Anything You Can Do, I Can Do Better!: RDM Guide

Distract/Frazzle can last 20+ minutes on non-NMs :D

Someone needs to adjust the Debuffed timers then!

A lot of the time you need to do that yourself. I've never found the default Debuffed timers to work correctly without manually telling it what multipliers I'm using from gear

Maximizing Savage Blade:

SB is a 2 hit weaponskill, but I'm going to assume that since this isn't an FTP replicating WS and the scaling is so high on the first hit that stacking additional multi-hit isn't going to do much. Is the target then to max str/mnd and attack to cap, then PDL as needed? Where in the weaponskill damage calculation are multi-strike, attack, pdl, and wsd applied? Does WSD continue to multiply things at damage cap or will you get diminishing returns from stacking wsd if you're already there over pdl?

It's hard to give definitives here because it depends on the values you're dealing with (10 WSD vs 5 STR or whatever) but here's how I would evaluate stat priority for SB:

PDL (if attack capped) > WSD > STR > MND > atk >>>>>>>>>>>> multi-attack

Multi-attack procs do next to nothing for SB, you'll probably never notice them unless you're watching your TP return.

Damage = Base Damage × Ratio × WSD

Base Damage = (Weapon Damage + WSC + fSTR) × fTP

Ratio(pDIF) = Attacker Attack/Defender Defense

Multiattacks add 1 to fTP each a maximum of two times per weaponskill and a maximum of eight total hits, only if the hit connects. One hit WS will only ever be eligible for one multiattack proc.

PDL from gear and aftermath multiplies the pDIF cap (for a RDM w/ a sword, this is 3.35) and PDL from traits adds a straight amount to the pDIF cap.


Idk what you're really asking here but you can't go beyond 99,999 damage with weaponskills so any stat becomes meaningless if you're already doing that. There's also no diminishing returns for WSD, the stat doesn't get worse the more you have of it at different soft caps or breakpoints. But due to it being a multiplier of everything else, increasing the base damage or pDIF may get you more overall damage sometimes because WSD is just multiplying base damage, i.e. multiplying 25000 by 23% is more damage than multiplying 24000 by 26%. PDL is often better than WSD if you can support the increased attack requirement because it is a multiplier for a multiplier, leading to it just being a stronger stat, not because WSD is worse because you have "too much" of it.

I would go so far as to say that the only PDL rdm should consider outside of VR ring choice is going to be crepuscular pebble and sroda ring. The Bunzi body doesn't have enough PDL or WS related stats to be worth swapping out Nyame for it.

Ran WS calculator using Nyame body vs Bunzi body (R25 vs R30) and Nyame still pulls ahead even while attack capped.

Crep pebble & Sroda Ring do quite well for Savage Blade though, so I agree with the assessment above.

The way this is worded is potentially a bit misleading: multiattack only gets a chance to activate on the first two hits of a WS, but they can add more than two hits (i.e, a QA proc will add 3 hits), up to the 8 max
Also, each of those hits does bring +1 fTP unless the WS is fTP replicating, in which case each additional hit will have the same fTP as the first one - and both Evisceration and CDC work that way (not that RDM really uses those in 2024)

He was specifically asking about Savage Blade optimization, so I didn't bother including anything about fTP replication since it's irrelevant in this circumstance. He also already seems perfectly aware of how fTP replication works. I guess I can see how someone who isn't literate in the mechanics could misread what I said but in my head the "maximum of 8 hits" already covered the fact that any multiattack will work up to a maximum of two times per weaponskill within the maximum of 8 total hits for the weaponskill.

Where does the Naegling multiplier come into play? Is that additive with WSD or multiplicative at a certain step?

You're right on all accounts but I have to deal with way too many people who don't read previous posts :D


Naegling's Savege Blade specific WSD is a separate multiplier that is applied at the end of everything else and works on all hits (regular WSD only applies to first hit)

So the '"Savage Blade" Damage +15%' changes:
Damage = Base*Ratio*WSD
into:
Damage = (Base*Ratio*WSD)*1.15?

So then would 'Weapon Skill Attack bonus based on the number of upgrades' apply at:
Ratio(pDIF) = (Attack*(1+(.01*buffcount))/Defense)?

