Final Form Magian Weapons

[Darkyrm]

The procrate shown in his sample is close enough. You would know this if you have done this stuff before. if the rate he got was like 10% I would say something was off, but he's within 4~% of the previously assumed procrate with a rather low amount of hits. It's the same.

Its still closer to 40% and since there still no bigger sample size Magian test or Joy test to prove "It's the same" its not valid.

If you are so sure just wait till more extesive testing or just say "They should be the same" trying to use Math reasoning without actual tests to back up is a moot point.

[NynJa]

Beating hagun is one thing - plenty of weapon already do that.

Did no one notice this?

[Byrthnoth]

Well, idk if I'm right about this, but I don't see the OAT GKT being vastly superior to Hagun.

Assuming a 6-hit for both and discounting DA because we don't yet know how it factors into this, you're looking at 3.777 attacks rounds per WS for the new GKT and 5 for Hagun. New GKT has 28.3% higher WS frequency. DA not stacking Jailer weapon style will push this advantage down a bit little when you factor in /WAR and Brutal.

WS hard-ness assuming 144 total STR (80 base + 64) varies as follows:
D68, 464 delay, 89 from 75% STR, +15fSTR - 1.6625 fTP = 285.95
D75, 450 delay, 89 from 75% STR, +16fSTR - 1.975 fTP = 355.5
Hagun WSs 24.3% harder. Use of Sekka will push this down.

So you're looking a 3% increase in damage from using the new GKT, but you have to factor the benefit of that against all the extra TP you're feeding. 15% DA would increase Hagun's TP gain significantly more than the new GKT assuming it isn't Jailer style, but Sekka makes my WS estimations a little exaggerated in favor of Hagun. Worse WS gear will favor Hagun to a point, but not if carried to an extreme. I'd be entirely unsurprised to find out Hagun wins against this GKT in most situations.

I didn't factor in differences in Haste, because if I'm not mistaken now people advocate Hagun with a full Haste build and Sword strap? I also didn't include the 5-hit mentioned above because it sacrifices a potential 5% Haste and Sword strap and isn't worth it (3.35 rounds/100TP vs 3.777, and >50% Haste the tradeoff doesn't work.)

As a tip to those looking to check my math, the easiest way to do the probability distributions is to calculate the probability of a 3-hit 100TP case (3 doubles, 2 doubles and a single) and the probability of a 5-hit 100TP case (4 single hits), then subtract them from 100%.

Edit: I guess what I can conclude is that /DRG with the new GKT (5-hit it) would be better than /DRG with Hagun.
Doubledit: If I factor 15% in DA not stacking like a Justice Weapon and instead acting like Joyeuse, then the WS frequency and WS strength differences almost perfectly cancel and Hagun = the good OAT GKT (not counting WS delay or Sekka). Looks like this is a weapon mostly for /DRG, not that I wouldn't sell my hypothetical Hagun if my hypothetical SAM got it.

[Thorny]

Except it's not that good. And Mandau/X's isn't even king-crud either (evenish with blau at best). 31/211*60*1.624 = ~14.3 DPS, and that's assuming 50% non-jailer OAT. That's barely more than several alternatives and doesn't even go into att/str/acc/etc. If you do the tp-gain math, it's barely faster than Sirocco for this. Plus single-wielding for DMG on mobs merit level and below, even with Mandau, is fail for DPS. So I wouldn't single-wield this .

Ridill is great for thief (probably more useful than with any other job). It doesn't cease to surprise me how often I'm fighting something EM/DC or below. Even in merits with current gear (cuchulain's, anwig) it's very close to x's/blau for mandau offhand, tho I don't use it.

So no I can't see it having very much use . Just hoping it's jailer-style OAT but I doubt that will be the case.

didn't mean mandau + x's DWed, meant own either of them(and x's is debatable here anyway)

blau/sirocco's dps beats it single wielded(~17.5)

tp gain for athame is 0.04387 per 1 delay
blau/sirocco's is 0.03498 per 1 delay
c/p from calculator so more precise than needed, but eh.. it comes down to 1 haste, 2 chr, 27.9% ws frequency(before considering haste) vs 22.3% damage while tping, 16 att, 3 acc

calculating x-hit and probability of overflow may slightly skew results, but from what i can see athame single wield is the option of choice for people without a mandau before even considering the option to use subs other than nin if viable

[Thorny]

Did no one notice this?

nobody pointed it out because soboro beats hagun easily and pachi/av gk can pull ahead depending what gear is available to you

[Shamaya]

lolWut. Don't you mean Tomoe or Amano (buffed attack merit-burn style situations), Thorny? Soboro for salvage style fodder mobs, tho, f'shizz. Also isn't rindo or w/e more viable than pachi for voodoo 5-hit setup (I'm not a sam so Idk this bit)?

