Forum Settings
       
Reply To Thread

(by request) X's/blau versus blau/sirocco comparisonFollow

#102 Feb 21 2009 at 12:44 AM Rating: Good
***
2,270 posts
Quote:
Fact: XB takes 10 successful rounds to get 102 tp. BS takes 11 successful founds to get 108 tp. XB uses tp more effiently. (this does not mean it gains tp faster or is the better combo, but it does mean the numbers are closer than anyone here (seems) to think).


The next time you get exactly 14 TP after every DE, hit twice a round only (no misses/doubles/triples) and land exactly on Xhits to 100 with any semblance of reliability please inform us. If you miss 1 hit in your scenario the table turns completely as it would take an additional entire round for X's to hit 100 and would waste MUCH more TP while sirocco would land right on loosing no time at all. I can counter your claim by pointing out that at the acc cap, you are VERY likely to miss 1 hit on the way up therefore making X's TERRIBLE. But you can counter that you will double or triple, or acc is farther under cap so you miss another one, while i say you can miss a hit or 2 on DE, and on and on.

Thf TP gain is FAR to unpredictable to depend on X Hits giving 1 build a lead over another when the difference is a single connected hit to 100. There are countless ways that Thf will get to 100 TP and with a single hit difference, it will favor one or the other changing constantly. If you triple Attack Blau/sirocco sits at exactly 100TP. If you miss your offhand hit on DE, the 102 vs 98 both drop under 100 meaning 1 more round. If you miss 1 hit on the way up. If you double a hit. If you multi hit that last round and "waste" TP. If you only land base hits +2TP from DE Etc.

Blau Sirocco can give you literally ANY TP number from WS from 1-16 in 1 tp increments. Using a stable base on a WS that can double/triple/miss in any combination of 1 or 4tp increments means you cant really build an Xhit build. Thfs high natural level of doubles/triples/misses during TP only exacerbates this issue. Increase this unpredictability exponentially because we swing in sets of 2 over 1 thanks to dual wield. With the variability of each build to 100TP, it may fall in Xs or siroccos camp.

Unless you want to do the rather extensive probability math to tell us what % of the time it favors X's vs sirocco, i have to call it a wash and will be forced to fall back on the avg TP gain of each set. Since Thf/nin terribly irregular TP gain, there will be both wasted TP and varying number of rounds to 100. Until someone can quantify the differences accurately, the ONLY measure we have to really compare dagger set TP gains is the speed they gain tp overall, not the exact Xhits to 100. We are not a Sam with 1hit WS and only a slight chance to double attack on the final hit as the only variables. Thf hits to 100 is notoriously difficult to accurately estimate since it has such a huge range of possible combinations starting with TP returns anywhere from 1-16. With so many variables, the avg will ultimately still show its face over time. In this case Blau/siroccos TP gain and ultimately number of WS will avg higher to a similar rate as the TP gain difference.

Edited, Feb 21st 2009 3:47am by Banalaty
Edit mah spelling sux

Edited, Feb 21st 2009 3:51am by Banalaty

Edited, Feb 21st 2009 3:53am by Banalaty
____________________________
Aanalaty, Mithra of San d'Oria
99Thf/99Drg/99Rng/95Pld/Secret Job 92
Bonecraft-100+3 7/29/08
Lu Shangs-5/28/08
Twashtar-90
Ryunohige-75: 1/7/2012
Mandau-75: 3/9/2012

Maat-1/1 Thf Meleed (pansies steal)
Some of mah gearz: http://www.ffxiah.com/player/Lakshmi/Aanalaty#item-sets
Join 'People Against The Heart Snatcher' and help put Thfs on the right P.A.T.H.S. today!
#103 Feb 21 2009 at 1:56 AM Rating: Decent
Thief's Knife
*****
15,049 posts
Quote:
All these numbers assume 1.3 pDIF mod for X's crits and 8/dmg average wind for sirocco. Mandau extra effects (triple dmg, poison, crit rates) are ignored.


Then it tells us nothing about Mandau/X's
____________________________
Final Fantasy XI 12-14-11 Update wrote:
Adjust the resolution of menus.
The main screen resolution for "FINAL FANTASY XI" is dependent on the "Overlay Graphics Resolution" setting.
If the Overlay Graphics Resolution is set higher than the Menu Resolution, menus will be automatically resized.


I thought of it first:

http://ffxi.allakhazam.com/forum.html?forum=10&mid=130073657654872218#20
#104 Feb 21 2009 at 6:40 AM Rating: Decent
***
1,274 posts
Banalaty: completely disagree.

You're right that it does vary for each set, but the amount of TP you gain for each successful hit does not change (aside from the 1's on extra ws connects).

If you're at 80% accuracy (and you should be) you should be reliably hitting at least 4 of the 5 hits. So if one dagger ended at exactly 100 tp (mandau/blau does) you may have a valid point, but since neither does, you don't (and as you mentioned its likely a random DA/TA may make up for this). So assuming 4 or 5 connects doesn't not affect this scenario. (assuming 0 extra hits is stupid though, and relying on a DA/TA process during the ws is equally dumb).

What we know: the amount of Tp gain for each hit, the amount of swings we'll make over time, our DA and TA attack rate and our accuracy.

What we can calculate: The average amount of time it takes for a hit to connect.

You simply take the amount of swings multiply in DA/TA process then cut it down by the amount of accuracy. This is the number of hits that connect over that period of time. Now take time, divide it by that number and you get average amount of time it takes for a hit to connect, now we can multiply back in the amount of tp each of those hits gives.

Now if I take Mel's setup and take out the subs and use sole sushi to get 95% acc. we get some interesting numbers:
X/B: real time/swing 3.45 seconds using the adjustment above 3.46 seconds
B/S: real time/swing 2.98 seconds using the adjustment above 2.95 seconds

Its interesting to me because at 95% acc. the B/S is fast enough to get enough DA/TA process to actually swing faster than normal delay through an average.

to reliably get a DoT calculation and tp gain I use the 2nd number (at 80% acc. both are longer) I fully recognize if you sat down with a stop watch, counted the rounds, and timed each one the amount of rounds would vary (the amount of hits would not - aside from perhaps the occasional really awful or good DE for connects) and the amount of time would probably never match, however if you did that 10 or 20 times the averages you come up with should match my numbers very nicely.

So I'm sure my numbers are off at a small sample size (say under 10 minutes), I mean I'm sure I can flip a coin and get 3 heads in a row, but if I do it 100 times I'm sure i'll get 50 heads and 50 tails. So we say its 50% chance. I only care about the averages. Arguing such and such might happen is dumb, because it won't happen with any consistency, and is pointless in an average over time comparison.

Fact of the matter is the tp formulas are:
Zx + y >= 100

Z=number of hits
x=tp/hit
y=extra tp from ws.

Z>=(100-y)/x

You're just comparing x, which would work if y and Z were the same, but they're not.



____________________________
FFIX Melee Damage Comparitor
Brimstone
#105 Feb 21 2009 at 6:43 AM Rating: Decent
***
1,274 posts
Lobivopis wrote:
Quote:
All these numbers assume 1.3 pDIF mod for X's crits and 8/dmg average wind for sirocco. Mandau extra effects (triple dmg, poison, crit rates) are ignored.


Then it tells us nothing about Mandau/X's


I would tend to agree, nothing about that calculation was optimized for it. But it took me 4 seconds to run it and post it.

I don't know what the crit rate mod is nor how long it lasts. If I knew either of those I could rerun the numbers, its probably also important to aim for high spike dmg. Which is clearly not the case in the numbers I ran.
____________________________
FFIX Melee Damage Comparitor
Brimstone
#106 Feb 21 2009 at 7:33 AM Rating: Good
***
3,775 posts
Quote:
What we know: the amount of Tp gain for each hit, the amount of swings we'll make over time, our DA and TA attack rate and our accuracy.

What we can calculate: The average amount of time it takes for a hit to connect.


Sorry brim, but I have to side with Banaltay on this one. The averages will add up to the tp expected but getting there means the random number generator, and I'm familiar with that enough to say you can't rely on that in the way you are trying to say. Banalaty explained the differences very well, so I'm not going to repeat it all, but I'll say this.

If you consider the evasion cap is 80% and I can easily hit this on most targets with my build, your argument of 80% accuracy can be tested with a mirror example. When I'm in my evasion build I can dodge pretty regularly for a while, but eventually the random number generator throws out **** at me where for a while I tend to go longer periods of time without getting hit, followed by a flurry of bad luck where I don't dodge for sh*t. Coomparing accuracy at 80% to evasion at 80% then means those misses would be hits, and hit's would be misses, but it still means that tp to 100% is unpredictable and accuracy will not change this.

