Parses after the patch

[LordMagnus]

Obviously you don't get it.. I went to GC camp with 0 pure Acc gear w/o a bard and I broke Acc cap (96.8%). You really think 15 more accuracy is going to matter? heck the DRG in that parse was getting a free useless madrigal too.

He never said he didn't have gear with accuracy+ on it, he said that he didn't have any gear with just accuracy+ on it (though that also isn't true if he was wearing the "accuracy ring" he mentioned ).

Also, correct me if I'm wrong, but I think the "0" next to "Polearm Merits" in the OP was supposed to mean "yes, he has them capped", not "no, he has 0 polearm merits".

[NynJa]

Also, correct me if I'm wrong, but I think the "0" next to "Polearm Merits" in the OP was supposed to mean "yes, he has them capped", not "no, he has 0 polearm merits".

merits: Axe/Sword 8/8, Berserk/Agg 5/5

shouldnt change numbering convention then :\

[LordMagnus]

Also, correct me if I'm wrong, but I think the "0" next to "Polearm Merits" in the OP was supposed to mean "yes, he has them capped", not "no, he has 0 polearm merits".

merits: Axe/Sword 8/8, Berserk/Agg 5/5

shouldnt change numbering convention then :\

word up

[Pix!]

Umm.. fixed.. they had no combat merits into their used weapons.
DRK had only 265 skill @ the start of the parse and got it capped during the PT

[Xajii]

Was I even fucking talking to you? I could have sworn I was responding to maxx about the ares body and ignoring whatever the fuck you're going on about.

This is why i love BG .. appeal to authority/ad hominem > all. All you need to do is agree with them and you're a f'ing genius... disagree and "wow.. are.. you.. really.. that.. dumb"

You're trying to take place in a numbers game type argument, and here's what you're saying:

+2.5% chance of Triple Attack outparses a raw 5% increase in attack speed?

Since your job doesn't have triple attack you have a 2.5% chance to hit three times instead of hitting 5% faster, adding much more than the rare triple attack does. Whether it's on a skill or not is such a fluke that it's almost negligible.

I would agree the BG veterans have an elitist attitude that makes them feel obligated to act like a douchebag to everyone instead of engaging in civil argument, but yeah, what you're trying to say is pretty dumb.

[fantasticdan]

I don't see the big difference between 2.5% triple attack and 5% haste on drg. Every 1000 attack rounds should net you 25 triple attacks, which is two extra attacks each, increasing your total swings to 1050, which I think is the same thing essentially as double attack + 5%? I always figured double attack 1% > haste 1% on slower weapons, especially in the case of drg where you have instant attacks where DA/TA can proc but obviously not affected by haste. Even if haste is slightly ahead when stacked due to increasing returns, it can't be by any discernible amount.

And whether or not your job has native triple attack has no bearing on the benefit of adding another fixed %, unless you are hitting some kind of cap. JA job with native 5% triple attack will triple 2.5% more with homam body than without, a job with 0% native triple attack with triple 2.5% more often with homam body than without.

[Kaylia]

And whether or not your job has native triple attack has no bearing on the benefit of adding another fixed %, unless you are hitting some kind of cap. JA job with native 5% triple attack will triple 2.5% more with homam body than without, a job with 0% native triple attack with triple 2.5% more often with homam body than without.

Wrong

5% base trait:
100 swings = 110attack average
+5attack from gears
=4.5% more hits (or increased dot)

No trait:
100swings = 100 attacks
+5attacks from gears
=5% more hits (or dot)

5% haste:
100 swings = 100 attacks
+5% haste
=105 swings for 105 attack in the same period

However, haste effect is cumulative, and most job have 10-20%, +brd song,+haste spell, so you would generally benefit more from haste than triple.

[Amele]

not picking on Xajii particularly but his post is a good jumping off point.

I would agree the BG veterans have an elitist attitude that makes them feel obligated to act like a douchebag to everyone instead of engaging in civil argument, but yeah, what you're trying to say is pretty dumb.

it's attitudes like this that lead people on this board to be douches to some others.


argue intelligently, with good evidence or at least clearly marking speculation, address the entire argument or indicate why you did not, avoid straw men, appeals to emotion, and especially ad hominem, be prepared to change your mind, and most importantly: you are not your argument.

do these things and you will have successful (if occasionally heated) discussions and debates and get the civil argument you are looking for, both here on the boards, and in life in general

[Amele]

I always figured double attack 1% > haste 1% on slower weapons, especially in the case of drg where you have instant attacks where DA/TA can proc but obviously not affected by haste.