Also, the whole WS-pause thing we discussed last time completely throws x-hit build vs hagun comparisons for a fuck-loop. The math needs to be completely reworked if people think it's a close call. But TBH that really deserves its own thread.

[Thorny]

lolWut. Don't you mean Tomoe or Amano (buffed attack merit-burn style situations), Thorny? Soboro for salvage style fodder mobs, tho, f'shizz. Also isn't rindo or w/e more viable than pachi for voodoo 5-hit setup (I'm not a sam so Idk this bit)?

Well, I assumed everyone knew Tomoe/Love/Amano will destroy Hagun for situations where you can use your attack. Soboro/Rindo/Pachi/AV all have potential to beat hagun in gekko situations depending on your specific gear available.

[ringthree]

Please tell me others agree with this. Need to know if I should sell Hagun =P

It's mdkuser. I would not accept what either pchan or mdkuser say about anything.

As a reference see the Rindomaru thread. mdkuser does not consider scaling factors that favor Hagun in that discussion.

In general, pchan and mdkuser start from a conclusion and then back track through the data to make the hypothesis. The fact that mdkuser still says that Rindomaru is 3% better than Hagun should throw all of this into question.

Rindomaru IS favorable over Hagun against, and ONLY against, Greater Colibri. As defense and levels skew against the player, the Hagun quickly surpasses Rindomaru. You can not make a conclusion about the quality of a weapon in all situations from just one example.

I have not yet done the math for the new OAT GKT, but I will work on that one day next week. I presume nothing either way about each of these weapons, I just know that I don't trust mdkuser or pchan's math, because it never adds up.

[Shamaya]

For reference, since more people should know about this, here's a link to most of what we know so far about JA/WS pause. And it's highly relevant to damage comparisons. Especially for very high WS-frequency builds, like the new OAT weapons we're talking about. I'd definitely give it a bit of a read b4 you do any math on OAT GK, Ringthree. Basic run down though is that JA/WS use = 2 seconds of freezed delay. Stacking 2/more JA/WS together (within a very short period of time) means a 2 second delay freeze for the first use, and 1 second for subsequent JA's in the stack.

[pchan]

It's mdkuser. I would not accept what either pchan or mdkuser say about anything.

Opinion

I have not yet done the math for the new OAT GKT

Knew it.

I remember you did maths of destroyers vs faith and found destroyers to win, so I don't expext anything serious out of you..

[mdkuser]

Well, idk if I'm right about this, but I don't see the OAT GKT being vastly superior to Hagun.

Assuming a 6-hit for both and discounting DA because we don't yet know how it factors into this, you're looking at 3.777 attacks rounds per WS for the new GKT and 5 for Hagun. New GKT has 28.3% higher WS frequency. DA not stacking Jailer weapon style will push this advantage down a bit little when you factor in /WAR and Brutal.

WS hard-ness assuming 144 total STR (80 base + 64) varies as follows:
D68, 464 delay, 89 from 75% STR, +15fSTR - 1.6625 fTP = 285.95
D75, 450 delay, 89 from 75% STR, +16fSTR - 1.975 fTP = 355.5
Hagun WSs 24.3% harder. Use of Sekka will push this down.

So you're looking a 3% increase in damage from using the new GKT, but you have to factor the benefit of that against all the extra TP you're feeding. 15% DA would increase Hagun's TP gain significantly more than the new GKT assuming it isn't Jailer style, but Sekka makes my WS estimations a little exaggerated in favor of Hagun. Worse WS gear will favor Hagun to a point, but not if carried to an extreme. I'd be entirely unsurprised to find out Hagun wins against this GKT in most situations.