I HAVE done this too. I have gone out to run errands for an hour and geared full evasion and let a level 1 bee swing at me while I was gone (not that I needed evasion here >.>, but I did anyway for fun). I came back and reviewed what I saw. I could group any number of consecutive swings generated at me and this could theoretically be a live scenario of tp to 100%. I had streaks where in 20 swing attempts at me I got hit 8 times (which would be only 12 successful hits in reverse) and I had streaks where I got hit once (which would be 19 hits landing on the way up). I have had streaks of 16 consecutive misses and streaks where I got hit 5 times in 8 swings. One hit..... that's all you need to sway the verdict one way or the other, and the sample size of each tp to 100% is only 10-11 rounds.

Statistically the random number generator will favor both dagger combos EQUALLY in the long run. There will be streaks where the blau/sirocco gets extra weaponskills and streaks where x's/blau gets extra skills, but they will average to 50% on each side, or no change at all to the ratio of number created. Your argument is moot because you invoked the random number generator, and you need an extremely LARGE sample size for them to average out, which will eventually be 50% both ways.

Edited, Feb 21st 2009 11:04am by Melphina
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#107 Feb 21 2009 at 8:00 AM Rating: Excellent
***
3,775 posts
EDIT: To clariify--- This post and the one above are addressing separate points. This post has nothing to do with the discussion of the RNG, It's here to clarify a math error I made for completeness.
------------------------------------------------------------------


For the record also

Quote:
-----------------------------------------
Calculating number of swings per round

Triple attack is merited 5 times, and I have brutal thus in 100 rounds

100 x 0.05 == 5 extra attacks from brutal
95 x 0.1 x 2 ==== 19 extra attacks from triple attacks
(100 x 2) + 5 + 19 == 224 attacks per 100 rounds
2.24 attacks/round


Since this was not in my OP I did not feel it necessary to change it, but I realized two days ago that this is wrong. The correct formula should have been

100 x 0.05 == 5 extra attacks from brutal
95 x 0.1 x 2 ==== 19 extra attacks from triple attacks
(100 + 5 + 19) x 2 == 248 attacks per 100 rounds
2.48 attacks/round

In my example I calculated the number of generated swings from 100 rounds, but when I went to combine them I effectively only gave ONE hand multi proc rounds. In truth, since 100 swings from each hand gets 24 extra attacks, and both hands comrpise a round, you get 1.24 swings per hand, and 2.48 swings per round in the long run. This doesn't change the verdict of anything and since it's not in my OP I'll just leave it noted here. It just means even more tp in the same timeframe, and we all know what that would mean for this discussion so I needn't add the details for you to know how I feel about it.

Edited, Feb 21st 2009 3:30pm by Melphina
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#108 Feb 21 2009 at 11:03 AM Rating: Good
***
1,274 posts
Quote:

Blau/sirocco gets 4.7 tp/hit and 9.4 tp/round while X's/blau gets 4.9 tp per hit and 9.8/round. Your accuracy is 84.5% thus

--------------
Blau/sirocco --- number of rounds x accuracy x tp/round

1877 x 0.845 x 9.4 == 14,909 tp/hr

------------------
X's/blau -- number of rounds x accuracy x tp/round

1621 x 0.845 x 9.8 == 13,423 tp/hr


a difference of 1486 tp/hr


If you can't create 2055 ACTUAL damage with over 12-13 weaponskills


Mel, I don't think you even understand my post. I'm throwing the random generator out by using totals. I'm just doing it in a way that accounts for all the subtleties of tp gain. You are not.

That may be the amount of tp gained, but its surely not 12-13 ws.

Lets assume you're using a single hit ws, and you do not get hit. This means the amount of tp you hold at any given time is a multiple of 4.7 or 4.9

so for B/S: 4.7 * 22 = 103.4 This is the minimum TP you can ws at, and depending on your rounds it may be 108.1

so for X/B: 4.9 * 21 = 102.9 possibly 107.8 (and maybe more likely if you have an even number of misses).

Obviously, this can change if TA goes off when you're between 95-100, but I agree that will even out either way. What I do not agree with is dividing by 100. What I've clearly noticed is it doesn't matter which job I'm on, with a given setup I will use the majority of my ws at the same tp. sometimes its 100 or 101 others its like 106-108.

In the 100-101 case I'm only wasting about 1% of my tp. In the other case I'm wasting about 7% of my tp. (and because of the setup, I don't really have a choice in the matter). This is fairly predictable and should be accounted for.

In my dumbed down example it actually doesn't matter, but the extra tp from a multihit ws does change it quite a bit (ie it makes up for the extra half of round on the X/B). But if you compare a 450 delay to a 463 delay its a big difference.

There is more than 1 way to get 'correct' numbers (or at least reasonably accurate). Totaling tp and dividing by 100 is not one of them, nor is taking a ratio of 2 numbers with an added constant in them (although this works better with a single hit ws, its will still be inaccurate if you try and compare a 450 to 463 delay for example).
____________________________
FFIX Melee Damage Comparitor
Brimstone
#109 Feb 21 2009 at 11:44 AM Rating: Good
***
3,775 posts
Quote:
Mel, I don't think you even understand my post.


I understand your point entirely, I separated those two posts because they address two entirely different aspects. The first post was to address your point, the second had nothing to do with the first. I discussed the concept of RNG separately from the math calculation, my second post was there just to clarify an earlier error. My apologies if you misunderstood, I edited the second one to point this out.

Quote:
I'm throwing the random generator out by using totals. I'm just doing it in a way that accounts for all the subtleties of tp gain. You are not.


But you can't do that. THIS is where your mistake is. The random number generator is an inescapable part of our game, and will nullify your argument. You can't say "the random number generator doesn't matter because I'm ignoring it in my examples", IT DOES!! Your throwing it out is wrong, YOU CAN'T. You are saying that x's/blau uses tp more efficiently, but I disagree with you and am in agreement with Banalaty. MY argument is that the random number generator will favor both combos equally, because each tp to 100% is too small a batch for statistical averages to be present EVERY time. You may tp in 11 rounds after weaponskill one time, but then miss a lot and need 14 rounds to tp again due to whiffed attacks. I'll give a scenario where you can see my point more clearly below

----------------Example:----------------------------

tp is 11 from previous ws

10 attack rounds generates 15 hits

blau/sirocco -- 4.7 x 15 == 70.5 + 11 == 81.5
X's/blau -- 4.9 x 15 == 73.5 + 11 == 84.5

one more round added of no misses

blau/sirocco -- 81.5 + 9.4 == 90.9
x's/blau 84.5 + 9.8 = 94.3


And the final round with no missed swings

blau/sirocco -- 90.9 + 9.4 == 100.3 -- In this instance blau/sirocco has a perfect efficiency 100% ws. Maximizing potential

X's/blau -- 94.3 + 9.8 == 94.3 + 9.8 = 104.1 tp -- In this instance x's not only gained tp slower it also wasted more

The difference in my example makes blau/sirocco be MORE efficient than x's/blau by 3.8 tp.
--------------------//End Example---------------------

You could have altered that to give the latter two attack rounds a triple with no missed swings and tp'd in one round instead of two, or a double with one missed swing, or three missed swings with a triple and a double and the result would have still been the same.

The ONLY difference in the blau/sirocco versus x's/blau getting tp more efficiently is then going to 100% with an odd or an even number of swings. Statistically you will tp to 100% in an odd number half the time, and an even number half the time. Because of this your argument is null and void. If tp was a static gain EVERY time to 100% THEN you could throw that card, but because of the RNG it has no standing grounds to fall upon since the only difference is tping in an even or an odd number of attacks, and that favors both combos equally.


Edited, Feb 21st 2009 3:36pm by Melphina
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#110 Feb 21 2009 at 1:11 PM Rating: Good
Sage
***
1,293 posts
There is no way to account for being at 95% tp and then having a round of double triples and then being at 124% tp. When I do my calcs, I always factor in about 10-15% wasted TP, no matter what, which is probably a little bit of overkill. Just because you do waste TP. But any set up wastes TP, even if you have made it so it theoretically doesn't.
#111 Feb 22 2009 at 12:50 PM Rating: Decent
***
1,274 posts
Meldi wrote:
There is no way to account for being at 95% tp and then having a round of double triples and then being at 124% tp. When I do my calcs, I always factor in about 10-15% wasted TP, no matter what, which is probably a little bit of overkill. Just because you do waste TP. But any set up wastes TP, even if you have made it so it theoretically doesn't.



