Why You Shouldn't Use Fracture

[Kitru]

So I've been doing some math and thinking while I was on suspension (I think they tagged a whole slew of us from the various tank discussion threads all at once, lol) and came to a pretty interesting conclusion: on any fight that lasts a decent period of time, using Fracture ends up being a net loss in total DPS.

The reason for this is pretty simple: Fracture is pretty expensive for a single target attacks given the damage it deals. It cost 80 TP, which is a net loss of 30 TP/GCD; at 300 potency per use, that's 10 potency per net TP cost. The BB combo (150 + 200 + 280; 70 + 60 + 60) is 15.75 (630 / (70 + 60 + 60 - 50 * 3)). On top of this, the value of that higher potency per GCD is diluted *very* heavily by the fact that, in order for it to actually provide that massive potency benefit, it can only be used once every 12 GCDs.

Here's some more math to chew on that gets a bit more involved.

The basic optimal rotation for a WAR is BB>BB>SE, which is 325.30 potency per GCD (math is in the DPS/enmity post in my sig) at a net cost of 14.44 TP/GCD ((580 - 50 * 9) / 9).

Adding in Fracture (one out of every 12 GCDs, which means that the rotation gets pretty long since that's once every 4 combos even though it's pretty simple) gets you an average potency per GCD formula that looks like this:

(((((8 * 630 + 3 * 300) / .9) + 4 * 610) / 39) * 1.0556 + 83.33 * 1.11) * 1.2 * 1.0476 * 1.0272 * .75 = 326.55 potency per GCD

The average net cost of this rotation is 15.64 TP/GCD ((70 * 12 + 60 * 12 + 60 * 8 + 70 * 4 + 80 * 3) / 39 - 50).

To get the Fracture rotation to reach resource parity with the non-Fracture rotation, you're going to have to include 1 empty GCD out of every 54, which means that you'd be dropping your DPS by ~1.8%.

So, all in all, you'd be dropping your DPS by nearly 2% to increase it by less than half of a percent. Fracture just really doesn't pull its own weight.

Considering that Fracture is stronger on WAR than on any other class, I'm reasonably confident that *no one* should be using Fracture at all. It's just not worth the cost over any fight of non-trivial duration (which, considering that it's a *DoT* and only effectively operates over non-trivial durations, it's kind of shooting itself in the foot).

[Phreak]

Correct me if I'm wrong fracture is even worse than in your proposed case because the damage over time cannot crit, so it's even less damage per TP.

[Sapphidia]

Interesting math. I posted a "why fracture sucks for paladins" post in a crossclass thread with similar findings, mostly based around the relatively low potency compared to halone comboing, and the high TP cost making it a loss in situations where you bottom out your TP (which admittedly is rarer for paladins).

The only thing to consider with Warrior though is that you've modelled your rotation based on a constant use of Fracture within the rotation... but that's not really the optimal way to use it. I agree with your findings, but for DoTs in this game you have to take into account Cooldowns.

When a DOT is applied it takes a snapshot of your damage stats and maintains that stat throughout the duration of the DoT. Fracture is a 300 potency attack for 80 TP for warriors, yes, which over sustained use is rather negligible... but what you need to factor in is using Fracture ONLY when you have dps cooldowns up. Pop it when Berserk, Unchained and Internal Release are up (we're assuming Maim will always be up for these tests). In these situations you'll get a relatively higher contribution to DPS for using Fracture at this point, and it SHOULD equate to a DPS gain. Suddenly it becomes a 360 or 420 potency attack with possibly boosted crit, and whilst the other attacks are also boosted, the larger difference should make the 10-20 extra TP worth it.

This is just napkinning of course. Fracture shouldn't be used without cooldowns up according to your maths, which I agree with. Add in a period of Berserk and Internal Release on cooldown and it -should- be a DPS gain. I hope.

[Terabyt3]

Take into account the SE and MAIM on fracture too?

I tend to go BB>BB>SE>Fracture>BB>BB>SE>Fracture (Obviously adding xtras in when needed but thats the general framework).

I couldn't follow your maths fully i must admit but are you taking into account crit rate on fractures initial hit? Maim buff afaik dots take a snapshot and apply that so fracture DoT with maim is more than without.

