Now that I have two “models” which are both updated to represent accurate Determination scaling, I’ve decided to test them and calculate some updated weights for the Dragoon. In all honesty, this is very impressive… One thing to note with Crafted Gear: The lower your Strength, the lower the relative value Determination has. At i130, your Determination Weighting will be pushed up to 0.43. But, you have a set-up similar to what I have, where you have 646 Strength due to crafted gear, the weighting of Determination falls down to around 0.41
Now now lads and gals, don’t be thinking that this is set in-stone. There’s still more work to ensure we even greater accuracy, but these weights are a rough estimation for what is to come.
Also, Weapon Damage’s relative value is lower with my calculation. Previously, we just used Weapon Damage to calculate Auto-Attacks, as AA_DMG/AA_Delay provides “roughly” 100 Potency like standard Ability skills (It’s closer to 98.22 I believe). But now that Weapon Damage is no longer factored into my Auto-Attack damage calculations (the final damage is AD+AA), an increase in Weapon Damage doesn’t affect the Weights as much as they used to. I might just revert it back to how it originally was, for simplicities sake.
Dervy does Theorycrafting 
Please please please Dervy, level up BLM and do the same thing. We are misinformed since 2.0 😥 😥 😥
LikeLike
I wouldn’t necessarily need to level up a BLM to 50 to do something similar, but I will have to do a lot of work to create a hybrid model/DPS calculator for the BLM similar to what I have for all other jobs.
I could just take a Puros thundercloud and fire starter formulas…
LikeLike
I’m not sure that take anything from puro’s work is a good idea to start something (credibility… xD)
BTW instead to level up a job, you could also create a specific methodology and others (like myself) could provide you datas (why not with a video that show every dmg and the equiped gear so you could double check that every value is correct) so it could be your work from the beginning to the end
LikeLike
The main thing is calculating Fire Starter procs and Thunderclouds. Once I have a formula for that, I can get a very good model done.
LikeLike
ok ok ^^
LikeLike
Good news is I just hit 50 BLM and got my relic, so I can get you some damage data after I take a short break and get dinner!
LikeLike
Hey dervy, I know this is an oldish post but I still thought it was the best place to comment, hope you get to reading it.
I don’t actually main a DRG (whm here), my Bf plays a dragoon and he follows all of your drg guides.
I was wondering, now with the increased emphasis on crit over det for a dragoon, might weaving in all non essential off gcd skills be a potential dps loss? I didn’t want to do the maths myself because that’s my day job, but I thought you might want to because it’s your thang
Short explanation of what I mean;emphasising crit means that you would want to lessen the number of skills that can crit so there is a greater chance that your high potency skills crit. I don’t think this would be a major change for drg. Just loss of leg sweep unless you need it for mechanics obv, and possibly gierskogul on a single target, and the jump not affected by botd (dragonfire dive I think)
LikeLike
Sorry let me elaborate a bit on my reasoning coupled with my suspicion on the crit rng – what I noticed was that even in a 1 min parse my bf crits 25% of the time every time. If rng was a pure 0.25 chance of critting on every skill, his crit shouldn’t come out at 25% every time, for such a short parse. Therefore I had a suspicion that se has a probability formulae that increases crit of the next skill after a non successful crit, in order to make sure crit values are reached even in a short duration.
Again this is speculation and I haven’t done any calcs nor do I have enough data: it’s just what I’ve observed and am making rough estimations on.
LikeLike
That’s all completely down to RNG. FFXIV uses the Pseudorandom Number Generator algorithm in FFXIV to generate, non-random, random values.
LikeLike
Hm, I think there might be more to it – I’m not disputing the rng factor at all, I’m disputing the flat 0.25 chance on each skill. The likelihood of getting the exact same crit rate over and over with n=30 (30 gcds approx per minute) with p=0.25 is kinda unlikely, but yes possible. I’m talking probability distribution maths here.
Unless all the parses were just “lucky” and ended up at exactly 25% everytime. Well, I’ll look into it then sometime, when I’m feeling lively. The data shouldn’t be too hard to get I hope.
LikeLike