1. I agree that does happen, and any instance it does will throw my current calculations off, additionally sometimes you look away and just wait an extra round or hold it because the mob is almost dead.
2. I disagree you can't predict it. Odds are it will happen at the same rate your DA and TA rates are (as in combined, so 5% DA and 5% TA will lead to a 10% chance you end up with too much tp. So you can factor the approximate amount out).
3. I think its relatively pointless to try and predict it anyways. If you're goal is to predict your damage, then it is. If your goal is to compare 2 pieces of gear in a theoretical environment so you know which to slap onto your character then all that matter is the performance difference between them, statistically they have to same odds (barring a DA or TA mod on the gear) to do this, ergo its a non-issue, the amount calculated will be overestimated, but overestimated equally for both.
4. In a THF/NIN discussion its wise to just throw TP gain from getting hit out the window, as its not difficult to go for very long periods of time from not getting hit, but in any discussion it can be a relative aspect of TP gain, its much harder to predict but still not impossible to develop a model. (you take into account amount gained per hit, compare evasion and enemy delay, decide if you have shadows, if so remove the delay spent casting from your normal swing delay, etc...) Its a level of **** retentiveness I don't care to get into myself, because you're essentially playing the game with equations at that point (and I'd much rather just play). But those weapons that like to stop at 99TP and force you to wait another round, getting hit once fixes that issue. I pretty much calculate everything like you're fighting a campaign wall and assume the player is paying attention, this gives you the relative information you should be looking for: which piece of gear (food, buffs, etc..) gives me the most potential to perform well over time. A large part of that is paying attention while playing, so why should any calculation assume someone is lazy? I've seen people with inferior gear out parse those with superior gear in an xp party simply because they consistently engage the mob soon, ride their JA timers and use TP more efficiently. Gearing yourself to the hilt and not doing those things merely wastes your potential, so why get all that gear?

When numbers are very close, it will come down to play style and making a personal guestimation of those fudge factors for yourself is perfectly reasonable. It drives me a little batty to see players deride others because in their mind gear X is theoretically 1% better than gear Y, and player is wearing gear Y. (not that that is necessarily the attitude in this thread).

Which leads me to my last point: When doing any sort of estimation it should be done using the best methods known which are reasonable to execute. (ie Newtons equations are not technically correct (Einstien's are), but they're still in wider use and valid in most context).
____________________________
FFIX Melee Damage Comparitor
Brimstone
#112 Feb 22 2009 at 1:16 PM Rating: Decent
***
1,274 posts
Mistress Melphina wrote:

because each tp to 100% is too small a batch for statistical averages to be present EVERY time. You may tp in 11 rounds after weaponskill one time, but then miss a lot and need 14 rounds to tp again due to whiffed attacks. I'll give a scenario where you can see my point more clearly below


And here's your mistake the sample isn't 1 trip to 100tp. The question is how often can I get to 100 tp over a larger period of time.

You're basically arguing you can't predict the probability of a coin on 1 flip. (duh!) we already know its probability, we know if you flip it enough times you will get 50% heads and 50% tails we also know that you will get 5 tails in a row as often as you get 5 heads in a row. The sample size is not going 0-100, going 0-100 is only 1 instance in the sample. (now you can replace heads or tails in this paragraph and the 50% and adjust it for any factor of this game).

To put it another way one combo will lose a round for some reason as often as the other combo gains around for some other reason, it comes out in the wash. (I don't think you disagree with this).

Lets try a more simple example:
450 delay weapon vs. 463 delay weapon, on a campaign wall, 100% accuracy (just for simplicity since its the same for both). no store TP, no double/triple attack, single hit WS, each person WS as soon as possible always.

Calculate for me, over a 1 hr. period:
1. the total TP gained
2. the total # ws performed
3. which gains tp faster? (compare any way you'd like).

Based on what I've seen in the thread so far, I don't think you'll come up with the correct numbers, but I believe I can demonstrate what the correct numbers are easily enough.
____________________________
FFIX Melee Damage Comparitor
Brimstone
#113 Feb 22 2009 at 1:54 PM Rating: Good
***
2,270 posts
Quote:
You're basically arguing you can't predict the probability of a coin on 1 flip. (duh!) we already know its probability, we know if you flip it enough times you will get 50% heads and 50% tails we also know that you will get 5 tails in a row as often as you get 5 heads in a row. The sample size is not going 0-100, going 0-100 is only 1 instance in the sample. (now you can replace heads or tails in this paragraph and the 50% and adjust it for any factor of this game).


The trouble is that your not fliping a either/or single variable. Its more akin to playing a game needing 21 points to win then rolling 4 10-sided dice at once with over 10k possible combinations. The higest you can get is 40 (4x10) lowest you can get is 4 (4x1). Sure the AVERAGE is 21, but the way you are using the variables is that if the avg is 21, you can then use that and say you will get 21 all the time without going under or over. You will NOT get 21 with any frequency because 1 or more variables (dice) wont land correctly. Just because the avg of all combinations is 21 doesnt mean that it is useful in predicting what individual rolls will be.

The tp return from DE is one variable. Acc is another variable. Double attack, Triple attack, etc. Every possible thing that gets you TP or limits TP is a variable.

If it takes exactly 19 hits to get 100 TP (example) and you get 5tp a hit you cant just say: "Well you miss 1 in 20 hits on avg, so you swing 20 times, miss once, and TP in 20 hits. 10 rounds exactly." *insert calcularion* All of the numbers we are using are averages to begin with. 10% triple, 5% double, XX% acc. Your aggregating averages to a single avg and applying them to smaller sample sizes. It doesnt work that way, its backwards.


Quote:
1. I agree that does happen, and any instance it does will throw my current calculations off, additionally sometimes you look away and just wait an extra round or hold it because the mob is almost dead.


Yes, but this is the primary point. Things do happen that will mess up the TP advantage in the method you have described it. The thing is that between odd DE return/double/triple/acc/wasted TP all affected further by dual wield since it brings ALL these variables together in sets of 2 that every round SOMETHING will happen. One of these "exceptions" that mess with your model WILL happen between 0 and 100 TP. It is the norm not the exception as you are more likely to have some combination of multi hit proc/miss/strange DE return than you are to have an uneventful trip that follows your model. If any of those happen, it can force tp to land favorable to either sirocco or to X's randomly on any occasion. Combined with wasted TP thanks to dual wield and multi hits, a fair amount of TP will be lost.

Simply because of the individual trips to 100 have such a HUGE variance (sometimes 15 sec, sometimes 45 etc) and can favor either build on an Xhit basis at any moment and completely randomly, the only avg we have left that is meaning full is the actual speed each set gains TP.

You are using the mean without accounting in any way for the HUGE variance within the data. However, over a LONG period of time (a party for example) the sample size grows and leans towards the TP gaining speed mean. Your measuring TP in 0-100 increments which has much to high a variance for mean averages to be very useful. Looking at the rate one set gains TP over another will be more useful for jobs like Thf with TONS of variables that have huge variation in smaller sample sizes.
____________________________
Aanalaty, Mithra of San d'Oria
99Thf/99Drg/99Rng/95Pld/Secret Job 92
Bonecraft-100+3 7/29/08
Lu Shangs-5/28/08
Twashtar-90
Ryunohige-75: 1/7/2012
Mandau-75: 3/9/2012

Maat-1/1 Thf Meleed (pansies steal)
Some of mah gearz: http://www.ffxiah.com/player/Lakshmi/Aanalaty#item-sets
Join 'People Against The Heart Snatcher' and help put Thfs on the right P.A.T.H.S. today!
#114 Mar 01 2009 at 12:27 AM Rating: Default
*
139 posts
BTW, I've done some very basic preliminary testing on comparing overall damage output between X's Knife vs. Sirocco Kukri myself now. (Eventually my goal is to single out a handful of the most common situations we THFs find ourselves in and determine which one actually gives an edge above the other.)

I parsed about a half-hour/45min worth of data in 2 series. It was me and another THF in my LS. Most of our gear was identical, with a few exceptions. The target mobs included the colibri, spiders, morbols and saplings near the Whitegate/Thickets tower, so they all con'd easy prey and decent challenge. (fairly low level mobs.) The first parse I went X/Blau and my friend went Blau/Sirocco. The second parse we both went Blau/Sirocco. Given that it was a quick job and we did have a small handful of armor differences between us, the data isn't perfect and was prone to show some variability that could have been nominalized had we gone for a longer period of time. I could therefore compare data between my own X vs. Sirocco output, my X vs. other Sirocco differences, and my Sirocco vs. other Sirocco differences. This short parses' result? All 4 data sets (2 from me and 2 from the other THF) were within less than 1.5% of each other at most, and that fluctuated either way depending on who WS'd last. At one point there was only a 0.07% difference between my X and his Sirocco output. Between my own X's output and my own Sirocco output.. there was almost no difference. There was one notable difference though: WS output vs. Melee output. X's increased WS damage output, and Sirocco outperformed X's in Melee damage output. Interestingly enough, they nearly perfectly offset each other and total WS+Melee damage combined came out to a virtual wash.