Also DoT from fracture generates Enmity each tick it appears so it's an extra constant boost to enmity generation ove those 30 seconds.

P.S. I got banned too. Seems someone in that thread didn't like us much. What can you do when mathematical fact offend peoples dream world?

[Zell]

What about the fact that some encounters require constant movement. If you apply fracture before the movement, you're still doing damage/threat where you wouldn't be otherwise.

Case in point, putting up fracture before Ifrit Jumps, or when garuda teleports. On titan it doesn't matter since you're stationary, but it could work just as well when Titan jumps as well.

There are a lot of mobile fights where if Fracture isn't up, you're not getting any threat. I understand that for stationary fights you have better use of TP in the rotation, but if i know a boss is going to move where I won't be hitting it for 5-7 seconds.... then Fracture is probably the BEST thing you can do at that moment.

[Sapphidia]

There are a lot of mobile fights where if Fracture isn't up, you're not getting any threat. I understand that for stationary fights you have better use of TP in the rotation, but if i know a boss is going to move where I won't be hitting it for 5-7 seconds.... then Fracture is probably the BEST thing you can do at that moment.

This is fallacious thinking. Just because a DoT ticks when you arent attacking something doesnt mean it does more threat than hitting a high threat attack and then doing nothing.

You really shouldn't use -threat- when talking about Fracture's use. It's a 300 potency attack that deals zero bonus enmity and has its potency spread over 30 seconds. Skull Sunder does 600 potency of threat if comboed, Butchers Block does 1400 potency worth of threat.

Threat in this game does not Decay. If Ifrit is about to jump away, and you have a single global cooldown to do something before he leaps, in THAT situation, a fracture would be better than a Heavy Swing... but if you've already done a heavy swing, you should hit Sunder and Butchers. Sunder does 600 potency of threat if it hits comboed, all at once. Hitting a Sunder and then having ifrit run away for 10 seconds doesnt do -less- threat than hitting a Fracture and having it tick when he's away. The 600 potency of threat has already been done. It wont be lost.

Fracture is only worth hitting when a boss jumps away if you don't care about threat and literally the ONLY move you can do otherwise is a Heavy Swing. THEN fracture is worth hitting for threat.

[FAVisceglia]

I never did any math and this is the first thread on the subject I've seen, but when tanking, I almost never use Fracture. Just from my experience, the use of a GCD and the loss of TP does not make sense given the paltry DPS and zero bonus threat that Fracture provides. Fracture, to me, always seemed like a throwaway skill, so I threw it away, more or less. I think the only way they could make Fracture interesting would be to add some debuff component to it (like reducing a target's resistance or something).

[Makeda]

I've pretty much moved Fracture over to one of those keybinds that's a little harder to press - the spot where I put things because I'm running out of buttons I can reach without moving my hands...

I have to wonder... is it useful at any point in the class? Is it like WoW's old school (pre-Cata) Rend? Something you get as a lowbie before they give you a real skill?

I still hit it from time to time when I'm in a "oh I haven't hit that button yet this fight" mood... but when I actually care about what I'm doing, I find it just doesn't occur to me to even slide my fingers over to that keybind (its on shift-7 at the moment, might move it even further out. Might even place it on the action bar I use for minions when not in dungeons... ).

[And why do these forums STILL think I'm only level 32, when I've been above 40 for 2 weeks now... :O]

EDIT: Maybe what it needs is what happened to Rend in WoW after Cata. It became something you could apply, that had a built in threat component, and other skills renewed it. You could have its timer reset anytime you re-applied "Storm's Path".
(Or Overpower could spread it - but then you're just making the FFXIV Warrior into the WoW warrior...)

[HiirNoivl]

I use Fracture when I'm about to hit Berserk Pacification.

Don't use it for anything else.

[Kitru]

Take into account the SE and MAIM on fracture too?

Yes. I applied the SE buff to everything except for the SE combo (the SE buff fades as you start up that third combo) and Maim to everything (because it would have 1 GCD crossover).

I tend to go BB>BB>SE>Fracture>BB>BB>SE>Fracture (Obviously adding xtras in when needed but thats the general framework).