I didn't factor in differences in Haste, because if I'm not mistaken now people advocate Hagun with a full Haste build and Sword strap? I also didn't include the 5-hit mentioned above because it sacrifices a potential 5% Haste and Sword strap and isn't worth it (3.35 rounds/100TP vs 3.777, and >50% Haste the tradeoff doesn't work.)

As a tip to those looking to check my math, the easiest way to do the probability distributions is to calculate the probability of a 3-hit 100TP case (3 doubles, 2 doubles and a single) and the probability of a 5-hit 100TP case (4 single hits), then subtract them from 100%.

Edit: I guess what I can conclude is that /DRG with the new GKT (5-hit it) would be better than /DRG with Hagun.

Its easy to factor in Double attacks as follows:

compute p0, p1, p2, the probability to land respectively 0, 1 or 2 hits. This depends on hit rate, DA rate, and OAT rate by formuals ive got here on a worksheet ( can include zanshin rate as well).

once you've got that you do your computations as if your new DA rate was p2. For example under caped accuracy ( p0=0), you've got p1 chance to do one hit, p2 chance to do 2 hits. p1+p2=1.

I'm guessing you end up with the same formula as mine?

capped accuracy: the probability distribution is

-case of 5 single attacks in a row, probability= p1^5 and R=5 rounds.
-case of one double attack, that is not last: proba=4*p1^3*p2 and R=4 rounds.
-case of one double attack last: proba= p1^4*p2 and R= 5 rounds
-case of two double attacks not last( none wasted): proba=3*p1*p2^2, R=3 rounds
-case of two double attacks, one wasted: proba=3*p1^2*p2^2 R=4 rounds
-case of three double attacks: proba= p2^3 R=3 rounds.

average rounds formulas:

5*p1^5+4*4*p1^3*p2+5*p1^4*p2+3*3*p1*p2^2+4*3*p1^2* p2^2+3*p2^3

the formula for p0>0 is not much different:

http://i224.photobucket.com/albums/d.../round6hit.jpg

so if I plug 100% accuracy 50% double attack into the formula:
P0=0, p1=0.5, p2=0.5

I find

average rounds= 3.56 not 3.77

so you forgot some cases!

[ringthree]

For reference, since more people should know about this, here's a link to most of what we know so far about JA/WS pause. And it's highly relevant to damage comparisons. Especially for very high WS-frequency builds, like the new OAT weapons we're talking about. I'd definitely give it a bit of a read b4 you do any math on OAT GK, Ringthree. Basic run down though is that JA/WS use = 2 seconds of freezed delay. Stacking 2/more JA/WS together (within a very short period of time) means a 2 second delay freeze for the first use, and 1 second for subsequent JA's in the stack.

I actually considered that fact, but left it out in the Rindomaru discussion because it favored Hagun, but I did make mention of it at some point.

Many of the discussions about X-hit builds leave out this important function, thank you for bringing it up!

[Nameless]

Well, idk if I'm right about this, but I don't see the OAT GKT being vastly superior to Hagun.

Assuming a 6-hit for both and discounting DA because we don't yet know how it factors into this, you're looking at 3.777 attacks rounds per WS for the new GKT and 5 for Hagun. New GKT has 28.3% higher WS frequency. DA not stacking Jailer weapon style will push this advantage down a bit little when you factor in /WAR and Brutal.

WS hard-ness assuming 144 total STR (80 base + 64) varies as follows:
D68, 464 delay, 89 from 75% STR, +15fSTR - 1.6625 fTP = 285.95
D75, 450 delay, 89 from 75% STR, +16fSTR - 1.975 fTP = 355.5
Hagun WSs 24.3% harder. Use of Sekka will push this down.

So you're looking a 3% increase in damage from using the new GKT, but you have to factor the benefit of that against all the extra TP you're feeding. 15% DA would increase Hagun's TP gain significantly more than the new GKT assuming it isn't Jailer style, but Sekka makes my WS estimations a little exaggerated in favor of Hagun. Worse WS gear will favor Hagun to a point, but not if carried to an extreme. I'd be entirely unsurprised to find out Hagun wins against this GKT in most situations.

