Faith vs Spharai Parse data? Anyone have?

[Lahan]

I have a semi related question, what is a weapon rank? And throw out some examples please.
(Use drg and sam as they are the other DD besides mnk I ever really use)

What is a rank 0, rank 1, rank 2 ect. weapon?

[Thunder]

I get the "lossy" tp importance when doing the actual math. It changes the ratio of melee/ws damage. I'm not arguing about how relevant it is to the actual calculated damage of faith as a stand alone weapon.

We're arguing different things. I'm only talking about "which is better", regardless of slight 100 tp overflow.

And why couldn't you say this earlier instead of arguing with me if lossy tp matters in math because I tried pointing out that's all I was doing several times now. The whole, "KC = extra DA/kick attacks = faith's lossy tp" was pretty much where I was getting you're a dumb ass from, by the way.

[Byrthnoth]

Weapon Rank - FFXIclopedia, the Final Fantasy XI wiki - Characters, items, jobs, and more
It determines what the cap for fSTR is.

Amano (D88 ) has a Weapon Rank of 9 [ FLOOR(88/9) ], so its fSTR2 cap is 17.
Hagun (D75) has a Weapon Rank of 8 [ FLOOR(95/9) ], so its fSTR2 cap is 16.

If you WS in enough STR, Amano gets an extra base damage from fSTR2.

[pchan]

6) Sorry pchan, I didn't really look at your math (or read the rest of this thread). I was just annoyed that people kept bumping it so I threw some math at it and hoped it would go away.
.

My math was not accurate since nobody knows the procrate of spharai, and also I assumed the 6% crit to proc on kick which doesn't seem to be true. The only way to account for "lossy" tp is to have a program that emulated H2H weapons; too many random number appear : miss/double attack/kick attack/extra attack from ammo, also ... getting hit by monsters, be it magic damage or melee damage. Nonetheless relic should be ahead of faith without the counter effect and the random tripple proc. Also , I'm not sure how faith really work when combined with double attacks. Graphically DA only occur on the left hand, while extra attack proc on both hands, so anyone has ever seen left fist > right fist > left fist 3x with faith which would be 2 normal fist, 2 doube attacks, 1 extra attack (i'm not mentioning 1 DA and 1 extra since you wouln't be able to differentiate it from 2 extra attacks).

[Lahan]

Weapon Rank - FFXIclopedia, the Final Fantasy XI wiki - Characters, items, jobs, and more
It determines what the cap for fSTR is.

Amano (D88 ) has a Weapon Rank of 9 [ FLOOR(88/9) ], so its fSTR2 cap is 17.
Hagun (D75) has a Weapon Rank of 8 [ FLOOR(95/9) ], so its fSTR2 cap is 16.

If you WS in enough STR, Amano gets an extra base damage from fSTR2.

was that supposed to be floor (75)/9 for hagun?

so it is the weapons dmg / 9 = its weapon rank rounded down

so a 95 dmg lance is a rank 10 weapon.

(And what is floor)

[Byrthnoth]

Yeah, sorry. I originally did the math for Valkyrie's Fork vs. Gungnir, but then I realized how completely unapplicable that comparison was relative to Hagun vs. Amano.

[Onorgul]

Regarding the whole premise of WS delay (and ability delay), I've estimated to be a lot closer to 1 second than 3. The animation for Asuran Fists and many other weaponskills is fairly long, but the change to swing delay is substantially less, as can be readily seen if you fire off a weaponskill just before your next attack round. You won't see your normal melee swing but you'll get the message in your log.

[pchan]

Regarding the whole premise of WS delay (and ability delay), I've estimated to be a lot closer to 1 second than 3. The animation for Asuran Fists and many other weaponskills is fairly long, but the change to swing delay is substantially less, as can be readily seen if you fire off a weaponskill just before your next attack round. You won't see your normal melee swing but you'll get the message in your log.

But boosting and swapping hands is necessarily adding delay

[Byrthnoth]

With Soboro it was pretty easy to calculate the lossy TP because you could just assume a 33/33/33 distribution and SAM/DRG sans Brutal. Originally I was trying to calculate it to see if there was an argument for SAM/WAR and Brutal Earring with Soboro/Yoichi, but making the probabilities of a single, double, and triple hit uneven was just not easy for me. I think I did it once for Fortitude Axe, but I either deleted or didn't save it because I can't find it.

I would say that the numbers my spreadsheet comes up with are the minimum pDIF difference +20 Atk would have to make for Spharai to be better than Faith. Also note that it doesn't attempt to factor together WS and DoT mostly because the "lossy TP" issue would have made any factoring I could do pretty invalid. Either way, Faith beats a D23 Delay 86 statless weapon and might beat Spharai under the right conditions.

