Acc Dakini x2 VS Acc Dakini/BlauFollow

Is it actually possible to get two Acc Dakinis? Not trying to be snarky but I hear conflicting opinions on the possibility of having two of the same weapon.

Edited, May 9th 2010 2:49pm by Lucinus

In that case I lean very heavily towards Acc/Acc, assuming Pizza, but am willing to be proven wrong.

Edited, May 9th 2010 3:23pm by Lucinus

I'd still take the TD/Blau over TD/TD. Meldi's conclusions were the same as my own. The DoT aspect of TD/Blau is several percent higher than that of TD/TD and SA is a wash. Remember that 1 base damage equates to 0.5 base damage averaged per round but 12 delay and 16 attack is considerably better than 5 accuracy in pure DoT stats. WS can be iffy but it isn't too terribly sketchy if your accuracy is in the ballpark of where it should be (especially if you're considering ManStab since TD/TD leans to favoring multi hit whereas TD/Blau favors everything well). When considering weaponskill frequency the increased tp gain of the TD/TD through 5 accuracy is countered by the faster swings of 12 less delay from TD/blau, and in weaponskill damage itself dex mods and accuracy is weighted against the approximate 5% average pDIF increase of 16 attack. The offhand swing getting 1 base damage more is so microscopic it's irrelevant to this calculation. TD/TD can pull ahead in weaponskill damage, but overall the margin is too small for me to count on it with the bigger picture in mind.

I've played with numbers myself and I'll agree it's pretty close, but I'll take the TD/Blau personally. The closing argument is that the comparison REQUIRES the use of the extra accuracy. If your accuracy overshoots the cap the TD/Blau will be hands down better than TD/TD in every department save SA (again, being a wash). I prefer the consistent performance increase of the TD/blau over the TD/TD because I know it'll be useful 100% of the time whereas TD/TD only stands a chance when acc is uncapped, and even when it is my "big picture" figures lean towards TD/Blau anyway. I consider TD/TD a gamble whereas TD/Blau is a sure thing and I'd prefer the sure thing myself. Maybe that's just me though? But I doubt I'm alone with that train of thought.

Edited, May 10th 2010 12:48pm by Melphina

Azoth was close to SK with +9 dmg +4 acc and 210 delay. Thunder Dakini is hands down better than Azoth, so it should beat SK. If we are comparing it to Blau and saying "Oh, its close!," there really shouldn't even be a discussion about whether it is better than SK.

Well, the reason Azoth lost out to SK, is because there was a 60 delay difference between it and SK. TD loses 1 base damage on Azoth but makes up 33% of the delay between Azoth and SK. So the difference in damage is 8 (31% higher or ~28% with fSTR), the 10 acc would more than make up for that. If you are talking about capped acc, well then that definately pushes the scales to SK. But if you make use of the acc, then TD is going to win in the long run.

Basically, SK = 27% lower delay vs. TD = 28% higher damage and +5% hit rate.

Meldi, if you're refering to the imfamous (and often refered to) azoth vs sirocco thread some moons ago Melph did, the initial conclusion favored azoth iirc, but completely ignored TP gain in the equation. It was then was updated later in the thread and the pure TP speed of sirocco just obliterated all the other advantages. This was also long before mirke with another 3% DW and no change to TP values which will just add to the weight of sirocco. It didnt end up really being very close once things were better accounted for. It seems this thread is mentioned a lot referencing the original conclusing which was turned on its head later.

This is better than azoth, but azoth wasnt really close once the speed factor for TP is considered. I feel about the same as melph on this one. If you REALLY need teh acc it will proably be pretty close, but favor blau anyway due to the speed TP/DPS/Atk edge it has. And in many situations the acc is overkill anyway and completely variable based on gear/mob/merits/buffs. Blau is always a sure bet and generally edge out dakini anyway. Bonus to SA is nice and all, but its such a small part of our damage, it just doesnt add up well. Our damage comes from DPS and TP gain primarily. Our WS numbers are good, but the strength of our WS is more focused on the frequency based on our super fast TP gain. Dakinis are good all-round daggers to be sure, but Blau/Sirocco just compliments thf so well in every way its still the top dog with the slight edge lost to blau/new-DA-Dagger which has the same concept of speed just at a new level.

