I personally think the GCD is a bit much at 2.5 default, they should shave it down to 1.5. You're still limited by TP so I dont see it breaking anything. As far as the skillspeed stat itself, I would much rather see an overall haste stat replace it.
I personally think the GCD is a bit much at 2.5 default, they should shave it down to 1.5. You're still limited by TP so I dont see it breaking anything. As far as the skillspeed stat itself, I would much rather see an overall haste stat replace it.


In theory every single TP using class will run out of TP in a reasonable timeframe if they require constant uptime on target.
Thus, it is worthless if fights were like, 10-15 minutes of constant fighting.
Since they are not, we cannot conclude that skill speed is "worthless" but its pretty damn bad.
The other issue is "skill speed lets you get more attacks in for a cooldown". My take is it takes way too much skill speed, and you're better off with another stat.
The other main consideration is whether more skill speed will let monks (maybe dragoons too?) get another attack in inside their maintain rotation. I do not play either so don't know the breakpoints.
But the easy way to think about it is this: would you rather have all your offstats in skill speed? Or all your skill speed in another offstat?
And the answer to that question is easy. Thus, skillspeed is terrible. I won't say worthless, but its pretty bad.

I think Skill speed determination and crit rate need a rework. I find it difficult to notice any change in in my speed / damage / crit rate when stacking them.


That's because you are in lowest tier raiding gear in a "vanilla" environment.
The reason its hard to get stats now is as they release more content, power creep will shoot up your offstats to absurd levels.
As for clipping DoTs increasing mana costs, yes this is true.
However, it is also true that your 3 main DoTs are by far your most efficient spells.
Again, clipping your DoTs by 1 second or so will not affect you in any way at all. It is at most something like a 0.8% dps loss per DoT type that you routinely clip. DoTs are roughly half your damage, so thats 0.4% max loss per DoT you routinely clip by 1 second.
If you clip all your DoTs, you will lose something like 2% DPS. Now, this is significant enough to not want to do, but it is not significant enough that you need to beat yourself up for doing it.
It is OKAY is all I'm saying. Don't worry about it. It is BETTER to clip your DoTs by 1 second then to use that GCD for something else, or to delay that GCD in order to "perfectly" line up your DoT.
I know we are all in favor of maximum numbers here, but we're also all human. Max dps is great on paper but in a fight you have much more important things to worry about (such as positioning, mana, and mechanics). Furthermore, if you know a movement is coming up soon and you won't have boss uptime, I would clip DoTs by as much as 3-5 seconds.
Furthermore as for DoT clipping, summoners have 1 cooldown every 3 minutes, and you use that with contagion anyway so clipping is a very, very minor issue.
Ah, the art of min-max at work

It's only pretty bad for Bards because Crit is good for Bards.
For DRGs and MNKs, skill speed is relatively on par for DPS considerations (although still typically slightly weaker), although it's de-prioritized because it comes with the slight TP burden.
Melding 101 crit will always increase your critical rate by 7%, while melding 101 speed will always reduce your GCD by 0.096.
Not considering procs, you would get 3.5% more DPS by melding crit.
Assuming you have a GCD of 2.4, you would get 4.175% more damage on your GCD abillities by melding speed.
Assuming you have a GCD of 2.3, you would get 4.364% more damage on your GCD abillities by melding speed.
The effects of critstats stay linear, while the ones on speedstats will get slightly better the more you stack them.
Speed has larger effects (usually more than 20%, which is alot) than crit but don't work on AA and dots. Speed leads to more TP consumption.
The mentioned 4.~% for speed are in practice also diminished, due to animationlocks when using offGCD abillities between GCD abillities.
Speed has also something like 'breakingpoints'. When you stack enough, you can fit one more action into timeframes of buffs.
That could allow the usage of superior rotations.
These are my thoughts on speed vs. crit.
I have no conclusion yet.
If the statvalues in question are low, I would probably take crit. But speed might become objectively better for certain classes, when you
stack the stat to very high amounts.

You would get [baseline crit %] less than 3.5% more relative DPS by melding crit.
You would get [AA+OGCD contribution%] less than 4.175% more relative damage from GCD abilities by melding speed. Sidenote: which you mention later.Assuming you have a GCD of 2.4, you would get 4.175% more damage on your GCD abillities by melding speed.
DOT fallacy: skill speed improves DOT uptime. The actual damage delta (due to server DOT timing) is variable and more of a simulation exercise than anything. For a modeling exercise, DOT damage is increased directly with increased uptime.Speed has larger effects (usually more than 20%, which is alot) than crit but don't work on AA and dots. Speed leads to more TP consumption.
Edit: Clarification: the fact that SS does not increase tick rate would be more impacting at higher values of skill speed -- e.g. after the SS stops improving uptime. However, for practical values of SS, it improves uptime.
As a practical matter, animation locks only exist for DRGs where the jump animations exceed the slack time between GCD animations.The mentioned 4.~% for speed are in practice also diminished, due to animationlocks when using offGCD abillities between GCD abillities.
Edit: Or fringe circumstances where you intentionally try to throw a second OGCD between GCDs, but this is a utility topic (Bard silence), not a DPS output topic.
Edit2:
The only major takeaway is that skill speed does have accelerating returns at high values, countered by the accelerated TP drain. Generally speaking I'd expect to hit some pretty hard TP walls before a player sees any notable DPS accelerated returns.
Last edited by EasymodeX; 10-22-2013 at 12:07 AM.
Thx EasymodeX for your annotations.
I see what you mean with the relative DPS increase with crit being dependent on the critchance you already had before increasing the chance.
Didn't thought about that.
The relative DPS increase would be in my example 0.5*7%* 1/(1+prevCritChance) -> rel. DPS increase for melding crit ~3.04% at 15% baseline critchance
To evaluate the 4.~% in terms of relative DPS increase. I could calculate a factor with the knowledge of my AA DPS and my GCD abillities DPS at 100 potency multiplied by an guessed value based on my average potency on GCD abillities I use while fighting. That guessed value would be dependent on the used rotation and the conditions of the fight (interrutions etc.).
So as DRG I would guess values between 1.84 and 2.6~ for example, assuming I perform better than spamming Heavy Thrust and worse than the 27-actions rotation (without offGCD's) while fighting.
-1000 chars-
|
|
![]() |
![]() |
![]() |
|
|
Cookie Policy
This website uses cookies. If you do not wish us to set cookies on your device, please do not use the website. Please read the Square Enix cookies policy for more information. Your use of the website is also subject to the terms in the Square Enix website terms of use and privacy policy and by using the website you are accepting those terms. The Square Enix terms of use, privacy policy and cookies policy can also be found through links at the bottom of the page.

Reply With Quote