When it comes down to it though, the I found this thread so silly because one of Monk's primary advantages is how little TP it feeds, and that is somewhat countered by Faiths.

[Vyvian]

Couldn't you figure out a best case TP scenario (like 101), and a worse case TP scenario (landing on 99 and needing an entire attack round, so like 165), figure out the average number of hits and figure out an average TP you'd land on, like 125. Then readjust the formula to say that on average you usually WS at 125 tp on Faith, so instead of hits to 100 you calculate hits to 125. Sometimes you'll go over and sometimes you'll go under, so it averages out more often.

I thought I remembered earlier in the thread there was an assumed average number of hits per attack round that can be used to calculate this.

[Tulun]

When it comes down to it though, the I found this thread so silly because one of Monk's primary advantages is how little TP it feeds, and that is somewhat countered by Faiths.

Isn't this a misnomer? Even with high Subtle blow gear on, the amount of TP fed to get to 100 TP is worse than melee two handers, no?

[pchan]

Isn't this a misnomer? Even with high Subtle blow gear on, the amount of TP fed to get to 100 TP is worse than melee two handers, no?

yes if the 2 hander subs nin, otherwise no.

[Vodou]

I would say that the numbers my spreadsheet comes up with are the minimum pDIF difference +20 Atk would have to make for Spharai to be better than Faith. Also note that it doesn't attempt to factor together WS and DoT mostly because the "lossy TP" issue would have made any factoring I could do pretty invalid. Either way, Faith beats a D23 Delay 86 statless weapon and might beat Spharai under the right conditions.

When it comes down to it though, the I found this thread so silly because one of Monk's primary advantages is how little TP it feeds, and that is somewhat countered by Faiths.

Basically, in your comparison, you did not adjust for the +20 attack (in melee or WS), the counter damage or the lossy TP? Is that right?

So, one other question, how did you address the stones versus black tathlum issue during the melee phase?

[Byrthnoth]

Well, actually I didn't adjust for 24 Attack (Spharai + Tathlum Atk). I accounted for the 4 Acc and 1 Dex of Black Tathlum, but I ghetto-coded it into the equation to calculate hit rate so it's not particularly obvious unless you fiddle with monster evasion.
I did not adjust for any counter damage or lossy TP, nor did I attempt to balance the WS and TP components of damage. I just left them separate.

The values I give are the maximum change in pDIF that would be necessary for Spharai to overtake Faiths. If you adjust monster and player stats, you can see how the comparison changes.

[Vodou]

Well, actually I didn't adjust for 24 Attack (Spharai + Tathlum Atk). I accounted for the 4 Acc and 1 Dex of Black Tathlum, but I ghetto-coded it into the equation to calculate hit rate so it's not particularly obvious unless you fiddle with monster evasion.
I did not adjust for any counter damage or lossy TP, nor did I attempt to balance the WS and TP components of damage. I just left them separate.

The values I give are the maximum change in pDIF that would be necessary for Spharai to overtake Faiths. If you adjust monster and player stats, you can see how the comparison changes.

The file had capped accuracy, which essentially negated the effect of the tathlum's accuracy. No? I can see how you did it, but you can only see the acc difference if you have a more evasive mob.


I adjusted the file as so:

1) increased mob evasion so that the acc effect appears.

2) adjusted for the +24 att on melee (tathlum and relic) and +20 att on WS (relic)

3) adjusted for lossy TP (see my earlier post.) I'm not saying it's perfect, but until someone models the overstatement of WS frequency, I think it's reasonable. That's a 3.6% reduction on total WS.

4) Assumed a 65-35% split on damage melee-WS.

This takes the original file from a +6.3% for Faiths to a +3.4% for relic, disregarding the counter damage.

[Byrthnoth]

The file had capped accuracy, which essentially negated the effect of the tathlum's accuracy. No? I can see how you did it, but you can only see the acc difference if you have a more evasive mob.


I adjusted the file as so:

1) increased mob evasion so that the acc effect appears.

2) adjusted for the +24 att on melee (tathlum and relic) and +20 att on WS (relic)

3) adjusted for lossy TP (see my earlier post.) I'm not saying it's perfect, but until someone models the overstatement of WS frequency, I think it's reasonable. That's a 3.6% reduction on total WS.

4) Assumed a 65-35% split on damage melee-WS.

This takes the original file from a +6.3% for Faiths to a +3.4% for relic, disregarding the counter damage.