I didn't factor in differences in Haste, because if I'm not mistaken now people advocate Hagun with a full Haste build and Sword strap? I also didn't include the 5-hit mentioned above because it sacrifices a potential 5% Haste and Sword strap and isn't worth it (3.35 rounds/100TP vs 3.777, and >50% Haste the tradeoff doesn't work.)

As a tip to those looking to check my math, the easiest way to do the probability distributions is to calculate the probability of a 3-hit 100TP case (3 doubles, 2 doubles and a single) and the probability of a 5-hit 100TP case (4 single hits), then subtract them from 100%.

Edit: I guess what I can conclude is that /DRG with the new GKT (5-hit it) would be better than /DRG with Hagun.
Doubledit: If I factor 15% in DA not stacking like a Justice Weapon and instead acting like Joyeuse, then the WS frequency and WS strength differences almost perfectly cancel and Hagun = the good OAT GKT (not counting WS delay or Sekka). Looks like this is a weapon mostly for /DRG, not that I wouldn't sell my hypothetical Hagun if my hypothetical SAM got it.

It all look good but you forgot that OAT GK will rape Hagun on DoT, and in the end the two won't even be in the same discussion.

[mdkuser]

The fact that mdkuser still says that Rindomaru is 3% better than Hagun should throw all of this into question.

the fact you still think rindomaru doesn't beat hagun proves you're still stupid. its been said many times that what makes rindomaru beat or not beat hagun is not "scaling" but the AFK factor: the more you afk the better the hagun w... doesnt really make it better. If you're full time on the mob, rindomaru pachipachio etc win.

[Byrthnoth]

The probability of getting TP in 3 rounds is: 35.2%
.4 (2) * .4 (2) * .6 (1) * 3 = .288 ( the *3 takes care of the 3 order-specific permutations that are possible)
+
.4 (2) * .4 (2) * .4 (2) = .064

The probability of getting TP in 5 rounds is:
.6^4 (4 consecutive 1-hit rounds is the only way to not get TP before 5 rounds) = 12.96%

Therefore, the probability of getting TP in 4 rounds is:
1 - 35.2% - 12.96% = 51.84%

35.2*3 + 51.84 * 4 + 12.96*5 = 3.7776

I may be wrong or may be systematically making the same calculator error, seeing as how I got about 3 hours of sleep last night, but I'm sleepily confident in that math. Oh, well the difference is that I'm using 40% OAT rate (per Mont's parses?) and you are using 50%.

PS. Sleepiness evident from DoT forgetfulness. Yeah, new GKT would win by approximately 23% in the TP phase.

[Shadida]

About to finish my evasion dagger! Oh wait.

http://files1.guildlaunch.net/guild/...ins7514Mid.png

FML

[pchan]

AFK factor: the more you afk the better the hagun w

We should redo the math of faith vs destroyers by including afk factor too. The more you afk the more your chi blast damage, so the mnd+7 on faith is helping more.

[Shamaya]

blau/sirocco's dps beats it single wielded(~17.5)

tp gain for athame is 0.04387 per 1 delay
blau/sirocco's is 0.03498 per 1 delay
c/p from calculator so more precise than needed, but eh.. it comes down to 1 haste, 2 chr, 27.9% ws frequency(before considering haste) vs 22.3% damage while tping, 16 att, 3 acc

calculating x-hit and probability of overflow may slightly skew results, but from what i can see athame single wield is the option of choice for people without a mandau before even considering the option to use subs other than nin if viable

Gotta state what conditions you're usin. ACP 3% DW? StoreTP included? If so, Skadi or no skadi? Did you assume jailer-style OAT or not? 40% D.A. assumed or something else?

You don't wanna single-wield this new dagger. The math involves a lot more than very skeletal TP/gain BD/delay considerations. I appreciate the contribution though. I'll edit in some of my own wonder mathz in this post in a bit.

Editz:

Just for comparison's sake: How does the new dagger compare to Sirocco's increase in TP Gain?
From the standpoint of starting from 211 delay. With 178delay 2nd dagger, DW&ACP, &+13sTP.

211&178 TP/hit = 5.3
211&178 TP/time = 0.0356tp/delay
150&178 TP/hit = 5.0
150&178 TP/time = 0.0397tp/delay

So in this scenario, going from a 211->150 delay offhand is increasing TP/time by 397/356, or 11.5%. (Calculating b4 sTP, this was 357/315 or 13.3%).

