DPS Spreadsheets

[Martel]

I finished my Jump ODD Test. Confirmed ODD doesn't work on jumps.

So, that's a valid issue to be fixed in the spreadsheet. Poor Gungnir.

While I'm at it, a couple suggestions.

Gear to add: Ghillie earring +1, Phos belt/+1, Rancorous mantle, Wanion belt, Drachenhorn, Vulcan's pearl(+4 str), Herja's Fork.

Add sonic thrust to the WS list.

Merkira Oto+1 latent. I just added a helm called Mekira Oto+1(Latent) then added it to the fTP gear cell in the data tab.

Oh, could also set jump/high jump to auto-crit when using Ryunohige. I can account for it by putting +100% crit in the set bonus cell, but it'd be nice to not have to redo it everytime I change polearms.

[Dotsudoku]

Is there a reasonably accurate way to simulate performance against some of the sturdier content currently offered? Provenance and legion, for instance.

[Byrthnoth]

I've been bumping the Qilin level up to 120 to simulate newer content, but I'm not sure that's accurate. The best way might be to go RNG on the test server and waste Legion entries taking single shots at monsters using Flashy shot to remove level penalties. You could repeat this using different Ammo to approximate your fSTR and conclude what pDIF and target VIT is.

[Martel]

Unless someone figures out legion mobs stats, there probly won't be.

Oh, and the spread sheet acts a bit weird when you put -crit rate in large amounts. Was trying to simulate Mul wave 2 and their crit- trait. Even after I should have had a 0% crit rate, putting more crit- would continue to reduce my dps.

[Motenten]

Drg:

Removed ODD from all jump damage.
Added gear.
Added Sonic Thrust; had to guess at 200 TP value.
Jump/High Jump have 100% crit rate when using Ryunohige.
Crit rate can't go below 0%.

Mekira latent not added. Have to see if there's a better way to do it.



And yes, need some work done on finding stats for new high-end mobs.

[Motenten]

Speaking of mobs, I'm going to move Fodder down several levels on the new spreadsheet updates, to hold a more appropriate position for its name.

[Taint]

I use

Target Ig
Level 120
pDif Correction 1.05
Defense 500
Final Def 500
Evasion 530
Vit 120
Agi 120


For high end mobs. The Eva should be very close for Ig based on my parses on DRK. (agressor,DE,SE barely capping me)

[Martel]

Drg:

Removed ODD from all jump damage.
Added gear.
Added Sonic Thrust; had to guess at 200 TP value.
Jump/High Jump have 100% crit rate when using Ryunohige.
Crit rate can't go below 0%.

Mekira latent not added. Have to see if there's a better way to do it.



And yes, need some work done on finding stats for new high-end mobs.

Awesome, thank you. ^^

[Churchill]

Ig is evasive enough that without madrigal, an Ukon will not be capped (usually around 85-89%) and closer to 93-95% with it.

[Byrthnoth]

I parsed a 95% hit rate on him last night over 12 kills without Madrigals and I know he Blindga'd me once or twice.

[Motenten]

Added Madrigal to the existing revised spreadsheets (dnc/drg/sam/war). Also rearranged the Cor roll block to make room.

[pchan]

You called them minuet instead of madrigal.

[Motenten]

You called them minuet instead of madrigal.

Whoops, typo on the war spreadsheet. The others were named properly. Fixed.

[Motenten]

Pulling data from the mobs mentioned in the mob data thread (web page: http://www7b.biglobe.ne.jp/~fit/memo/st.htm)

Decided to add in Hydra as a target mob, since they have notably different stats than other stuff presently in the spreadsheets.

Took the highest level for each. Was looking at giving the full range, but it got too complicated, and probably not worth worrying about the lower levels if you're estimating against a general target.

Used an estimation of Agi by taking Hume stats, looking at the ratio between lvl 95 and 75, applying that ratio to the lvl 95 stats (remember this difference), taking the ratio of lvl 75 vit vs lvl 75 agi, then readding the difference from 95 to 75. This is on the basis that stats post-75 are basically linearly constant.

I have nothing to go on for evasion, so just carried over the value used from Nightmare mobs. These need to be tested if possible.


Replaced 'Fodder' with 'Seaboard Vulture', as it represents roughly the same position I was aiming for (mid-80s mob). Vit, Def and Eva are known. Agi estimated:

War type mob(?), 274 skill at level 84 == 266 evasion
Total known evasion: 351
Evasion from Agi and traits: 85
So:
Agi = 170, or
Evasion Bonus 2 and Agi = 126 or
Evasion Bonus 3 and Agi = 100 or
Evasion Bonus 4 and Agi = 74

The most likely seems to be the last of that list, given Vit=75.