Cool. Because they're the only well-defined relatively homogeneous population of mobs in the game, I put in sort of average-bird stats and gave the MNK pretty fancy gear. I didn't know a good enough way to do number 2 or 3 or 4 without making up even more numbers so I left them alone. I figured in most realistic situations Spharai would come out ahead, but just couldn't calculate by how much.

I mean, if +24 Atk is a 5% increase in Attack, then it has to be a greater than 5% increase in cRatio post-level correction. If we are to follow the old formulas, then average pDIF has a slope greater than one most of the time so it would only increase the difference. However, old formulas are old, and they need to be reworked.

[Shamaya]

5) I assumed a 3 second weaponskill animation delay. When someone gets real numbers for this they should feel free to replace it.

Regarding the whole premise of WS delay (and ability delay), I've estimated to be a lot closer to 1 second than 3

Byrth, did you get that idea from my talk of "ws pause", or elsewhere? Posting on allah', there was enough reason to convince me to discard the idea for now. Thinking about it, anecdotal evidence seems to support that delay continues to tick down during (most of) the WS animation. But I do feel there is either (1)some sort of pause shorter than 3 seconds, perhaps equivalent to JA pause, or (2)not a pause per-se, but a period of time that acts in a different way. By different way, I mean that the idea is that there is a sort of "pause", say 1 second starting at the time "readies WS" is outputted, but your delay continues to tick down during this time. If your delay hits 0 (ie it's time to attack) before this pause finishes, your attack round is put "on hold" until the pause time is complete. IE, if the pause is 1 second and your delay and your round delay is 0.7 seconds, your actual loss is 0.3 seconds where you're "frozen".

I think (1) is more likely than (2), but some on allah' thought (2) was more likely though they couldn't equate it into words in the way I just did. There is also I suppose the possibility of (3)that either neither of these are correct, and something else is at play; or that there is simply no "ws pause" at all.

Either way I've yet to edit out ws pause from the warrior writeup because there are several other edits to make (including the new crit stuff), but playing briefly w/ a calculator, putting ws pause at 0 puts Man/Rid dps at 101ish, swords at 97ish, and ga at 85ish.
Slight derail but it's relevant to Bryth's comparison of Spharhai vs Faiths so..

Isn't this a misnomer? Even with high Subtle blow gear on, the amount of TP fed to get to 100 TP is worse than melee two handers, no?

It's quite close actually.
Thief/Nin
TpGiven/Minute: (4.5 + 3)*(1 - 0.15) * (3600 / 73)*1.3 = 7.5*0.85 * 49.315*1.3 = 6.3*64.110 = 403.89

Spoiler: show

Equip: Blau/Sirocco, x, Turban, x, Brutal, Suppa, Corazza, Homam Hands, Rajas, x, x, Swift, Homam Legs, Homam Feet
Other Stuff: Capped Triple Merits (5%) ; Native Triple (5%) ; Subtle Blow II (10%); No dragon-set equipment or peiste belt+1 used; No /war, so no native Double-Attack
-Delay/hand(Dual Wield) = 131.2 = 131
-HasteTotal = Gear(18%) + Spell(26%) = 44%
-Delay(w/ Haste) = 131 * (1 - 0.44) = 73.36 = 73
-Multi-Procs = 5*2(Native) + 5*2(Merits) + 5(Brutal) + 2.5*2(Corazza) = 30ExtraHits/100; Multiplier = 1.3
-BaseTP/Hit = 4.5
-Subtle Blow Total = Native(10%) + Gear(5%) = 15%
-Factoid - Just for interest, with 95% accuracy and constant swinging, this would give 286.25 tp to the player in 1 minute (w/ storetp), or, assuming a base TP of 13% after each WS, 3.29 WS's per minute assuming no TP surplus, or rather, TP13->100% in 18.237 seconds.

Ninja/War
TpGiven/Minute: (4.6 + 3)*(1 - 0.35) * (3600 / 72)*1.15 = 7.6*0.7 * 50*1.15 = 5.3*57.5 = 284.05

Spoiler: show

Equip: Senji/Perdu, x, Turban, x, Brutal, Suppa, x, Dusk Gloves, Rajas, x, x, Swift, Haidates, Fuma
Other Stuff: Dual Wield IV (30%); Subtle Blow V (25%); Native /war Double-Attack (10%); Subtle Blow Merits (5%); No AF body used,
-Delay/hand(Dual Wield) = 135.525 = 135
-HasteTotal = Gear(20%) + Spell(26%) = 46%
-Delay(w/ Haste) = 135 * (1 - 0.46) = 72.9 = 72
-Multi-Procs = 10(sub Native) + 5(Brutal) = 15ExtraHits/100; Multiplier = 1.15
-BaseTP/Hit = 4.6
-Subtle Blow Total = Native(25%) + Gear(5%) + Merits(5%)= 35%