Reiterating b4 and revisiting the occ. atk. 2x on the new offhand dagger, we're increasing hits/round from 2.48 (2 daggers & 24% triple attack) to 2.86. So the increase in swings/time or TP/time = 15.3%.

So new dagger has a clear lead in speed ahead of all our other options (That aren't M.Kris, Ridill, or Kclub). In this example, the TP gain is 3.4% faster than using Sirocco .

/nin
This is some shiz I posted on the allah' forums. But if we're talkin single-wield,

Athame base TP/hit = 5.7. If we're talkin brutal/rajas/skadi, we get 6.4. Assuming 45% non-jailer D.A., brutal and merits, our multi-proc multiplier is (1.45 + 0.55*0.24 =) 1.582. So new avgTP/round = ~10.1. Or about 0.0479 tp/1delay.
As for shields, that's another matter I suppose, but if we do Assume Airy buckler and only factor in the haste for these purposes, assuming 20% equip haste, haste, 2xMarch, we get a marginal 2.2% increase in speed which'd turn it into 0.0490 for our purposes.

Blau/Athame (using this cuz Athame > Sirocco) TP/hit is 4.7. 5.3 after sTP. After multi-proc, ~7.48tp/hit or double that per round. TP/delay being ~0.050, or higher/faster than single-wield Athame.

***

Ugghh. Well obviously sub job is a big thing. Hmm. Normally /drg just doesn't cut it over /war, but Ima rethink this a second.

/war
If /war, multi-procs/round go up from 1.24 to 1.32. For athame single-wield 45% non-jailer, it'd go up from 1.582 to 1.626, and would change final tp/1delay after airy buckler haste to ~0.050, which at this point is conveniently equal to ablau/athame.

/drg
If /drg and we replace suppa w/ wyvern earring (we'd have to replace Dusk+1 w/ Homam or shield w/ something else), we'd get a marginal 4% haste, or 10% increased speed w/ the merit style settings used. This would bring comparative tp/delay to ~0.054. This would estimate Athame/Drg at 8% faster tp/time than Blau/Athame in ideal meripo settings.

[Thorny]

assumed suppa but no ACP DW(rapparee), took your DPS calculation for athame, and assumed 5/5 TA merits and brutal for calculating blau/sirocco's DPS, giving priority to triple attack over double attack if both proc

also forgot to give athame hollow earring, which would negate the last few acc from blau.. but it was just a quick calculation not a full evaluation

they're very close if you're using /nin regardless, if you don't need shadows and can /drg or /war then it's no contest

x's shouldn't change anything since x's/blau is very close to blau/sirocco, mandau might, haven't looked into it yet

[mdkuser]

The probability of getting TP in 3 rounds is: 35.2%
.4 (2) * .4 (2) * .6 (1) * 3 = .288 ( the *3 takes care of the 3 order-specific permutations that are possible)
+
.4 (2) * .4 (2) * .4 (2) = .064

The probability of getting TP in 5 rounds is:
.6^4 (4 consecutive 1-hit rounds is the only way to not get TP before 5 rounds) = 12.96%

Therefore, the probability of getting TP in 4 rounds is:
1 - 35.2% - 12.96% = 51.84%

35.2*3 + 51.84 * 4 + 12.96*5 = 3.7776

I may be wrong or may be systematically making the same calculator error, seeing as how I got about 3 hours of sleep last night, but I'm sleepily confident in that math. Oh, well the difference is that I'm using 40% OAT rate (per Mont's parses?) and you are using 50%.

PS. Sleepiness evident from DoT forgetfulness. Yeah, new GKT would win by approximately 23% in the TP phase.

ah ok you take 40% OAT not 50%, then we aggree my formula is the same as yours with p0=0, p1=0.6 p2=0.4

so if you want to include double attack work it as follows:

with 40% OAT and 17% double attack rate, your effective DA rate is

p2=0.4+(1-0.4)*0.17

if both DA and OAT cannot stack. if they can stacks itsmore complicated as tripple attacks can occur.

using the p2 above you find:

3.55 rounds untill 100 TP ( under 100% acc)

and 3.73 rounds with 95% accuracy. ( =almost 3.55/0.95)