Hydra Mnk and Hydra Rng gives a decent range for their mob type: both have fairly high def, mnk has high stats and good evasion, rng has low stats and very poor evasion. Rng is pretty similar to Brd, so dropped Brd. Left Drg in, as it covers a bit of a middle ground between the two.

So this gives a spread of:

Fodder [Seaboard Vulture] -- easy mob of a type that may still show up in content (eg: Neo Nyzul)
Nightmare mob -- most commonly killed mob in the game right now, this gives stats that roughly approximate the ranges available
Hydra mobs -- DC mobs with very high Def and defensive stats. Tough mobs, but without level correction.
Bluffalo -- EM/testing mob with crit defense bonus. (Note: adding mob's crit defense bonus in). Need to get Eva/Agi numbers (evasion is easy; should be able to derive agi from evasion).

I think that decently covers the needs for mobs EM and below.

Added in crit rate mods for mobs, so that can be applied to Legion targets later.

I would appreciate it if people would be willing to help revise the data for Bhukis, Pil and Qilin, as well as get data for Ig-Alima, Provenence Watcher, and any Legion target of note (your choice).


Current table (implemented in Sam, will expand to others):

Mobs                    Level   Defense Evasion Agi     Vit     CritDef CritRate
Seaboard Vulture        84      360     351     71      75      0%      0%
Nightmare mob           96      435     397     90      85      0%      0%
Hydra Drg               95      648     430     125     133     8%      0%
Hydra Mnk               95      661     464     120     150     0%      0%
Hydra Rng               95      642     335     110     124     0%      0%
Bluffalo EM             99      516     411     77      118     14%     0%
Bukhis                  101     560     430     100     110     0%      0%
Pil                     110     560     470     110     115     0%      0%
Qilin                   110     560     485     120     120     0%      0%

[Martel]

I have a fairly massive amount of Mul data for legion. But I'm not sure how reliable the data would be, considering the chaos of actual combat situations.(Buffs wearing/varying cor roll potency/debuffs, etc)

I'm also not sure what to do with said data to get mob stats, assuming it's usable.

[Taint]

I parsed a 95% hit rate on him last night over 12 kills without Madrigals and I know he Blindga'd me once or twice.

DEX merits helps a lot.

My stat guesses were when Ig first came out.

[Motenten]

I have a fairly massive amount of Mul data for legion. But I'm not sure how reliable the data would be, considering the chaos of actual combat situations.(Buffs wearing/varying cor roll potency/debuffs, etc)

I'm also not sure what to do with said data to get mob stats, assuming it's usable.

It'll be difficult, but a few things can probably be gleaned from that.

First you need to find the vit. For that, you want the weapon damage and str of a given DD, and a significant number of hits below 1.5 cRatio. In a heavily buffed environment, this is more likely to occur for a one-hander than a two-hander (though with dual-wielders you also have to account for possibly different damages per weapon).

What you want to look for is a major frequency spike in the Offense Details page for a given player/mob combo, to indicate the 1.0 pDif value. Cross-reference that with the weapon damage and player str, and you can get a good estimate of the vit. Multiple confirmations should be made, if possible.

Once you find the vit, you can then determine the mob level. This will require that a given player will be able to cap cRatio, and so you most likely want to look at buffed Wars and Drks. Within that, find their maximum damage done, along with weapon damage and fStr. We can then figure cRatio from that, and from cRatio determine the level correction, and thus the level.

Evasion will be tricky. Might be able to get rough estimates by using the Buffs By Time tab, assuming they used some variable acc buffs over the fight. Ideally, you want someone with a constant accuracy over the entire fight, but I suspect that won't easily happen. With Buffs By Time (and just one accuracy buff at a time; combined buffs messes it up), you can see if the difference between with and without the buff approximates the expected gain from said accuracy buff. If it does, you can use those values to get an approximate evasion, and then factor out level correction.

[Martel]

Is there anyway to combine multiple parses in one go? Doing it one at a time gets kinda old, and I start to lose track of which I've added. <,<

[Motenten]

Not right now, sorry. Would take a bit to put that together; will get that added in when I can, but won't be immediate.

[Martel]

k, np. One by one it is!