All Formulas are up in the “Damage Mechanics and Formulas” page. If you click on the 3 lines on the top right corner, it’ll open up a menu which you can use to navigate my posts. Anyway, here’s the current and most updated damage formulas:
Critical Hit Chance
((CRT-354)/(858*5))+0.05
Critical Hit Damage
((CRT-354)/(858*5))+1.45
Skillspeed on GCD
2.50245-((SS-354)*0.0003776)
Dervy does Theorycrafting 
Hi,
Some days ago, I seen a case where your formula prediction wasn’t match my current GCD.
So, today, I use my time to collect data and try, in game, each GCD cap for SS < 888 (not enough stuff for better ^^), and it seems to me your formula fail for some SS values.
So I worked on a new formula to match failure cases, what is following :
GCD = 2.51 – (SS – 334) / 2641
(rounded at lower, validated for all SS value < 888)
My data, maths, and comparaison with your formula : https://docs.google.com/spreadsheets/d/1bpgNXe_g1nWHSO2GC0ubOxT8awd5BCsm_p012FsMMDs/edit?usp=sharing
Hope i'll be helpfull !
LikeLiked by 1 person
Unfortunately, the doc is a PDF and unable to view it correctly on my phone.
Two questions.
What’s the main difference between Dervy1 and Dervy2 column formulas? Is it just the starting GCD?
And secondly! How comes you’re doing SS-334? I assume it’s because it takes 7 SS @ 60 for us to see the first shift in GCD, where as we have an average of 26.5 per GCD shift, so 354-(26.5-7) = 334.5? But as the second shift is 26, you did 354 – 20?
Good fucking job on fixing my work!
LikeLike
Dervy 1 is formula I seen from DRG thread in official forum : 2.50256-(0.01*(SS-354)/26.5) http://forum.square-enix.com/ffxiv/threads/238845-How-to-Dragoon-Per-Second-A-DPS-Paradigm
Dervy 2, formula upper this thread
Two different step of your work, I guess.
Basically, I choose 334 because it seems to do the job.
Shift is sometime 26, or 27, but it’s more complicated than a 26.5 shift. Sequence from data I got is something like 26, 27, 26, 26, 27, 26, 27, 26, 27, 26, 26, 27, 26, 27… (If I remembed good)
So I do an empirical search for a shift factor and a gap value matching this sequence, I found 26.41 and 334.
Maybe we could found a shift factor matching SS-354, I didn’t try it.
LikeLike
Hey mate.
So, I decided to try out your new formula fix along side 2.0 SS values, it’s a *near* perfect fit.
The only issue, is that instead of starting from a base of 334, I start from a base of 333. The shift per GCD in 2.0 is 10.5.
This is the rule that applies to 3.0, or level 60. The first shift after 2.49 GCD is 26, so the previous is a 27 shift.
334 + 27 = 361, where our first shift happens.
Here’s the thing. After we hit 2.49 GCD (344 SS) at level 50, the next shift is 11 SS. Meaning, the previous should should be 10… or, 334.
But, it isn’t. Using 334 gives me errors, but 333 doesn’t.
My current formula for 2.0 SS is:
2.51 – (SS-333)/1050
Could you see if your script could identify the issue?
The level 50 SS values are all here:
https://docs.google.com/spreadsheets/d/1GyDql_asLEufO7Spduvn4SLrqUlcxSFj_ipy6Dt6agk/edit?pli=1
LikeLike
I worked on 2.0 formula for a while.
But I’m confused, I took one by one all game data from 341ss to 609ss, and some data I got don’t match purostrider spreadsheet. I’m afraid he made extrapolation with not enough data and his formula don’t match perfectly game data. I added my 2.0 data and formula I found in my spreadsheet : https://docs.google.com/spreadsheets/d/1bpgNXe_g1nWHSO2GC0ubOxT8awd5BCsm_p012FsMMDs/edit?usp=sharing
According my data, and same formula than 3.0, following constants works good with 2.0 :
– base ss : 333.35
– shift : 10.48
But I agree with you, lvl50 and lvl60 formula use probably both the same value as base, so I’ll try to work on a formula with the same constant as base ss.
LikeLike
Another thing I noticed, was if I were to use the same base value of 334 for both 2.0 and 3.0, the base GCD, 2.51 would not work for either. It would be 2.51 for 3.0 and like 2.5094 for 2.0.
But when I use 2.5904 for 3.0, I get errors.
LikeLike
For what it’s worth, my intuition is 2.5 is a static constant. Or a variable, because there’s spells for 1.5 to 4.5 seconds casting time, but I would’nt have to use something like 2.5xxx
By the way, I changed my formula to use 2.5 instead of 2.51, what is look like more natural for a computer formula :
2.5 – int(SS – base) / shift) / 100
I made a program to test MANY (probably all possible) base and shift values matching game data for 2.0 and 3.0, according topper formula. I have some possible results, but no common base value between 2.0 and 3.0 constants. Numerical results are in my spreadsheet.
Is my formula wrong, or is there really different value for base and shift ?
For now I will work on non-2.5 seconds casting time by collecting new data from game, to find a casting-time formula including a base casting time variable, what is 2.5 for a GCD.
I hope It will give more ideas to me 🙂
LikeLike