The damage output breakdown of my little "test" leads me to believe that an X's/Blau combo will perform better on very high level mobs, and Blau/Sirocco will perform better or roughly the same on other mobs. My next target are merit mobs for testing.

I should have saved that data. I'll do a better job of posting results and relavent data next time. That was just a quicky.

I liked all of the math work you guys have done on these comparisons. Very illuminating. Its a topic of interest to me, and I'm going to try to add some actual resulting comparison data on it as well.
____________________________
"An expert is one who knows more and more about less and less." -Japanese proverb

75's: THF BLM WHM SMN [Up and coming: DRK(66)]
Windy, Bastok, Sandy, Zilart, CoP, ToAU, WotG: Complete
Assault: Captain
Crafting/Hobbies:
Woodworking: 100+3+Signboard
Alchemy 100+3+Signboard
Bonecraft 100+3
Goldsmithing 78+2
Chocobo Digging: A10
(TARUTARU)
-----IFRIT SERVER-----
#115 Mar 03 2009 at 10:45 AM Rating: Default
**
452 posts
I have a question for Pudgy or anyone else with X's.

Does the knife affect Trick Attack before or after factoring the added AGI boost of Rogue Armlets +1? It only makes a diffence of the damage inflicted by a few AGI, but was just curious.
____________________________
F*ing polar bears.
www.myspace.com/thewintercomes


#116 Mar 07 2009 at 5:06 PM Rating: Good
*
139 posts
Well, not exactly sure Shamanfox. I still don't have the grasp on the damage calculations as some around here have; maybe one of them can use the data I'll post below and we'll see. Or maybe I can set up a different experiment to test that specificaly later. I just completed my first comparison test between X's Knife/Blau Dolch VS. Blau Dolch/Sirocco Kukri on a Greater Colibri Merit Party situation. Using K Parser, I collected the following data. Both data sets were taken in the same party, which I think represents a fairly standard merit party situation (on Greater Colibri), back to back.

X's Blau VS. Blau Sirocco Comparison, Colibri Merit Camp, Attempt 1:
(For every set of numbers displayed, the left figure is with X's/Blau, and the right figure is with Blau/Sirocco. Sorry, but the spacing I tried to put in to line everything up and make it all neat just didn't work.)

Main: X's Knife Blau Dolch
Sub: Blau Dolch Sirocco Kukri

Time Elapse of Parse: 29 min 41 sec 30 min 26 sec
Actual Time Spent Fighting: 25 min 52 sec 25 min 54 sec
# of Fights: 36 37
Total Damage delt by Party: 172685 182982
Total Damage delt by Thief: 60438 64219
% of total damage done in party: 35.00% 35.10%


MELEE:
Melee Damage: 35836 37886
% of Damage Melee: 59.29% 59.00%
Hit/Miss: 647/103 755/160
Melee Accuracy %: 86.27% 82.51%
Melee Low/High: 20/76 18/72
Melee Average: 47.02 42.14
# Critical Hits: 75 86
Critical Hit Low/High: 63/491 57/458
Critical Hit Average: 118.84 112.7
Critical Hite Rate: 11.59% 11.39%

# of Melee Attacks: 750 915
# of Melee Rounds: 307 350
# of Double Attacks: 79 117
Double Attack Rate: 13.21% 17.36%
# of Triple Attacks: 30 52
Triple Attack Rate: 5.02% 7.72%


WEAPONSKILLS:
Weaponskill Damage: 24602 29217
% of Weaponskill Damage: 40.71% 40.82%
WS Hit/Miss: 24/0 28/0
WS Accuracy Rate: 100.00% 100.00%
WS Low/High: 338/1444 233/1295
WS Average: 1025.08 936.32

Dancing Edge:
Dancing Edge Damage: 8466 14823
% of Dancing Edge Damage: 34.41% 56.54%
Hit/Miss: 8/0 15/0
Dancing Edge Accuracy: 100% 100%
Dancing Edge Low/High: 620/1340 513/1295
Dancing Edge Average: 1058.25 988.2

Shark Bite:
Shark Bite Damage: 16136 11394
% of Shark Bite Damage: 65.59% 43.46%
Hit/Miss: 16/0 13/0
Shark Bite Accuracy: 100% 100%
Shark Bite Low/High: 338/1444 233/1274
Shark Bite Average: 1008.5 876.46

Sneak Attack:
# of Sneak Attacks: 14 18
# of Solo Sneak Attacks: 7 10
Solo Sneak Attack Avg: 413.33 396.8
# of Sneak Attack stacked with WS: 7 8
Sneak Attack stacked with WS Avg: 1165.57 952.88

Trick Attack:
# of Trick Attacks: 13 16
# of Solo Trick Attacks: 4 5
Solo Trick Attack Avg: 395.5 382
# of Trick Attack stacked with WS: 9 11
Trick Attack stacked with WS Avg: 1038.44 956.82


CONDITIONS (Same for both):
Party Composition: WAR RNG THF SAM BRD WHM
WHM Influence: Haste
BRD Influence: 1 Accuracy Song + 1 Attack Song
WAR Influence: Warcy

Food Used: Meat Mithkabob
Thief Race: Tarutaru
Sub Job: Ninja
Relevant Merits: 8/8 Dagger
4/4 Critical Hit Rate
5/5 Triple Attack Rate
5/5 Feint
1/5 Assassin's Charge
1/5 Ambush

Gearsets used: If someone really wants me to post this, I will. Just getting a bit tired of typing. I'll summarize though: TP/Melee gear: Used haste set (3/5 Homam, Swift, Rap Harness.) WS gearsets: Used 4 different macros: Solo SA, SA+WS, Solo TA, TA+WS, with slight gear variations between them. Again, I'll post it if anyone needs/wants to know exactly what was used.

Conclusions:

Wow talk about a close call. In comparison to the rest of the party, the damage output between the two weapon sets only varied by 0.10%. While more actual damage was done with the Blau/Sirocco set, keep in mind that everyone in the party showed an increase in total damage output on that round,(1 extra mobs worth of HP?) and that the relative percentages scaled almost perfectly. Also, it might be worth noting that I did end the X's/Blau data set with 100% TP that went unused/unrecorded, and that was not the case with the Blau/Sirocco data set. So really, the X's Knife set got shafted 1 WS's worth of damage that it had accumulated the TP for. Seems to me that this adjustment would have made the comparison even closer.

I like this sort of testing because it takes into account player style and local inneficiencies. For example, in compiling the data to compare I noticed that my SA and TA usage counts were much lower than I thought they would have been. (14-18 SA's and 13-16 TA's in roughly 30 minutes.) And yet I thought that I had been using them as often as I could. Those few seconds delay waiting to hit 100 TP again or for whatever reason sure add up.

You can see some fairly predictable results: Blau/Sirocco showed a significantly higher number of attacks due to the lower delay of the weapons, and X's/Blau showed a higher damage per attack in almost every situation. (hi/low/avg, melee/ws, etc.) Its interesting that the two differences balanced out so closely.

Perhaps if SA and TA were spammed resulting in less wasted time when the JA's are ready, (and this would mean almost never stacking SA and TA with WSs- just using each as soon as the timer or TP is ready,) the X's/Blau set would pull ahead. (Just a theory, but I think I have an idea now for my next experiment.)

Its also noteworthy that no Wind Damage was factored in on the Sirocco set, since the Colibris took no wind damage. How much more damage would the Sirocco have done on a mob with comparable stats but minus the wind resistance, in the same time frame? I'd estimate an additional 6,040-7,550 damage, which would definitely give it the edge over the X's/Blau combo. (755 landed melee hits x 8-10 damage per wind proc.)

Hmmmmmm just a closing thought. I couldn't figure out how I did an extra 3,781 damage using the Sirocco set over X's when there was only 1 more mobs worth of HP to consider, and only a .1% overall damage increase of total party damage. The Wind Damage.. actually healed the Greater Colibri? (I had heard this before, but it doesn't show up in the chatlog, so I wasn't completely sure.) Thus giving the mobs fought more HP in that data set than in the X's/Blau set, and thus resulting in more damage done (by everyone.) This seems to be supported by:

182982 - 172685 total damage delt = 10297.
10297 - 1 Greater Colibri's worth of HP(~4000?) = ~6297
Estimated amount of HP healed by Wind Damage(same as what would have been dealt otherwise: 6040-7550.