Banalaty: I thought the final conclusion of that thread was that Azoth was decent, but it just couldn't quite catch the SK because of the delay difference. I will go look again, but I thought that it was still close, just everything favored the SK. If it is really miles apart, then there should be no way that X's knife can compare to SK either.

Kalisa, if you look down the page a bit, I did an analysis of TD vs. FD and the basic conclusion was that FD doesn't win until you have capped acc. So your test would have to be at capped acc. At that point, I would say FD has a better chance of beating SK than TD does, but I can't say for sure.

Banalaty: I went back and found the Azoth vs. SK thread, and this is Melph's Final Conclusion



It looks like it is still close, but SK edges it out. I would say that that is a good argument for the possibility that TD will beat SK, but someone with more time would have to do the math.



I don't mention it much, because I do not have one, and probably will never have one and I just don't know about that aspect of it. I would like to know though, what kind of damage procs on things like HNMs and stuff, cause that is mostly what I fight, with some limbus and einherjar which should for the most part have good procs on the damage.

Edited, May 11th 2010 10:04am by Meldi

Actually, I usually go /nin to most things because due to collaborator, I usually end up tanking at least for a little bit on protracted fights.

It probably does, your definition of things is usually subtly different than what other people think.

Anyway, here is some info on B/SK vs. TD/B.

Base, 17% Haste, 90% acc, 1.000 cRatio, +6 sTP, +3 fSTR. Hypothetical WS @ 100 tp exactly every time. No TP return, 1st hit Always lands.

Time = 3600 seconds (1 hr)
Hits:
B/SK = 1028 rds x .9 = 925 rds /hit
TD/B = 923 rds x .95 = 876 rds /hit

TP:
B/SK = 925 hits x 2 x 4.7 tp/hit = 8695 tp
TD/B = 876 hits x 2 x 4.8 tp/hit = 8409.6 tp

DoT:
B/SK = 60125
TD/B = 63948

WS: Dancing Edge Acc 90%, WSC 40, fSTR 10, cRatio 1.000
B/SK = 86 WS, 48 6-hit, 23 5-hit, 10 4-hit, 4 3-hit, 1 2-hit
Damage = 38885
TD/B = 84 WS, 61 6-hit, 18 5-hit, 4 4-hit, 1 3-hit
Damage = 41574

Conclusion:
DoT: TD/B @ +3823 [+6.4%], (TD/B @ +123 if Wind Damage proc avg is +4)
WS: TD/B @ +2689 [+6.9%]

Caveats:
1. Not factoring TP return favors the slower TP gain of TD/B, could skew results of WS in favor of B/SK
2. Neither set can get exactly 100 tp in TP phase, so WS total is not exactly accurate.

Case 1: 32% Haste, 90% acc, 1.000 cRatio, +6 sTP, +3 fSTR. Hypothetical WS @ 100 tp exactly every time. No TP return, 1st hit Always lands.

Time = 3600 seconds (1 hr)
Hits:
B/SK = 1250 rds x .9 = 1125 rds /hit
TD/B = 1125 rds x .95 = 1068 rds /hit

TP:
B/SK = 1125 hits x 2 x 4.7 tp/hit = 10575 tp
TD/B = 1068 hits x 2 x 4.8 tp/hit = 10252.8 tp

DoT:
B/SK = 73125
TD/B = 77964

WS: Dancing Edge Acc 90%, WSC 40, fSTR 10, cRatio 1.000
B/SK = 105 WS, 59 6-hit, 28 5-hit, 13 4-hit, 4 3-hit, 1 2-hit
Damage = 47642
TD/B = 102 WS, 74 6-hit, 22 5-hit, 5 4-hit, 1 3-hit
Damage = 50501

Conclusion:
DoT: TD/B @ +4839 [+6.6%], (TD/B @ +339 if Wind Damage proc avg is +4)
WS: TD/B @ +2859 [+6.0%]

Case 2: 52% Haste, 90% acc, 1.000 cRatio, +6 sTP, +3 fSTR. Hypothetical WS @ 100 tp exactly every time. No TP return, 1st hit Always lands.*

Time = 3600 seconds (1 hr)
Hits:
B/SK = 1773 rds x .9 = 1595 rds /hit
TD/B = 1585 rds x .95 = 1505 rds /hit

