Why You Shouldn't Use Fracture

[Kitru]

950 / 14.44 = 72?

Yeah, was misremembering a Berserk adjusted time that I did a while back rather than doing the math like I should have. If you Berserk adjust (only consuming for 16 out of 18 GCDs) the Non-Fracture rotation, you get a 7.28 net TP loss/GCD and 8.34 for the Fracture Rotation, which gives you 130.5 and 113.9 GCD until you run out. With those, Non-Frac won't ever actually overcome Fracture like you said, though the difference is nowhere *near* impressive or even remotely impressive (.27% increase in total damage). At 192 GCDs (8 minutes), the difference between GCD and non-GCD is .62% ((114 * 103.3 + (192 - 114) * 78.7) / (130 * 100 + (192 - 130) * 77.5)).

Of course, for all of this, we're ignoring the fact that it clashes with buff durations causing either itself or one of your other high damage attacks to not benefit from SE. I've been *inflating* the contributions of Fracture by a small margin for the entire course of this discussion. At the marginal improvements that we're getting to, that inflation is going to actually have an impact. The difference might not be as stark as I originally conceived it to be, but the fact remains that Fracture doesn't even get close to making a noticeable impact upon anything except for your TP costs (unless you're so good that you can notice differences of fractions of a percent). Rounding error on a parser, RNG variation, and lag variability are going to have a greater impact than Fracture on total DPS.

Additionally, this just makes no sense. 950 is a purely arbitrary number.

950 isn't arbitrary. As I explained before, it's the point wherein you absolutely have to start hitting empty GCDs because none of your abilities have enough TP. The actual number is slightly higher than that thanks to the average cost being in the 60s, but 950 is the point where your total net consumption prevents you from using *any* of your attacks and forces you to wait to get more TP. It's not arbitrary. You're almost never going to actually bottom out your TP when you've got empty GCDs; you're going to get empty GCDs when you've just got too little which is where the 950 comes from for the reasons previously mentioned.

[Coramac]

950 isn't arbitrary. As I explained before, it's the point wherein you absolutely have to start hitting empty GCDs because none of your abilities have enough TP. The actual number is slightly higher than that thanks to the average cost being in the 60s, but 950 is the point where your total net consumption prevents you from using *any* of your attacks and forces you to wait to get more TP. It's not arbitrary. You're almost never going to actually bottom out your TP when you've got empty GCDs; you're going to get empty GCDs when you've just got too little which is where the 950 comes from for the reasons previously mentioned.

It is completely arbitrary as you've been using averages for everything up to this point and suddenly don't want to use averages. Additionally, it doesn't necessarily stop at 50. You need to run out a cycle to determine where it does stop. If you have 60 TP and the last ability usage is a 60 TP ability, it stops at 0. If you have 60 TP left, and the ability costs 70, it stops there. Furthermore, there is no net consumption. It is a fictitious creation you've conjured up to make it easier to work with numbers. TP is 60 / 3 seconds. You can't net a real cost against an income that hasn't been realized.

Just stop, Fracture is worth using. Stop changing and twisting your reasoning to support the exclusion of Fracture.

[Kraze]

Just stop, Fracture is worth using. Stop changing and twisting your reasoning to support the exclusion of Fracture.

Thank you. Stop raiding the forums, and go raid the game. 332 insignificant posts is scary.

[Kitru]

It is completely arbitrary as you've been using averages for everything up to this point and suddenly don't want to use averages.

And it's effectively an average as well. It's the point where you stop being able to use attack. I could use 1000-(mean TP cost), but it doesn't really change anything: you get 129.5 GCDs for Non-Frac and 112.9 for Frac. You'll still have Fracture maintaining a 2.3% lead on damage for the first 113 GCDs that fades to nothingness at 130 (a .08% advantage) and proceeds to remain functionally nonexistent thanks to the dilution of the previous 130 GCDs (a .48% lead at 8 minutes).

Stop changing and twisting your reasoning to support the exclusion of Fracture.