Base is weapondamage+(str*.5)+(mnd*.5)?

Do you happen to know good writeups on how magical weaponskill damage works?

https://www.bg-wiki.com/ffxi/Weapon_Skill_Damage

It's multiplication, it doesn't really matter what step you do it in as long as you're only multiplying the things that need to be multiplied (i.e. for multihit and multiattacked WS you need to multiply by WSD before adding the extra fTP from additional hits). Very rarely there will be fractional rounding issues if you do it out of order but that practically never matters for weaponskill damage since it crops up as a result of flooring.

(Base*Ratio*WSD)*1.15 is effectively the same as writing

Base * 1.15 * Ratio * WSD or

Base * WSD * Ratio * 1.15.

You will get the same answer every time.

10000 * 1.15 * 3.35 * 1.4 = 53935

10000 * 1.4 * 3.35 * 1.15 = 53935

(10000 * 3.35 * 1.4) * 1.15 = 53935


For multihit (like Savage) you'd calculate base damage with only the base fTP, no +1 for the extra hit, then multiply by WSD. Then you'd add the extra 1.0 fTP worth of base damage and multiply by everything else.

That math only works for single hit

I already covered this, idk what the problem is. For emphasis, I've added what you do for a multihit WS at the end

((Base * ftp * wsd) + (Base * # extra hits)) * ratio * 1.15

Edit: the only problem is that for someone that is clearly not knowledgeable on basic ws math. Not specifying an equation that separates first and extra hits is confusing

Edit: I’m referring to the person you’re replying to as not knowledgeable. Not you. Just want to make that clear

Also note that not all WSD is the same. This isn't something you need to consider for RDM typically but for other jobs you'll want to know that WSD from job traits (i.e. /DRG trait, some types of augmented WSD) applies to every hit of the WS unlike gear WSD.

Yes, I understand the commutative property of multiplication, i added an unnecessary set of parenthesis to emphasize when those bonuses are applied as stages in a formula that's processed in stages.

I am not arguing with anyone. I'm just trying to improve my understanding of exactly how the multi-step calculations are occurring, if they are processed in separate steps, and what order those happen in, because that will help me interpret descriptions on other pieces of gear.

Investigation into various aspects of Frazzle III

In this post I hope to bring some interesting details about Frazzle III to light.

First, the magic accuracy for Distract I/II and Frazzle I/II was adjusted in the Mar 26, 2015 (JST) Version Update. A couple of months later, Distract III and Frazzle III are introduced without any mention of magic accuracy adjustments for these spells. Testing has been done with Distract I/II and Frazzle I/II using the automaton scanner method and has been quantified at +150 magic accuracy. Regarding Frazzle III (or even Distract III), I have been unable to find anything to confirm or deny the existence of a potential magic accuracy bonus. I wanted to know for sure so I got to work.

The target: Lvl 133 Apex Toad in Woh Gates, Darkness Rank G (100%), INT-299, MND-228. Monster stats found here.

The test: With a specified amount of magic accuracy, repeatedly cast Frazzle III and record the results. According to this magic accuracy table, 986 magic accuracy is needed to reach 95% hit rate. Using the stats/equipment below, if Frazzle III has a magic accuracy bonus then hit rate should fall somewhere in the 60-70% range. If not, then hit rate should fall far below 50%. No testing done on Light/Darksday, only neutral weather present. Enemy level verified by casting Thunder and finding the ones that took the least damage.

Relevant Stats:
M lvl. 33 RDM/NIN
509 Enfeebling Magic skill
Job point gifts Magic Accuracy +70
Job Points Magic Accuracy +20
5/5 Magic Accuracy Merits Magic Accuracy +25

Equipment
Range: Ullr Magic Accuracy +40
Neck: Erra Pendant Magic Accuracy +17
Left Ear: Crepuscular Earring Magic Accuracy +10
Body: Crepuscular Cloak Magic Accuracy +85
Ring 1: Shiva Ring
Ring 2: Crepuscular Ring Magic Accuracy +10
Back: Sucellos's Cape INT +30, Magic Accuracy +20
Waist: Witch Sash Magic Accuracy +2
Legs: Zoar Subligar R15
All other slots empty
Total INT: 136+163=299, 0 dINT Magic Accuracy Bonus
Total MND: 132+96=228, 0 dMND Magic Accuracy Bonus
Total Magic Accuracy: 808/958 if +150 innate bonus applies to Frazzle III