TP:
B/SK = 1595 hits x 2 x 4.7 tp/hit = 14993 tp
TD/B = 1505 hits x 2 x 4.8 tp/hit = 14448 tp

DoT:
B/SK = 103675
TD/B = 109865

WS: Dancing Edge Acc 90%, WSC 40, fSTR 10, cRatio 1.000
B/SK = 149 WS, 83 6-hit, 40 5-hit, 19 4-hit, 6 3-hit, 1 2-hit
Damage = 67556
TD/B = 144 WS, 105 6-hit, 30 5-hit, 7 4-hit, 2 3-hit
Damage = 71245

Conclusion:
DoT: TD/B @ +6190 [+6.0%], (TD/B @ -190 if Wind Damage proc avg is +4)
WS: TD/B @ +3689 [+5.5%]

*I found that if I truncated too much in the delay, it heavily favors SK, so in the last calculation I extended sec/rd to the 100th decimal place. I.E. Blau/SK @ 23% DW, 52% Haste is 2.03 sec/rd, not 2.0 sec/rd.

FINAL CONCLUSION: If you are making use of TD to its fullest, even in a worst case senario (90-95% acc) it should win against SK. If you have capped acc, I would stick to SK. Higher haste favors SK if you are making use of the wind damage proc. But if not, then TD widens its gap.


Edited, May 11th 2010 1:14pm by Meldi

Meldi Sage wrote:



Meldi, I'd like to have this checked by someone else. As far as I understand my thief from watching tp, I always end up with 9 tp after 2 attacks. I may get 4 tp if I miss one hit in a round but the next hit will always put me at 9 tp.

So, after 925 hits, I will not have accumulated 8695 tp and after 876 hits i will not have accumulated 8409.6 tp. Your calculations is favoring TD/B a lot more since the game doesn't count tp the way you posted.

Meldi wrote:



As something in this game to obtain for thief, I think Sirocco Kukri is up there just because paired with Blau it makes you go O.o

I remember when I first put it on, it was such a huge difference on the speed. Thief was actually my first DD job. I stopped exping just so I would ding in SK and on BD as well. It was a huge difference from past daggers.

It doesn't make one gimp by not having it, but in my opinion it makes thief a lot more fun to watch.

Is JoL not an HNM? Kirin? Odin? T4ZNM? T3-4 VNM? Wyrms/KV/Khim(no one "camps" those. Its Mule watch+who can gather fastest :P) (Insert Boss/Event/Popped mobs of high level here) etc. "High" NM doesnt have anything to do with the pop conditions. Its just a 'boss' type mob that is generally very high level and therefore more difficult. My LS virtually never touches HNM 'camps' because they usually end up being a monumental waste of LS time and most drops are available through other means save a few 'trophy' items like Dring/Ridil or are overated junk (hi2u lolAlgol/Haute etc). But we DO kill lots of HNMs. We just dont go through the dumb hours/drama/stress of camps only to loose to the heavy bot shells.

So yeah, the definition actually does affect your "rightness" on HNMs. I RARELY need to have to /job for teh sake of claiming but still fight plenty of hard stuff.

I had 50% written out a more detailed response about DK/B/S combos but my laptop ate it and i gotta get up real early tomorrow. So.....to bad :P maybe later.

Edited, May 11th 2010 11:23pm by Banalaty

It isn't 925 hits, it is 925 rounds hit, so actually 1850 hits, hence I did 925 x 2 x 4.7 tp/hit.

As for the TP, the reason you see the 4 -> 9 phenomenon is because the TP display truncates to the nearest integer on TP, but the actual TP stored is truncated to the 10th decimal place, so one hit is 4.7 tp, the display truncates it to 4, but when you hit that second hit it it goes up to 9.4 then truncates it to 9 for the display. The next hit will raise it to 14.1, truncating to 14 and so on.

Regardless, Noodles is right, in that Tp/hr is not an accurate model. But really, there isn't an accurate model. Unless someone can accurately predict how much exact TP is lost every time due to over TPing. Yes I am not accounting for JA pause or anything else, but this at least gives you a good idea of the potential. Of course, Noodles, if you have a better way, I would welcome it.

And my view of HNM more closely follows Banalaty's view. I knew what you were saying, and I knew what I was saying, but you seem to like to argue, so I just left it at "probably our views differ."