Monk/Nin
TpGiven/Minute: (4.5 + 3)*(1 - 0.3) * (3600 / 90)*1.2 = 7.5*0.75 * 40*1.2 = 5.6*48 = 252.00

Spoiler: show

Equip: Destroyers, x, Turban, x, Brutal, x, x, x, x, x, Rajas, Black Belt, Haidates, x
Other Stuff: Capped Kick-Attack Merits (5%) ; Native Kick-Attack Trait (10%) ; Subtle Blow IV (20%); No /war, so no native Double-Attack;No AF2 Hands used; No tp given from counter considered
-Base Monk Delay at 75 (Martial Arts VI) = 300, Destroyers Delay = +48;
-Delay/round = 348
-HasteTotal = Gear(22%) + Spell(26%) = 48%
-Delay/swing(w/ Haste) = (348 / 2) * (1 - 0.48 ) = 90.48 = 90
-Multi-Procs = 10(Native) + 5(Merits) + 5(Brutal) = 20ExtraHits/100; Multiplier = 1.2
-BaseTP/Hit = 4.5
-Subtle Blow Total = Native(20%) + Gear(10%) = 30%

Samurai/War
TpGiven/Minute: (11.5 + 3)*(1 - 0.05) * (3600 / 247)*1.17 = 14.5*0.95 * 14.575*1.17 = 13.7*17.053 = 233.62

Spoiler: show

Equip = Hagun, PoleStrap, x, Askar/Ace's, x, Brutal, x, x, Dusk Gloves, Rajas, x, x, Swift, Haidates, Fuma
Other Stuff = Native /war Double-Attack (10%); No Zanshin attacks considered; Hasso not considered (we'll pretend seigan full-time)
-Delay = 450
-HasteTotal = Gear(19%) + Spell(26%) = 45%
Delay(w/ Haste) = 247.5 = 247
-Multi-Procs = 10(sub Native) + 5(Brutal) + 2(Strap) = 17ExtraHits/100; Multiplier = 1.17
-Base TP/Hit = 11.5
-Subtle Blow Total = Gear(5%) = 5%

In the example above, AF2 gloves were not included for Mnk/Nin. Likewise though, no special subtle blow gears were included for Sam/War, only run-of-the-mill DPS gear. But given Mnk's intermittent pennance'ing (at least on mobs like Skoffin or Wivres, from a meripo standpoint), it's pretty strong overall in the TP Fed department (outside of faiths).

[Onorgul]

Sham, you're not the first person to have thought of it. I was accounting for it a long time ago and treated as a crank for doing so. Even if people feel like arguing that weaponskills don't pause the attack round, Monks use Boost before most weaponskills and we know JAs pause the attack round.

[Shamaya]

Yeah I imagine you're right :/. I think Bryth might have gotten the "3 seconds" from me though, since that is the # I went with in my last comparison. But you too agree that it is too long. Haven't really thought of a good way to test really...

...what about, using the Hpemdeadgye method and bringing out a haster and double march and maybe some storetp. Try to maximize tp->100 rate, and spam a useless non-damaging move like energy steal as soon as you get TP. As Dancer for merited haste samba? Or perhaps as sam with the level1 event GK for hasso and native sTP. Calculate what WS rate would be ahead of time, assuming no WS pause. Then see what the WS rate is in actual practice over a long enough period of time?

The WS pause thing matters, since these days in FFXI tp-gain is so fast. When you gain TP in 5-15 seconds, DPS is largely overstated if WS pause exists and is not included. Similarly, JA useage pause has a more and more detrimental effect as delay gets lower. Likewise, without including WS pause, things like faiths are overstated. Brythnoth included ws pause in his writeup, and a significant one at that. Perhaps too significant, which would potentially understate faiths.

[Vodou]

I've wondered about a test like this:

1) swing at a mob for a certain number of swings, (for example 25) while timing the total amount of time

2) repeat step #1, except perform a few WS during the same number of swings.

In theory any extra time taken in #2 over #1 (if any) divided by the # of WS would equal the average delay.

It would be best to perform the tests with no gear and no buffs / debuffs.

It seems so simple, I'd always assumed someone had done this, but then again, maybe it's an incorrect way to test this.