Test Data
Frazzle III, 808+150 Magic Accuracy
Full Duration: 65 (65%)
Partial Resist: 22 (22%)
Full Resist: 13 (13%)

As can be seen in the data above, hit rate indicates Frazzle III has an innate +150 magic accuracy bonus just like it's lower tier counterparts. This also means Distract III likely has the same bonus as well. Now, let's see if we can build upon this data.

It is known that Frazzle uses dMND for potency purposes, but what does it use for dSTAT magic accuracy correction? In doing some research, it seems there is some conflicting information.

There was some preliminary testing done a long time ago by Lute with Distract. Despite the small sample size, there appeared to be a clear correlation between higher MND and fewer resists when casting Distract.

There was also some testing done using trusts that found boosting Koru-Moru's MND had a positive effect on his AI causing him to cast Distract.

Additionally, there was testing to verify Frazzle I's magic accuracy bonus. It was also found that lowering enemy MND with Shock had a positive effect on the automaton's AI causing it to cast Frazzle. It is then stated that although caster dSTAT has not been confirmed, caster MND to target MND is the likely dSTAT correction for Frazzle. In their next post, they stated that caster MND to target MND is the dSTAT correction for Frazzle. A few days later this is edited to state that caster INT to target MND is the dSTAT correction for Distract/Frazzle without any real explanation or data for the change. Hmmmmm.

So, does Frazzle rely on a player's dINT, dMND, or maybe both for magic accuracy correction? In the next two trials I used the same setup as above with the addition of either Gain-INT (INT +55) or Gain-MND (MND +55) to each trial. With the above equipment giving us a dINT and dMND of 0, each Gain- spell will give us +26 magic accuracy. If either INT or MND contributes to the magic accuracy of Frazzle then hit rate should be in the low 90% range. If not, then there will obviously be no real change in hit rate when compared to the above data.

Test Data
Frazzle III, 808+150 Magic Accuracy w/Gain-INT
Full Duration: 67 (67%)
Partial Resist: 20 (20%)
Full Resist: 13 (13%)

Frazzle III, 808+150 Magic Accuracy w/Gain-MND
Full Duration: 91 (91%)
Partial Resist: 8 (8%)
Full Resist: 1 (1%)

There is a very clear difference here. This data set confirms that despite Frazzle being a black magic spell, player dMND is used for Frazzle's dSTAT magic accuracy correction and not dINT. This by extension also likely applies to Distract as well.

Lastly, Lute performed some verification testing of Frazzle III a long time ago using the automaton scanner method. Enfeebling magic skill potency values 476-625 were verified. Using this data, I plugged the numbers into the formula on BG and the results matched their testing. I would say the Frazzle III formula doesn't need verification tags anymore.

Fantastic work. Personally, I was still defaulting to INT for Frazzle macc. Black magic as a general rule works off dINT for macc, and just because the potency is MND based.. doesn't mean the macc dSTAT is as well. Now, we know. And can full focus on MND hitting potency and macc at the same time.

I've updated the BG wiki Frazzle III page with a reference to this test, and notes on the innate macc bonus and the spell's use of dMND for macc.

EDIT: Oh. I missed the bit about the skill formula verification. Removed verification tags and updated the reference title on the BG wiki page.

So Cali probably beats Murg when they are both maxed for these spells...Well done SE

VINDICATION!!!!

Does this mean that the older strategy of applying II and overwriting with III - because II went unresisted more often - is not actually needed? God knows I picked up that single nugget of info so it may have been wrong all along.

That strategy is based on making a set for Frazzle II that was pure accuracy while making a set for Frazzle III based on potency instead, in order to take advantage of Frazzle II lowering the required macc needed to land a full potency Frazzle III. At least, that was my understanding of it.

Now that both makes sense and still applies! Thank you.