisely) not keen on looking that term up on the forums which propagate its use, it simply expresses that a dual-wielding tank is much more likely to be parried by the boss. As every parry "hastes" the next attack by the parrying creature, being parried essentially increases the attack rate of a boss to some degree.

"Parry Gib" expresses the concept that: a) bosses like to hit for totals in the 5-digit range, and b) you are attacking quite quickly when dual-wielding, there is the "chance" that successive attacks would result in multiple, quick, high-damage hits (effectively "gibbing" you in the space of a second or two).

I find this concept... overly dramatic. What we must keep in mind is that bosses have a relatively low parry chance, so the odds of seeing two parries back-to-back - much less three or more - are quite low. Furthermore, at expertise cap, parry chance is even lower (don't ask me for numbers offhand, I don't have them).

Given that it's hard to swing a claymore without smacking into a plate item with expertise on it these days, expertise cap is very managable indeed. So the "parry gib" is essentially a non-entity. I hereby authorize you to smack anyone that talks about it in a "neener neener, you're gonna die" voice with whatever you have handy.

That said, dual-wield does have its issues. It's very hard to pull off without the right gear - it relies heavily on expertise and hit alike to land strikes and to avoid increasing a boss' damage. In addition, the absence of Rune of the Stoneskin Gargoyle means you'll have to make up some lost defense rating to stay over the 540 "cap." That's a lot of yellow gems - so expect your stamina to come down a bit as a result.

CONCLUSIONS

Dual-wield DK tanking is definitely a viable option, and it's something I believe I'll be doing as my preference here in the future. But if you're thinking about trying it out, keep in mind that this is a very gear-reliant option. Until you can expertise cap (26) AND def cap (540) without nuking the bejeezus out of your gear and stats, it's probably not worth it.

Fortunately, the ToC-5man gear is ideally suited to this spec, so hunt there for upgrades and give it a try!