The extra damage dealt overall, factoring out the 1 extra mob fought between the two data sets, falls within the Wind Damage/HP healed range that I estimated. This seems to fully account for the differences. More damage was dealt using the Sirocco, but the Sirocco caused the need for more damage to be dealt as well.




Bottom Line of Conclusions: In a standard Greater Colibri merit party situation with a BRD(acc/att songs) and a Haste from the mage, using a Haste/Meat build for TP'ing and decent WS builds, the difference in damage output between X's Knife and Sirocco Kukri is minimal at best. Play style matters, and depending on a persons play style, given the same equipment, one or the other could be more advantageous. Possible factors that could give one weapon set an advantage above the other:

1: If SA, TA, and WSs are taken advantage of more (as soon as TP or recast timers are ready), X's Knife may gain the advantage over Sirocco Kukri.

2: If the target merit mobs were different (anything but Colibri) then Sirocco Kukri will likely gain a significant advantage over X's Knife.
____________________________
"An expert is one who knows more and more about less and less." -Japanese proverb

75's: THF BLM WHM SMN [Up and coming: DRK(66)]
Windy, Bastok, Sandy, Zilart, CoP, ToAU, WotG: Complete
Assault: Captain
Crafting/Hobbies:
Woodworking: 100+3+Signboard
Alchemy 100+3+Signboard
Bonecraft 100+3
Goldsmithing 78+2
Chocobo Digging: A10
(TARUTARU)
-----IFRIT SERVER-----
#117 Mar 07 2009 at 5:13 PM Rating: Decent
******
22,693 posts
Smiley: yippee Thanks for taking the time to even perform something like this. Now if we can just got some 3 hour long ones for each sample size in the same exact py.. :Q___

Smiley: grin
____________________________
Dear people I don't like: 凸(●´―`●)凸
#118 Mar 12 2009 at 8:30 PM Rating: Good
*
141 posts
Guys, time to just toss in the towel and call it in favor of Heart Snatcher.
#119 Mar 12 2009 at 9:25 PM Rating: Excellent
***
3,775 posts
@ XxPudgyxX

That is a pretty impressive test, I definitely like that. It looks extremely accurate. I have to thank you for running that and posting it here. Looking at the data everything seems to correlate pretty well, and the conclusion is along the lines of my original assumption, albeit a smidge less difference.

I guess the best conclusion to draw from this is that unless you're in an end game shell where x's knife would be worth the amount of effort to get it (IE: Kings, Gods, HNM where you will be battling for prolonged periods of time, etc etc) the Blau/sirocco would be sufficient for all your daily needs. The X's knife really is that big of a pain in the rump to get with its once a week fight and drop rates so low, but in the aforementioned scenarios it would still be a good investment of your time, it all depends on what you do with your thief, or whether you will even take it to events such as these should you do them rather than another job.

The difference in performance is extremely close, but the wind damage would be a tipoff in sirocco's favor for TP burn, as I thought. For me that makes the decision easy, because I won't be seeing Cerb and Khim and Faf and Aspid etc etc again, and Odin I'll be coming to as Ranger. So I suppose I'll leave that as my closing thoughts. If you do fight these guys X's is a good time investment, otherwise you would probably be better off looking to Blau/sirocco and save yourself the effort of going through "Requieum of Sin". For most people sirocco is so much easier to get, and if it's going to perform equal to or slightly better than something you may not see in a years time the circumstances for each dagger are pretty clear in my mind.
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#120 Mar 13 2009 at 8:04 AM Rating: Good
Sage
***
1,293 posts
I have been browsing over to the KI forums occasionally, and interestingly enough, the topic of DE vs. Evis comes up quite regularly. Some interesting data has been showing up giving Evis more or less equal damage with DE in the long run (per testing on EP mobs by treelo, which favors DE imo). With X's, his eviscerations average about 50-100 more than his DEs over time. This may be something to consider down the road.

Treelo's opinion is that once you have X's Knife, you shouldn't even be touching DE.
#121 Mar 13 2009 at 8:27 AM Rating: Good
***
2,155 posts
Meldi wrote:
Some interesting data has been showing up giving Evis more or less equal damage with DE in the long run (per testing on EP mobs by treelo, which favors DE imo).


Just a quick input since I have to run to class soon, but wouldn't using EP mobs as a sample benefit Evisceration more than DE? I can't claim to know how significant the difference in AGI is between a VT mob and an EP mob, but I would think giving yourself a higher critical hit rate would benefit Evisceration a great deal. In fact, I don't see how using EP mobs would benefit DE at all (when compared to similar benefits that Evisceration will see. Ie: level difference, pDIF, Accuracy, etc).
____________________________
Chanchan - Mithra - Titan Phoenix
THF
#122 Mar 13 2009 at 8:42 AM Rating: Good
Sage
***
1,293 posts
Sorry, I meant that it was interesting data, but that it needed to be tested on more applicable mobs, IE merit mobs. I guess there is a trade off as far as the benefit, EP mobs are going to have a higher cRatio, which means the relative bonus of a critical is going to be lower, however, a crit is still a crit, and that is a significant boost.
#123 Mar 13 2009 at 10:24 AM Rating: Decent
******
22,693 posts
Hmm, that's an interesting notion actually.. have there been any tests that show crit rate increase on evisceration? That's the only unknown variable I can think of right now.
____________________________
Dear people I don't like: 凸(●´―`●)凸
#124 Mar 13 2009 at 10:56 AM Rating: Good
Sage
***
1,293 posts
The only info I have ever found on crit based WS is from the same source as the crit hit testing. Seems to indicate that 100 tp has a base +0% increase to crits on multi-hit crit based WS.

But this is the only info I have found: no testing to back it up.
Crit Info
#125 Mar 16 2009 at 7:30 PM Rating: Decent
**
951 posts
Quote:
Guys, time to just toss in the towel and call it in favor of Heart Snatcher.

lmao
____________________________
Khael of Lakshmi
Galka Beastmaster
=========================================================================
BST90/THF90/WAR90/BLM77/MNK66/DRK65/NIN62/PLD50/WHM50/RDM50/BLU50/DRG50/SAM50/DNC50/RNG43/SMN40/
ToAU 22/WoTG 4- Help Me Out!
Diabolos Fight- Help Me Out!
#126 Mar 17 2009 at 8:49 AM Rating: Decent
**
389 posts
Melphina, i hate you.

The only way i can find to disagree with you is "I dont believe you!" So much time wasted helping a friend getting his X for it to turn to naught... Its almost like having a Ridill on Warrior.


Going over these number for the Nth time already. Data parsing from the X friend who i enslaved also show pratically the same thing: X/Blau and Blau/Sirocco will favor a bit Blau/Sirocco when the mob dont take wind damage. With the wind procs its just that much more.

#127 Mar 17 2009 at 9:49 AM Rating: Excellent
***
3,775 posts
Quote:
Melphina, i hate you.

The only way i can find to disagree with you is "I dont believe you!" So much time wasted helping a friend getting his X for it to turn to naught... Its almost like having a Ridill on Warrior.


Going over these number for the Nth time already. Data parsing from the X friend who i enslaved also show pratically the same thing: X/Blau and Blau/Sirocco will favor a bit Blau/Sirocco when the mob dont take wind damage. With the wind procs its just that much more.


Smiley: lol I guess my assumption was right on then. But the point of this thread was NOT intended to trash the X's (it's still an awesome as **** dagger). The point was the show that blau/sirocco is an efficient alternative to x's/blau for those without the time to invest in "Requieum of Sin". I would still like an x's as a collectors item, and if someone was to offer to help me I wouldn't be above accepting that. I just don't want to form my own static. Blau/sirocco is an awesome combo, it's truly epic and a viable alternative. The two combos are neck and neck.

HOWEVER, there is a new development spawned from this thread that goes onto evisceration testing. It's still in its earliest stages, but I recommend you check it out. Banalaty's thread Evisceration Critical rate??.

The thread we just exhausted was focused on dancing edge, but if you have an x's knife you may want to consider Evisceration in its place, depending on the results of the new developments. I Still do NOT think x's/blau will be capable of outperforming blau/sirocco in a merit style fashion even with evisceration but the trend still shows situationality.

