How to Dragoon per Second: A DPS Paradigm

[Strident]

I did some math and in-field testing on the best in slot builds provided at the bottom of the original guide and have concluded, based on the (perhaps outdated?) provided Dragoon damage models, that Dervy's Best-in-Slot configurations is almost certainly not optimal.

To perform the comparison, I checked FFXIVGuild's provided BiS gear and found that the skill speed difference and accuracy differences were negligible (2 Skill Speed and 45 Accuracy in favor of FFXIVGuild's build). The big difference is in 116 CRIT vs 156 DET, and the difference based on the given CRIT and DET values comes out to ~4.4% extra damage in favor CRIT, which indicates to me that CRIT is grossly more powerful than DET point for point.

This also leads me to believe that Ariyala's build also cannot be BiS as it is down Crit and Skill Speed for DET (61 Crit and 27 Skill Speed vs 126 DET). I performed the same comparison with FFXIVGuild's BiS slot, while making the assumption the Skill Speed difference of 27 would not affect DPS (I highly doubt that). The difference finds that even ignoring Skill Speed, the FFXIVGuild BiS build deals ~2.1% more damage based on the current model.

Jack's build is one that I can't decide on. The difference is 30 Crit vs 29 DET and 44 Skill Speed (solely the Gordian Wristband of Slaying vs the Fabled Bracelet of Slaying). Of course the Crit beats out the DET but I am not sure on the Skill Speed. In any case, of the listed BiS load outs, Jack's is the best, with the exception of the Gordian Wristband being potentially interchangeable with the Fabled Bracelet.

[ish007xx]

a new thread again lol

[JackFross]

(snip)

You missed the discussions where all of these were brought up.

That was pretty much the consensus, but the damage differentials were all decided among most/all of us to be so small/negligible that they just sorta don't matter. The set that's attributed to me is the "technical" best in slot based solely on stat weights. Dervy's is a more consistent version (more det at the cost of ss and crit) of mine, and the other set is low on ss but technically higher weight than either of the others.

It's hard to settle on one set as being "best" when you're looking at numbers on the scale of 1500 and seriously coming in to say that one set will give you +30 dps over another. That +30 dps is literally one dropped Blood. It's one flubbed Wash Away backflip. It's not something that I feel comfortable saying definitively one way or the other, especially when people have pings such that 606 ss isn't enough for them to get 3 GSK per minute and other people who find 700 ss to just be unwieldy because of how fast you become.


The discussion of best in slot this patch is a lot more nuanced than just "which set is definitively the best" because that question doesn't exactly have an answer that is very satisfying, due to the way the different stats work and play with each other and the person using them.

I mean, this is the set that I'm personally building for, which is Dervy's set with the weapon swapped. It's weaker, yes, but I much prefer the grace I get from higher skill speed; it gives ME higher gains in the long run because I end up less self-conscious about losing Blood to a Geirskogul or two and end up actually squeezing them out.


That said, what set are you referencing? FFXIVGuild's?


///Oh, it's the same as mine but with 700 accuracy? That's silly. The det/ss is strictly better than the acc/crit. There's no question to be had there. There's not near enough crit to make up for the gains you get from the other secondaries. Otherwise, you're telling me that the raw difference in damage power between 30 crit and 29 determination is more powerful than the gains from 44 skill speed. I don't think that's something that should even be in question. There's no way that's the case.

[Strident]

Otherwise, you're telling me that the raw difference in damage power between 30 crit and 29 determination is more powerful than the gains from 44 skill speed. I don't think that's something that should even be in question. There's no way that's the case.

I think that asking whether or not the bolded statement is true is a valid question. Sometimes silly things turn out to be the most optimal alignment because it is more powerful in another specific area. This is why I wanted to avoid making any conclusions or hypothesis' on your build in comparison, because I am unsure of how to mathematically value Skill Speed. I would like to believe that 29 DET and 44 Skill Speed is superior to 30 Crit and an overload on accuracy, but I refrain from making that judgement. You seem to be confident that it is, and that's valuable information for me; how so exactly? I would be intrigued to know how I should value Skill Speed and in what context.

I also acknowledge that the mathematics I based my calculations on are (which were provided in the OP) are outdated; a more accurate assessment of the numbers would require more research.

That being said, this consideration was purposed solely to investigate flat hitting power in the comparison between Crit and Determination, which finds that every point of Crit is worth at least two points of DET for the purposes of DRG BiS. When comparing FFXIVGuild's configuration against the other two, I thus found that the flat hitting power of the former is significantly superior despite its Skill Speed advantage against both, which I thought was important. If we use 1500 as the baseline, the ~4.4% differential can lead to a 66 DPS increase.

Is this still true? I don't know and I would very much so like to hear about this.

[SunnyHirose]

Oh, we're talking BiS again? So, I made a chart, trying to be a cheapskate. Nothing doin'; Baked Pipira Pira is king.



How accurate are these DPS estimates? Well, Dervy's current damage formulas--which are used here, but were calculated from much weaker gear--seem to underestimate AAs by a bit and underestimate potency by a very small bit. It is at this point I wonder if there is merit to considering stats to be very slightly quadratic or something, since the crafter formula people seem to :S

And skill speed is still slightly mysterious (I haven't gotten around to testing it as thoroughly as I'd like), but perhaps by coincidence my rotational potency estimation (which depends on it) seems to be reasonably on point when compared to my performance using actual gear and measuring damage per AA/potency directly from the game.