I haven't changed my reasoning at all. My reasoning the entire time is that the TP cost on Fracture makes it an attack that you don't want to use because it doesn't add enough damage to your attack rotation to make it justifiable. I've refined the math, but the reasoning has remained the entire time.

[Vmage]

And it's effectively an average as well. It's the point where you stop being able to use attack. I could use 1000-(mean TP cost), but it doesn't really change anything: you get 129.5 GCDs for Non-Frac and 112.9 for Frac. You'll still have Fracture maintaining a 2.3% lead on damage for the first 113 GCDs that fades to nothingness at 130 (a .08% advantage) and proceeds to remain functionally nonexistent thanks to the dilution of the previous 130 GCDs (a .48% lead at 8 minutes).

I haven't changed my reasoning at all. My reasoning the entire time is that the TP cost on Fracture makes it an attack that you don't want to use because it doesn't add enough damage to your attack rotation to make it justifiable. I've refined the math, but the reasoning has remained the entire time.

which encounters are you referencing here?

I mean.. on any of the primals I have the excess TP and enmity lead to throw in fractures

same goes for any of the turns in coil..

I agree 100% that you shouldn't use fracture if you're in any danger of running yourself out of TP as it's a dps loss at that point... but when exactly does that happen ?

[Kitru]

I mean.. on any of the primals I have the excess TP and enmity lead to throw in fractures

same goes for any of the turns in coil..

Even on good Titan runs, I'll end up bottoming out TP when Titan is about 15-30%. On Caduceus, I'll run out during the split when my snake is at 20-30%. Ifrit give long enough empty time frames (Ifrit especially) that you're never going to run out of TP, but, if your DPS is mediocre on Garuda, the same can happen towards the end. Most of the end game fights, unless you have insane DPS, last long enough for you to run out of TP towards the end while the boss still has 1-2 minutes of life left. The fights are just that long.

[Phreak]

same goes for any of the turns in coil..

You must have the worst uptimes imaginable on bosses then or your DPS are insanely geared and well played or your healers somehow don't need MP ballad so your BRD plays around with TP instead.

I always bottom out on TP on every turn boss just using a standard BB > BB > SE rotation, outside of using Over Power on Phase 1 and 4 of turn 4 (which isn't a drain due to having TP Ballad running at that time).

[Coramac]

And it's effectively an average as well. It's the point where you stop being able to use attack. I could use 1000-(mean TP cost), but it doesn't really change anything: you get 129.5 GCDs for Non-Frac and 112.9 for Frac. You'll still have Fracture maintaining a 2.3% lead on damage for the first 113 GCDs that fades to nothingness at 130 (a .08% advantage) and proceeds to remain functionally nonexistent thanks to the dilution of the previous 130 GCDs (a .48% lead at 8 minutes).

I was wrong. It wasn't chosen arbitrarily. It was chosen incompetently. If we're using your net costs and pretending that there is 50 TP regen per GCD, then 50 is never the ending number. It is an impossibility. To net the cost, regen must occur after the ability is used. If an attack leaves you with 50 TP left, you will regen 50 TP before the next attack leaving the rotation in tact. You can't suddenly say I'm not going to use averages anymore after using them for everything. Actual vs. your averages leaves a real TP deficit 50% of the time (33% of the time you actually have more TP than you predict and 17% of the time reality and your averaging have the same TP). You can't average everything out to simplify things them sudden say "I'm going to stop using averaging because I feel like it." It doesn't work like that.

I haven't changed my reasoning at all. My reasoning the entire time is that the TP cost on Fracture makes it an attack that you don't want to use because it doesn't add enough damage to your attack rotation to make it justifiable. I've refined the math, but the reasoning has remained the entire time.

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.

...

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.

Fracture isn't worth it because it drops DPS. Now, Fracture isn't worth it because of it isn't at some arbitrary level that you conjured up when you were proven incorrect. Those represent two fundamentally different lines of logic.