My final conclusion is that Blau/sirocco and X's/blau are both seriously awesome combos, and each rivals the other so close it's hard to come up with a definite winner when used to its strength (IE: blau/siroocco De and x's/blau evisc). My original intet was to break the old stigma "X's is greater than all...... period!!". Having done that I am willing to change my stance on Evisceration as well and take it from here with whatever future developments may arise from the new testing.
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#128 Mar 17 2009 at 1:32 PM Rating: Good
Scholar
**
357 posts
Great post, love all the math. I have one question:


Mistress Melphina wrote:

--------------------------------------
Part E: Weaponskill Frequency and TP gain

Ahhh finally, the crux of the matter at hand. The biggest factor in this comparison is weaponskill frequency. The blau/sirocco with store tp + 5 of rajas will be getting 4.7 tp per swing (and two swings per round of course). The X's/blau combination will be getting 4.9 tp per hit with the rajas store tp + 5.

Blau/sirocco with no haste

823 x 2 x 4.7 x 0.85 == 6575 tp/hr

Blau/sirocco with 25% haste

1097 x 2 x 4.7 x 0.85 = 8765 tp/hr

X'a/blau with no haste
712 x 2 x 4.9 x 0.85 = 5930 tp/hr

X's/blau with 25% haste

950 x 2 x 4.9 x 0.85 == 7913 tp/hr

[/sm][/i]



Isn't this a rather inaccurate way to to judge how many WS they will get off in an hour?

Since X's/blau takes 21 hits from 0 TP and 18 with 15 tp and Blau/sirc takes 22 hits from 0 and 19 with 15 tp shouldn't we go with:
(doing both examples with haste only)

Siroc
((1097 x .85) - 22) / 19 = 47.91

X's
((950 x .85) - 21) / 18 = 43.69


Thus giving us only 4 less WS rather than the 8-9 that you predicted the other way. I also understand that my way is also not 100% right and maybe a 13-14 TP return be be slightly more accurate but this seemed to close the gap a bit.

I have no idea if X's would once again take the lead or if my math is even right but I felt your way was an inaccurate way of getting to WS/hr. Also its worth noting that your way probably isn't as inaccurate as I first thought because of times you sit at over 100% tp, but maybe 5-6 less WS (in the haste example) would be most accurate, averaging the two methods.

edit: I see some of this argument already posted, I only read to the bottom of the first page and didn't see the extra pages lol.


Edited, Mar 17th 2009 5:36pm by smallgod
#129 Mar 17 2009 at 1:42 PM Rating: Decent
******
22,693 posts
Quote:
Isn't this a rather inaccurate way to to judge how many WS they will get off in an hour?


This, along with the fact that we currently are unable to pinpoint exactly how often, how successful, and etc Sa and Ta contribute to our DoT and weapon skills is why this theorycrafting weapons is not an end all thing. It only shows that Blau/sirocco could potentially be as good as X's/blau, and vice versa. It's near impossible to really know.
____________________________
Dear people I don't like: 凸(●´―`●)凸
#130treelo, Posted: Aug 21 2009 at 2:50 PM, Rating: Sub-Default, (Expand Post) Excuse the necrobump.
#131 Aug 22 2009 at 2:13 AM Rating: Default
Scholar
**
485 posts
For everybody who is thinking: "Who on earth is Treelo and where can I find him/her?" Treelo is a frequent poster on KI and is, how shall I put it...passionate about thf. Beware if you get into an argument with Treelo he he.

@ Treelo: Welcome to the topic. We should just continue the discussion here since all the wrongness is already posted here. Besides at the moment I do not know what posts and/or posters in this 3 page topic you are referring to.
#132 Aug 22 2009 at 2:20 AM Rating: Excellent
*
150 posts
Quote:
Excuse the necrobump.

You are wrong on so many levels it hurts my eyes. You know where to find me.


He necrobumped a year old thread and that's all he had to say. I've never even heard of him and whatever he meant it's so vague you can't tell who he's disagreeing with and what about. Nor does he post anything to back up whatever that would be. All I get from that post is that he's arrogant and trolling. Even more so because he's from KI.
#133 Aug 22 2009 at 3:55 AM Rating: Default
Scholar
**
485 posts
/em gets out the popcorn and waits for Treelo to start tearing **** up.
#134 Aug 22 2009 at 9:18 AM Rating: Good
Scholar
***
1,678 posts
Yeah, the 'you know where to find me' part is made of win. I just pictured some goth kid with mascara, lipstick, a cheap black trench coat and fake broken-in work boots trying to act all tough in front of the jocks. Is this the part where we all go down to the coffee shop, interrupt his spoken word poem, and call him out on his ****
____________________________
Dooom wrote:
BG elitists <3 haste.

Allatards/uber casuals hate elitists.

Allatards/uber casuals hate haste, as valuing it would acknowledge that elitists are right


kerberoz wrote:
People don't hate emo kids because they're "misunderstood." People hate emo kids because they're useless.


Realix wrote:
PUP is an average to above average DD... when not in a zerg situation... or on particularly hard targets... and when properly configured... on windsday... with a RDM...
#135 Aug 22 2009 at 11:29 AM Rating: Excellent
Quote:
For everybody who is thinking: "Who on earth is Treelo and where can I find him/her?" Treelo is a frequent poster on KI and is, how shall I put it...passionate about thf. Beware if you get into an argument with Treelo he he.


Quote:
/em gets out the popcorn and waits for Treelo to start tearing **** up.


Who the fuck are you, president of his fan club?
____________________________
In simplicity, utility. Through utility, simplicity.
#136 Aug 22 2009 at 12:32 PM Rating: Default
Scholar
**
485 posts
lol relax. It is a frikking forum.
#137 Aug 22 2009 at 2:07 PM Rating: Good
**
269 posts
If he has more to say then "no ur wrong," I hope he posts. "forum rivalry" aside, I'm more interested in the data.
____________________________
Wolfheart 80 THF, DRK, MNK but not bard. No, no I didn't level bard.
Phoenix server.
( ')> Shadow Duck approves of this message.
#138treelo, Posted: Aug 22 2009 at 2:32 PM, Rating: Sub-Default, (Expand Post) I'm the president of my own fan club. I have a lot of fans, but nobody loves me as much as I do. Any complaints may be directed to me.
#139 Aug 22 2009 at 2:46 PM Rating: Decent
Did you know that carrying Sweet William, the main ingredient in dexterity potions, has a hidden latent of +1 dex? It's hard to tell because it wont be listed in your base or +stats but after extensive discussion we're not positive that it indeed has a latent dex bonus of +1.

Disclaimer: I do not own Sweet William and have never attempted carrying it in my inventory.
____________________________
Lex Talionis :: Honoris Causa
The Law of Revenge :: For The Sake of Honor
#140 Aug 22 2009 at 3:06 PM Rating: Good
Scholar
***
1,678 posts
Is there a reason all your parses are with gimps?
____________________________
Dooom wrote:
BG elitists <3 haste.

Allatards/uber casuals hate elitists.

Allatards/uber casuals hate haste, as valuing it would acknowledge that elitists are right


kerberoz wrote:
People don't hate emo kids because they're "misunderstood." People hate emo kids because they're useless.


Realix wrote:
PUP is an average to above average DD... when not in a zerg situation... or on particularly hard targets... and when properly configured... on windsday... with a RDM...
#141 Aug 22 2009 at 3:59 PM Rating: Good
**
269 posts
Well, I was hoping for numbers, not "feelings," but whatever. Experience counts for a lot, but knowing the raw potential of different dagger combinations has its merits as well. I have/use an X's and I'll admit I can't really tell if the higher WS numbers beat out the TP advantage of blau/sirocco.

In any event:

Quote:
I'd like to point out the mindless idiocy of following the advice of someone with no prior experience in the topic at hand.


Melphina has an X's, and seems to have the same opinion that I do. Unless we're both doinitwrong, or X's leaps ahead by leagues once you equip N-legs (which mel and I both lack), the overall damage difference between X's/blau and Blau/sirocco is difficult to eyeball.
____________________________
Wolfheart 80 THF, DRK, MNK but not bard. No, no I didn't level bard.
Phoenix server.
( ')> Shadow Duck approves of this message.
#142 Aug 22 2009 at 4:31 PM Rating: Excellent
***
3,775 posts
I read your thread Treelo but before I comment on it I will first note that I will agree with you that math and a calculator are not always the best way to do things and that firsthand experience is a better way to see what works for oneself. It is however a good start to get an idea of what you can expect. I've also analyzed parses and seen the difference myself, and I still stand by my original statement on blau/sirocco being superior to x's/blau in a tp burn scenario. I've read both of your posts in their entirety, so while it's a tad long I'd request that you do so of this one also before commenting.