TP gain via STP, and thus via Dual Wield delay- is a long winded calulation, but for accuracy, it is neede to be done.

I will give an example, one I used on BG:

This example uses 4.5% TP with BD/SK, up to 5.0 using Suppa/DW Mirke, and Rajas/Skadi/Brutal

This should show you why % increases based on TP are innaccurate and should not be concidered. If you really care about thief, I'm sorry to dissappoint you but you will need to take the long winded approach to find an accurate answer to the increase +STP gives.
DA/TA procs are not concidered within this calculation for the following reasoning: Over time, concidering the 95% Acc cap, 10% TA and 5% DA, let's assume you get eactly 5.0%TP per hit, you should average 100TP/WS and 105TP/WS about 50/50. Thus when STP saves one hit, not one round, the most realistic thing we cn do is calulate it as a round saved, and half the resulting % of it happening.

Here, now for STP uses, (which can be taken and applied to DW or dely/tp differences) is an example:
The main problem with TP/Hr, and all calulations derived from it, is that TP is reset when you WS. You can not use "X/Y rounds" where Y =/= 0. You never get (for example) 10.7 rounds. You either get 10 or 11.
This is a table showing the possible DE TP reutns, and the resulting TP/hit needed to reach 100TP with each +1STP added:
http://i22.photobucket.com/albums/b330/noodles355/tptbl1.gif
If we simply divide them by each other, at every starting TP, adding +1STP gives an increase of 2% > 0% > 2% in going to 14, 15 and 16STP.
This is what many would consider to be ccurate when they do not actually consider each individual case of STP doing anything (and then finding the % chance it will at that number and using that as an accurate means)

This second table, with the exact same values used, shows the actual numbers the game uses (because hits/10 can not be decimals, you can not have 9.5 hits to 100TP, 9.5htis is 10hits)
http://i22.photobucket.com/albums/b330/noodles355/tptbl2.gif
Here we can see in going ffrom 13>14STP, you save one hit twice, for 5.6 and 5.9% increase respectfully, overall then that is 1.64% increase.
At 14>15, it is 0%, obviously. At 15>16, it is 5.9 in one instance, or on average 0.84% increase.

So that's 1.64% > 0% > 0.84% increases, NOT 2% > 0% > 2%.

Except, because you're dual wielding and saving one hit, not one attack round, those hits saved will only happen half of the time, so the actual increases to WS frequency is 0.82%STP going from 13>14 and 0.42 going from 15>16STP.

So whilst you would calulate the increase in WS frequency in 13>14STP & 15>16STP in TP after dancing edge as 2% increase both times, they are actually closer to 0.82 and 0.42 respectively.

Store TP (And through it, delay and TP/hit in general) can not be simply calulated as TP1/TP2 because this is not accurate. In actuallity you must consider:
If it can save a round at all
If it can, how many times out of possible scenarios it will
The % increase for each case
And the total average increase to WS frequency through all those points

For example, if your dagger choice changes your delay per hit from 4.7 to 4.8, you can not do a simple calulation to see the % increase in TP. You must claulate every possible instance based off DE returns and calulate the % increase to DoT that will be gained if it does change your TP/hit, and then how many times out of each possible scenario it actually does, and then use that resulting % as a realistic value for the DoT change in changing the delay.

Edited, May 12th 2010 7:44pm by Noodles

It was nothing against your post in perticular, I think yours holds a lot of weight, it was against the general idea of TP1 / TP2.

However, my model is still not accurat as I did not concider the % chance of each insance occuring (s for example, the chance of a 11, 12 and 16% TP reutn WS happening are greater than a 9 or 15% happening (if you don't know why it's because with 95% acc, there's a greater chanc of you landing more hits. as for the 15 vs 16, 15=DA proc or ta proc with a miss, 16=TA proc, and with 95% accm theres a greater chance of 16% TP return than 15%)

OK, not too difficult to calculate average number of swings per weaponskill.