The explicit rotation that I used was BB>BB>SE>Frac>BB>BB>SE>BB>Frac>BB>SE>BB>BB>Frac>SE. It maintains the highest uptime on Fracture without breaking combos. It also maintains the buffs/debuffs as previously mentioned.

I couldn't follow your maths fully i must admit but are you taking into account crit rate on fractures initial hit? Maim buff afaik dots take a snapshot and apply that so fracture DoT with maim is more than without.

Crit is a static benefit that applies to all damage so it's ignored as part of the potency per GCD construct (since each potency per GCD is affected equally by the given crit chance). The only "accounting" needing to be done for it was in calculating the comparative percentage gains provided by Internal Release over time (a 20% increased crit chance isn't a 10% increase in damage because you've already got a 20% crit chance as a WAR with Wrath V, which means that it actually provides a 9.1% increase in total damage during its duration). I actually assumed that all damage dealt that Fracture landed at the time of activation, rather than over the entire duration, which increases the contributions it provides by a small margin since the SE debuff wouldn't have a 100% uptime but I was expecting Fracture to benefit from it all the time (since it would always be activated while the debuff is present).

Also DoT from fracture generates Enmity each tick it appears so it's an extra constant boost to enmity generation ove those 30 seconds.

It's still 300 enmity potency per use, which is the same amount of damage. It's not 300 at activation and then an additional 200 over 30 seconds; it's 100 and then 200, and, since it's delayed over that 30 seconds, it's actually *worse* for enmity generation because it's not up front.

P.S. I got banned too. Seems someone in that thread didn't like us much. What can you do when mathematical fact offend peoples dream world?

My best guess is that the mods were noticing that the WAR discussions were getting kind of heated and just decided to suspend everyone involved appreciably in the debate. They might've done it to shut up everyone for the week before the Live Letter since it's likely the devs had decided to address that question a good while ago as kind of a heavy handed "we got it, you can shut up now, stop making waves".

Correct me if I'm wrong fracture is even worse than in your proposed case because the damage over time cannot crit, so it's even less damage per TP.

I would be surprised to learn that DoTs are either incapable of critting or don't have some form of crit benefit factored into their damage. You would essentially be turning crit into a worthless stat for the classes that have DoTs make up a substantial portion of their total damage (BRD and SMN spring to mind).

This is just napkinning of course. Fracture shouldn't be used without cooldowns up according to your maths, which I agree with. Add in a period of Berserk and Internal Release on cooldown and it -should- be a DPS gain. I hope.

Berserk + IR + Unchained still aren't going to be enough to overcome the fact that Fracture is inordinately expensive for a single target damage attack. All of those buffs improve every other attack just as much as Fracture. The only way that Fracture would end up being a cost effective option would be if you have a single GCD to make an attack, the target is going to be vulnerable to damage for the next 30 seconds (when a target goes Invuln, the remaining DoT ticks are wasted; I'm not certain, but I'm pretty confident that when Ifrit/Titan vanish to do their special attacks that they go invuln and don't take any damage even from DoTs on them at that time) and you'd be starting off a new combo: Heavy Swing has a cost per GCD of 7.5 TP per potency (the efficiency is quickly made up by the cost per GCDs of Skull Sunder and Butcher's Block being 20 and 28 TP per potency, respectively). The fact that it's so ridiculously situational and, *even then* provides, at best, a marginal increase to total DPS means that you can pretty easily ignore it completely with no noticeable effect.

If you create a scenario wherein Fracture is as effective as possible compared to the other options, you have to gauge how common that scenario is and then average out the higher contribution against the rarity of it. Recall that I said that constantly maintaining Fracture by weaving it in as the 13th attack each time amounts to a .38% increase in total potency per GCD. Yes, it's an improvement, but only semantically. You're never going to notice an increase that small. Hell, it's likely below the game's rounding error such that how the RNG decides to place you within the damage range of your weapon will make a large impact.

Basically, Fracture provides a marginal improvement *at best* while causing a noticeable increased drain to your resource pool. Using Fracture is, for all intents and purposes, electing to *not* use another of your attacks later on down the line and replace it with an empty GCD.