Now, Evisceration is a good weaponskill with x's, and with x's it makes sense to use it. When I made my OP I did not yet own an x's knife, but since then I acquired one of my own to play around with. That happened in the middle of April, and I've spent the past 4 months experimenting with both dagger combos. I took the data to heart about evisceration being superior when one has an x's knife and I've gone out and used the **** out of it. I've had time to merit on mamools and colibris both and my weaponskill set is about the same as yours, I only lack my hecatomb subligar +1 which is just a matter of Odin and time. I currently have 83 dex with merits and my weaponskill set adds 49 more so I'm working with a total of 132 dex. That's enough to hit the cap on the silly birds and then some. I've seen similar results to yours on damage output and whenever I wear x's my evisceration sessions average to similar results in damage increases over dancing edge.

Now before I go on any further I want to note that I don't use direct parse, I use Kparser. I know kinematics in game, in fact he's still on my friend list from the time I spent on Unicorn and I keep in touch. The thief tab on kparser was added by my own request and is an invaluable asset to seeing how soloing sneak and trick attack affects your damage. I've analyzed not only my evisceration and dancing edge performance with x's, I've also analyzed my solo critical damage as well as the other factors that come from using a parser such as average hit, accuracy % and the like. Kparser gives much more data than directparse and if you analyze it all you see a bigger picture. Now, with that said I'm going to respectfully disagree with you that IN A TP BURN scenario the x's/blau using evisceration is superior to blau/sirocco using dancing edge. The reason is the same one that has been rehashed in this thread which you have failed to acknowledge, the speed of the blau/kukri and increased weaponskill frequency.

When I merit with my x's/blau combo the difference in speed from the blau/kukri combo is easy to see, and it's even more prominent when I add bard haste alongside magic. If I use x's/blau I'm going to use evisceration as my weaponskill, but when I use blau/sirocco I'm going to be using dancing edge. The number of attack rounds with the kukri is just THAT much higher, and my weaponskill count with blau/sirocco is approximately 15-20% higher than that of my x's/blau setup (the higher end with double march of course). While evisceration averages to a bit more in weaponskill damage, with x's I get a LOT fewer of them. My blau/kukri dancing edge set is strong enough to rival that of my x's/blau evisceration set that the frequency of dancing edges overtakes the average damage of evisceration. I don't get those OMGWTF e-peen shots to go as high, but I don't care about them anyway (and let's be honest, on merit mobs not many do anymore. That trend ended when most people finished meriting and got fed up with it a year ago or so). When I use kparser I can see the percentages of my damage that come from weaponskill solos, solo criticals (as I dub them my "Spike DoT") and my stacked weaponskill damage. I can also see what the wind damage of sirocco kukri does when I leave the birds behind and it starts to proc.

Now guess what? This may be a shock to you, but when I use blau/sirocco and analyze the data alongside similar x's/blau parses the blau/sirocco puts out equivalent and slightly superior numbers over the course of the parse. It merely alters WHERE the damage comes from. While with x's/blau I have a higher damage ratio in weaponskills and criticals and solo crits (surprise), when I use blau/sirocco my DoT goes up by an amount strong enough to more than make up for it and overtake the difference. We are talking TP BURN here, and that means haste, and lots of it!! In the case of a strong dancing edge thief with blau/kukri it's enough to close the gap and overtake the evisceration x's/blau.
-----------------------------------------------------------------

:Summarising that:

Treelo you are obviously a competent player who knows what he's doing. I can tell you love the job as much as I do and that you've poured your heart and soul into the job, and for that I respect you. I will honor anyone who puts a strong effort into the job I know and love. You are every way equal in dedication to bettering your thief as I am to mine, and I /salute a fellow thief for that. However, you have failed to acknowledge the difference between the two dagger setup as I have done in MY op. In your post you have explained that when you use x's knife evisceration makes you strong. DUH! That's a no brainer. What you have NOT done is post any data on the reverse side of the equation of your experiences using the blau/kukri combo and dancing edge. Also, though it's not a HUGE change while evisceration has a modifier to criticals dancing edge has a bonus to accuracy, which even if it's a small amount like 5 accuracy it will benefit the combo.

Also, while I DO respect you for your dedication to the thief job, I DON'T respect you for your attitude. What your x's knife thread comes off to me as is nothing more than a glorification of the e-peen that you so very much love. Quoting your own post

Quote:
Admittedly a lot less reliable than DE as you can see from the lowest WS, a result of missed hits and not many crits. The overall average is higher though, which is essentially what you are aiming for, then add the much vaunted e-peen shots you'll be looking for. An Evisceration that triple attacks and crits with every hit is a beautiful sight to behold.


You play for e-peen and you're full of yourself. You're so arrogant that you've become blinded to the fact that in a TP BURN scenario that x's knife of yours is SLOW. I'm sorry but it just is. The amount of haste available to us in a DoT oriented TP burn setup combined with magic haste and then march is an equally beautiful thing in my eyes as the e-peen is to yours. And I value the wind damage of the sirocco, it really adds up over time and tips the scales. If you went out and looked at an UNBIASED comparison between an evisceration x's/blau setup versus a dancing edge blau/kukri setup and compared the two I think you would see the difference is not quite the way you think. Also this comment holds merit

Quote:
Is there a reason all your parses are with gimps?


As you yourself said

Quote:
So since my arrival on the shores of Asura I've been able to get some serious meriting done; mostly just to show off, partly because my other jobs need the merits. Where are you guys hiding all the e.body/ridll wars eh?


Amidst all that bragging of yours you parsed against a war/nin with double march AND madrigal + minuet (dual wielding) and outdid him by 19% after the piercing bonus is ignored, and the sam/rng didn't do anything. Then you parsed against a great axe war/NIN and another sam/rng. Again, your parses are just throwing out e-peen and glorifying yourself. Treelo, you've fallen victim to "Big Stupid Number syndrome" and that doesn't bode well on your argument one bit. I understand you're passionate about the thief job but so am I. And if you took a looksie at all of the names on pages 1-3, I think you'd find that a majority of them are equally passionate about the job as well.

Edited, Aug 22nd 2009 9:04pm by Melphina
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#143 Aug 22 2009 at 4:48 PM Rating: Decent
Scholar
***
1,678 posts
Quote:
Also this comment holds merit


Hush you; you stole my bait.
____________________________
Dooom wrote:
BG elitists <3 haste.

Allatards/uber casuals hate elitists.

Allatards/uber casuals hate haste, as valuing it would acknowledge that elitists are right


kerberoz wrote:
People don't hate emo kids because they're "misunderstood." People hate emo kids because they're useless.


Realix wrote:
PUP is an average to above average DD... when not in a zerg situation... or on particularly hard targets... and when properly configured... on windsday... with a RDM...
#144treelo, Posted: Aug 22 2009 at 5:15 PM, Rating: Sub-Default, (Expand Post) You can't troll a troll.
#145 Aug 22 2009 at 6:15 PM Rating: Decent
Scholar
***
1,678 posts
Quote:
Evisceration certainly seems to favour an incredibly DEX heavy build.


Again, you're only factoring in one part of a THF's DoT.

Quote:
Outside of HQ Heca, I have virtually maxed out DEX for my WS build, with 5/5 DEX merits to back it up. Essentially, the more you have, the more useful X's Knife becomes.


The difference between a virtually maxed out Evisceration build and a virtually maxed out DE build are going to be maybe one piece different? Either way, both are going to be piling on as much DEX as possible. The difference is that the added damage to critical hits from X's, combined with the nature of critical hits, will make Evisceration perform better from an average damage perspective than Blau/SK using Dancing Edge. The trade-off for higher WS average is lower 'normal' DoT, which is where Blau/SK wins.

Quote:
the sheer weight of information in this thread alone would lead many people to believe it is the best combination possible outside of relic territory.


Actually, Melphina said that the two combos are pretty much identical, leaving it up to the reader to decide if the time spent on acquiring an X's Knife was worth it for them or not.

You should also take into account that evidence shows critical hits on WSs tend to follow the same guidelines as normal critical hits (this is assuming that Evisceration behaves the same way as RR and DB, which, by the description, it should). So for the 'average THF' with 'average gear', even being able to get 40 or more DEX than any worthwhile mob's AGI in their WS builds is fairly uncommon. In other words, yes, Blau/SK using DE is going to produce better numbers for most players, but that isn't due to lack of X's Knife so much as lack of other gear.