DA chance: 5%
TA chance: 9.5%
1 hit chance: 85.5%


DA mainhand: 5%
~ DA offhand: 5% -- 0.25% chance of 8
~ TA offhand: 9.5% -- 0.475% chance of 8
~ 1 offhand: 85.5% -- 4.275% chance of 7

TA mainhand: 9.5%
~ offhand doesn't matter; 9.5% chance of 8

1 mainhand: 85.5%
~ DA offhand: 5% -- 4.275% chance of 7
~ TA offhand: 9.5% -- 8.1225% chance of 8
~ 1 offhand: 85.5% -- 73.1025% chance of 6


Overall:
6 hit: 73.1025%
7 hit: 8.55%
8 hit: 18.3475%

Average number of swings: 6.45245


Chance of a full 8-hit connecting at 95% hit rate: 12.17%


Hits to 100 TP (taken from the non-Mirke options to keep the number of permutations down):

@4.7 TP/hit: 18 hits needed for full 8-hit, 19 for most others, +1 for each missed full-TP swing (0.77% chance of each). Avg of ~18.89 hits to 100 TP.

@4.9 TP/hit: 18 hits needed for 4-hit, +1 for each missed full-TP swing (0.77% chance of each). Avg of ~18.02 hits to 100 TP.

@5.0 TP/hit: 17 hits needed for 7-hit (24%), 18 for all others, +1 for each missed full-TP swing (0.77% chance of each). Avg of ~17.78 hits to 100 TP.

@5.3 TP/hit: 16 hits needed for 7-hit (24%), 17 for all others, +1 for each missed full-TP swing (0.77% chance of each). Avg of ~16.78 hits to 100 TP.


2.48 swings per round, so .74 over-TP swings on average. Rough guesstimate of 30% chance of an additional attack round landing before weaponskill, for another 0.744 swings.

1.48 swings added to the above for:

@4.7: 20.37 hits, or 8.65 rounds @95% acc
@4.9: 19.50 hits, or 8.28 rounds @95% acc
@5.0: 19.26 hits, or 8.17 rounds @95% acc
@5.3: 18.26 hits, or 7.75 rounds @95% acc


Time per weaponskill: Rounds * Round delay + 120

 
B/SK - 263 DW delay 
TP/Hit    17% haste   32% haste   52% haste 
4.7           2,008       1,667       1,212 
5.0           1,903       1,581       1,151 
 
TD/B - 295 DW delay 
TP/Hit    17% haste   32% haste   52% haste 
4.9           2,147       1,781       1,292 
5.3           2,018       1,675       1,217 
 
 
Ratio (TD/B) / (B/SK) 
TP/Hit    17% haste   32% haste   52% haste 
Base           6.9%        6.8%        6.6% 
Skadi          6.0%        5.9%        5.7% 

So the ratios there show the amount of additional damage per weaponskill cycle (total melee damage for the given number of hits, plus average weaponskill) that the TD/B config has to do in order to maintain parity with the B/SK config.

Simply plugging in the numbers that Meldi used:

Avg B/SK melee hit: 32.5
Avg TD/B melee hit: 36.5

Avg B/SK WS: 452
Avg TD/B WS: 495

Total damage at:

4.7: 20.37 * 32.5 + 452 = 1,114
5.0: 19.26 * 32.5 + 452 = 1,078

4.9: 19.50 * 36.5 + 495 = 1,207
5.3: 18.26 * 36.5 + 495 = 1,161

Giving ratios of 8.3% and 7.7%.

The maximum gain from SK's wind proc of 7 per hit yields a potential for up to 71 or 67 more damage per cycle on the B/SK side, reducing the gain to 1.9% and 1.4%, respectively. If the SK's wind damage does an average of 3.5 per proc (much more likely), then the overall gain is 4.5% - 5%.

From there you can then figure in the effect of the TD's accuracy. It looks like if you can use at least 4 of the 10 accuracy it puts the TD/B combo ahead.

~~ Just realized the original question included 2 TD option.. got distracted by the other discussion. Oh well.

So Fire Dakini/Blau will be superior to Blau/SK at capped acc then,

This QQ

Heh, so glad I went ahead and got both of these...although I had to sacrifice time for getting other important stuff done ; ;

Real question is...should I do another thunder dakini in anticipation of them expanding these. I'm speculating that the base dmg is gonna go up, delay down, and the stats will rise high enough that we may be saying goodbye to BD pretty soon...

no telling if that includes Blau, but i'm guessing BD will be dethroned in the near future.

So what the hell am I supposed to do with my Perseus's Harpe? ; ;

Lol hell might be too good for heart snatcher...

I'm gonna laugh my ass off if they put heart snatcher as a trial dagger in this update.