Anyway, BiS DPS might actually hit an average of 1500. Maybe.

[Hakmatic]

http://ffxiv.ariyala.com/RNWS that's what I'm going use for a4, one reason is I have all that gear except weapon and it's cheap food. I'll use baked pira for 1-3. I'm switching back to monk at some point in time so I'm not going be trying for any other Dragoon gear. I thought I'd share this since it does use a cheaper food which might appeal to some.

[Thendiel]

I did some math and in-field testing on the best in slot builds provided at the bottom of the original guide and have concluded, based on the (perhaps outdated?) provided Dragoon damage models, that Dervy's Best-in-Slot configurations is almost certainly not optimal.

To perform the comparison, I checked FFXIVGuild's provided BiS gear and found that the skill speed difference and accuracy differences were negligible (2 Skill Speed and 45 Accuracy in favor of FFXIVGuild's build). The big difference is in 116 CRIT vs 156 DET, and the difference based on the given CRIT and DET values comes out to ~4.4% extra damage in favor CRIT, which indicates to me that CRIT is grossly more powerful than DET point for point.

This also leads me to believe that Ariyala's build also cannot be BiS as it is down Crit and Skill Speed for DET (61 Crit and 27 Skill Speed vs 126 DET). I performed the same comparison with FFXIVGuild's BiS slot, while making the assumption the Skill Speed difference of 27 would not affect DPS (I highly doubt that). The difference finds that even ignoring Skill Speed, the FFXIVGuild BiS build deals ~2.1% more damage based on the current model.

Jack's build is one that I can't decide on. The difference is 30 Crit vs 29 DET and 44 Skill Speed (solely the Gordian Wristband of Slaying vs the Fabled Bracelet of Slaying). Of course the Crit beats out the DET but I am not sure on the Skill Speed. In any case, of the listed BiS load outs, Jack's is the best, with the exception of the Gordian Wristband being potentially interchangeable with the Fabled Bracelet.

Here is Dervy's reply:

DPS factors 1.03 party buff.

1. FFXIVGuid.com "BiS" set: http://www.ffxivguild.com/ffxiv-drag...&ct=clnk&gl=uk / http://ffxiv.ariyala.com/RJ5D

WD: 76 STR: 1079 ACC: 700 CRT: 930 DET: 328 SS: 606 Weight: 1721.240 DPS: 1508.56835902

2. Auirily's "BiS": http://ffxiv.ariyala.com/RL7Z

WD: 76 STR: 1079 ACC: 656 CRT: 869 DET: 454 *SS: 579 Weight: 1726.210 DPS: 1514.57267975

3. Jacks 650 SS: http://ffxiv.ariyala.com/RM4U

WD: 76 STR: 1079 ACC: 656 CRT: 869 DET: 454 SS: 579 Weight: 1724.852 DPS: 1514.01090884

4. My "Balanced" Set: http://ffxiv.ariyala.com/RLTG

WD: 76 STR: 1079 ACC: 655 CRT: 814 DET: 484 SS: 604 Weight: 1723.970 DPS: 1511.55580904

FFXIVGuild is the worst of the bunch. Not only that Strident. You can just use the stat-weights to get a rough estimation of the value of this 116 CRT vs 154 DET. 116 * 0.162 = 18.792 156 * 0.139 = 21.684 DET alone holds more stat-weight than Crit.

And no, Crit isn't x2 the value of Determination. Not even close. The problem is that you're calculating it one stat vs one stat. It doesn't work like that. Determination and Critical Hit Rating, although linear, scale from different base values and when you're calculating DPS of a set, you don't base it off 1 stat vs 1 stat. You compare the entire set together.
Not only that, I have a the biggest feeling you're doing something like (1 + 1.58 * 18.4%) to get the multiplier of CRT on damage. That's incorrect. You always do Critical Damage -1. Why? Because 1+ is a value used to convert the percentage, or, our coefficient, into a whole number value we can multiply with. Not only that. For arguments sake, let's assume we have a Critical Multiplier of 1.5, and a Critical Hit Rating of 20%:

Doing 1+1.5*0.2 = 1.3... No. We do not have a Critical Factor of 1.3. We have a 20% chance of getting that 1.5 damage boost. Our Critical Factor would be much lower. Using an expression like this gives the end result a multiplier of 2.5 if you had 100% Critical Hit Chance, which isn't the case. This isn't Diablo 3. We're not adding an additional 150% onto our base Critical Damage, which is 100%.

We should be doing 1+0.5*0.2 = 1.1

That, is most probably where you're getting your "Crit is 4% better!" Errors from. So if we compare my set and FFXIVGuilds set:

FFXIV Guild DPS contribution:
((930-354) / (858 * 5) + 0.05 = 0.184266
((930-354) / (858 * 5) + 0.45 = 0.584265734
(1 + 0.584265734 * 0.184266) = 1.107660155
(1 + 328/ 7290) = 1.0449931
1.107660155 * 1.0449931 = *1.157497264 Multiplier on WS Damage, or a 1.57% contribution to your base damage.

My sets DPS contribution:
((814-354) / (858 * 5) + 0.05 = 0.15722
((814-354) / (858 * 5) + 0.45 = 0.55723
(1 + 0.584265734* 0.184266) = 1.087610492
(1+ 484 / 7290) = 1.066392318
1.066392318 * 1.087610492 = *1.159819474 Multiplier on WS damage., or a 1.598% contribution to your base damage.

My Set is better by 0.028%.