Using Melphina's WS gear as an example, with 132 DEX, she is going to get about an extra ~10% chance at a critical hit proccing on any hit of Evisceration if the mob's AGI is 92, and an extra 20% chance if the mob's AGI is 82 or lower. Anything with an AGI higher than 92 is going to cause a sharp decline in the average damage of Evisceration, X's Knife or no. It's going to give the same percentage boost in damage regardless, it just won't be as effective against mobs with higher than normal AGI (I'd like to see numbers against Lurkers, but such a test would take forever to get anything approaching a reasonable sample size).
____________________________
Dooom wrote:
BG elitists <3 haste.

Allatards/uber casuals hate elitists.

Allatards/uber casuals hate haste, as valuing it would acknowledge that elitists are right


kerberoz wrote:
People don't hate emo kids because they're "misunderstood." People hate emo kids because they're useless.


Realix wrote:
PUP is an average to above average DD... when not in a zerg situation... or on particularly hard targets... and when properly configured... on windsday... with a RDM...
#146 Aug 22 2009 at 6:50 PM Rating: Excellent
***
3,775 posts
Quote:
Actually, Melphina said that the two combos are pretty much identical, leaving it up to the reader to decide if the time spent on acquiring an X's Knife was worth it for them or not.


This is exactly what I'm saying. There is no noticeable difference between x's/blau and blau/sirocco in a tp burn scenario that would clearly put either combo ahead. X's knife is an AMAZING dagger for gods, in fact barring relic it's THE dagger for gods. But in a tp burn where you're subbing /nin the combos are next to identical (/war would be blau dolch without question, the 16 attack alone will offset the difference, plus 5 accuracy and its monster DoT). I never said X's was BAD, I said that in a non God setup the blau/sirocco will perform equallly. I also considered that most thieves will NOT have this kind of gear, and considering this comment

Quote:

Using Melphina's WS gear as an example, with 132 DEX, she is going to get about an extra ~10% chance at a critical hit proccing on any hit of Evisceration if the mob's AGI is 92, and an extra 20% chance if the mob's AGI is 82 or lower. Anything with an AGI higher than 92 is going to cause a sharp decline in the average damage of Evisceration, X's Knife or no. It's going to give the same percentage boost in damage regardless, it just won't be as effective against mobs with higher than normal AGI


For the majority of thieves without the best stuff dancing edge will probably be performing similarly or even better because they lack the dexterity necessary to really push evisceration into the range it needs, which as you yourself said "needs a LOT of dex, increasing exponentially.

Now, you are a mithra with 5 dex merits, as am I a mithra with 3 dex. But what about elvaans with 5 strength merits and a base dex of 70, they have a nasty handicap right there of 15 dex toward you with even the BEST of gear (but they have 8 more racial strength and 5 merits, meaning that fSTR +3 is going to do a WORLD of good on their melee DoT with blau/sirocco). What about tarutarus who merited int for a mage, or people who don't have weaponskill gear with + 50 or more dex (and I don't care WHAT race you are, a mithra with a ws set of + 30 dex is limiting their performance in evisceration). And not everyone will have skadi's cuirie or etc, the hecatomnb harness is still a seriously GOOD weaponskill body for thieves who can't cap str (and even for those who can with it but need the accuracy) but it adds no dex. I used my heca harness +1 for ages, I only swapped when I got skadi's. I'd have used the nq if I had only that equally long. There are many many other examples of what would weaken a weaponskill set in terms of dex to add necessary balance elsewhere.

When you consider that even with the BEST of weaponskill stuff x's/blau with evisc and the blau/sirocco with the same ws set using dancing edge will be so close to identical it's hard to see a real difference, then add that that's elite thieves and a majority of thieves aren't elite I still recommend the blau/sirocco for most thieves if all they need is an everyday /nin combo.

By the way, Wolfhart isn't the only one who agrees with me. I know Lordtrey has an X's also and feels similarly as do others. We can't see a noticeable change one way or the other, it's just THAT close. I believe deadgye feels similarly, but he'll use x's just because of its rarity. Heck that's fine, I use x's a lot too!! It's a rare item and I own it so I'm going to use it (barring a really great augment on fane/blau, then I'd use THAT instead. I just need some strong luck... and maybe a good 3-5 years ; ;) I use x's/blau and blau/kukri interchangeably depending on my mood (barring puks of course, but that's another duh moment).

For most thieves the blau/kukri is going to serve them equally or better to the x's. And getting kreutzeuit to drop his dagger versus the mithran sisters ordeal will be no contest for 95% of people in terms of difficulty. Again, there will be exceptions, but most fall into this category. If you're one of the exceptions then by all means, go for x's!! Even if you aren't in the lucky boat X's is still GREAT if you own it when you fight gods and a strong dagger to strive for. But in /nin it's comparatively identical to blau/kukri. The math I posted isn't all that shows this, so do many parses I and others have seen as well.

Edited, Aug 22nd 2009 11:07pm by Melphina
____________________________
[ffxisig]56619[/ffxisig]

There are 10 kinds of people in the world. Those who understand binary, and those who don't.
#147 Aug 23 2009 at 10:21 PM Rating: Decent
Scholar
25 posts
Oops wrong thread

Edited, Aug 24th 2009 6:15am by MalicFayt
____________________________
Malicfayt
Mithra
75Thief<Main
75Drk/Brd/Smn/Dnc/Nin
#148 Aug 24 2009 at 12:10 AM Rating: Good
***
2,620 posts
I thought this Treelo guy was going to really be something from his first post in this thread, and the responce's in the couple of posts after, but all I can think to myself now is 'Is this guy for real'?

It has nothing to do with which combo is better, its how hes trying to prove X/B is better. He has severe BNS, parses against gimps, no mathematical knowledge, and an arrogant nooby attitude to accompany it.


I personally like to challenge myself when i parse, I merit with equally geared players (also to keep the balance of the meripo) and do my best to top the parse. Whether I do come first (or 3rd ><) theres never more than 4-5% between 1st and 3rd.

Mel did a great job at showing what kind of attitude this Treelo has in her post.

Treetop wrote:
Most people turn to KI for entertainment and don't tolerate the depressingly drab reams of figures found here (not to mention the god awful way this thing handles postings, I dunno how you put up with it.) I entertain while I attempt to educate, people tend to listen better.


Sounds like Killing Ifrit is full of ffxi wannabe's. Seems like we get the complete package here at allakazham.

____________________________
Taking a break.
#149 Aug 24 2009 at 8:36 AM Rating: Excellent
Scholar
***
1,678 posts
FYI, I went to check out their THF forums. Turns out treelo is just asking people like Shamaya how to get better damage out of Evisceration, anyways. So much for tough talk. And really, apparently we're all doing it wrong for not having a SCH to give us Hailstorm and a COR giving all your DDs Rogue's Roll (instead of Corsair's), just for our benefit. In case you wanted the real answer to why he merits with **** players... he is one.
____________________________
Dooom wrote:
BG elitists <3 haste.

Allatards/uber casuals hate elitists.

Allatards/uber casuals hate haste, as valuing it would acknowledge that elitists are right


kerberoz wrote:
People don't hate emo kids because they're "misunderstood." People hate emo kids because they're useless.


Realix wrote:
PUP is an average to above average DD... when not in a zerg situation... or on particularly hard targets... and when properly configured... on windsday... with a RDM...
#150treelo, Posted: Aug 24 2009 at 9:00 AM, Rating: Sub-Default, (Expand Post) Ohohohohohoho. How did I know that anything less than a vast line of numbers rivalling the Great Wall of China would be "less than satisfactory" to the math mad masses of Alla? As for parsing against "gimps", I'm new to the server, outside of my LS I know very few people and despise parties with random people. Unlike many of these people, I've spent years camping HNMs, and have had plenty of time to pimp myself out as a result, they have not. Some were good, some were awful. I pointed out in my post why the good ones lost, rather than plsying ghetto tank for DDs who don't sub NIN, I was able to play THF how it was meant to be played, and my performance increased drastically as a result. I wouldn't go so far as to say they were gimp (though some were), it merely demonstrates the difference between good and very good, not to mention the additional bonus of using piercing weapons.
#151treelo, Posted: Aug 24 2009 at 9:18 AM, Rating: Sub-Default, (Expand Post) Seems this popped up while I was writing my wall-o-text. I discuss quite a few things with Shamaya from time to time, if getting clarification on how some mechanics work and throwing ideas at one another is wrong, you're all as guilty as I am.
Reply To Thread

Colors Smileys Quote OriginalQuote Checked Help

 

Recent Visitors: 22 All times are in CDT
Anonymous Guests (22)