I won't address potential added damage on dragoon from jumps,

double attack 1% with 0% DA: 101 attacks in 100 unit time period. (dot gain 1%)
haste 1% with 0% haste: 101 attacks in 100 unit time period. (dot gain 1%)

double attack 1% with 10% DA: 111 attacks in 100 unit time period. (net dot gain from 10%DA: 111/110 = 0.9%)
haste 1% with 10% haste: 112.35 attacks in 100 unit time period. (net dot gain from 10%haste: 112.35/111.11 = 1.1%

the numbers get more drastic the higher you go. at 24% haste (haste+march) adding 1% haste gives you: 1.3% at 35% haste (double march) adding 1% haste gives you: 1.5% and so on.

there's never a time when adding equivalent double attack is better dot gain than haste. when you have 0% of both only the first point is equal.

[fantasticdan]

That kind of math only works if you are meleeing the same mob for an extended period of time. What if you don't get in any extra swings due to haste on high delay weapon on a portion of the mobs you fight? You can say if you gave everyone 5% increase in speed mobs would die 5% faster, regardless of weapon delay, but when looking at it individually the haste may not do anything against a portion of the mobs. If haste didn't cause your attacks to cause the mob to die at an earlier time, or do you to get an extra swing, your haste did nothing on that mob, there are infinite scenario for that to happen. No doubt you will still see overall increase in DoT/kill speed no matter how much haste you add to any job, however the numbers might not match up to the formula.

Where as increased DA/TA should always benefit. I know haste has increasing returns, I said that in the post you are replying to, however I don't see how 2h can always stand to benefit as much as 1-handers, unless you have absurd amount of haste like relic drk. I think it's also worth noting when you gear yourself that jobs like drg and scythe drk will be last priority to receive haste whenever mp conservation is a concern. Not saying it isn't worth it to wear haste gear on these jobs, but I don't think the benefit over equal amount of DA/TA is really there, especially for drg with jumps.

[Amele]

in a good stage pulling merit party, you basically *are* meleeing the same mob for extended periods of time.

you're also thinking about haste backwards. it has the -same- effect on a weapon regardless of delay (note that I never needed to reference the exact delay in my proof)

time variation is a poor argument - if you hit 10% faster, and mob death varies *at least* the increment of your original delay (spoilers: it does) then you are 11% more likely to get that additional hit in, which is (surprise to no one) just as much as the predicted dot gain I gave.

this is because time variation can screw you out of that last hit even if you -don't- have any haste, so it's a non-issue.


although this isn't the reason that ares and askar bodies are both better than homam body for dragoon now.

[Siatdiat]

All other things being equal, I'd say everyone can agree that haste is best for tp, DA/TA more useful to macro in at WS (Jump, etc.)

[Quizno]

Honestly, there hasn't been enough research done to really give a clear picture of haste versus double/triple attack.

I was in the middle writing a longer email but my group is heading out to lunch so I have to bail on this for an hour.

[Amele]

Honestly, there hasn't been enough research done to really give a clear picture of haste versus double/triple attack.

I was in the middle writing a longer email but my group is heading out to lunch so I have to bail on this for an hour.

what do you mean not enough research? the effects of both are (almost) perfectly understood and once combined with our (almost) perfect understanding of melee swing delay calculations and a little bit of math you can get an exact answer of which is better and by how much.

even then, it's (mostly) academic because except for a limited number of slots on a limited number of jobs, you never have a choice between the two period

and to my knowledge there isn't even a meaningful choice for any of those slots either (I mean, there is never in any way under any circumstances a reason to use 1% double attack over 5% haste on warrior legs currently.)

siatdiat is right: haste does nothing during an actual weaponskill, so macro in as much other stuff as you can.

[Nameless]

I always figured double attack 1% > haste 1% on slower weapons, especially in the case of drg where you have instant attacks where DA/TA can proc but obviously not affected by haste.

I won't address potential added damage on dragoon from jumps,

double attack 1% with 0% DA: 101 attacks in 100 unit time period. (dot gain 1%)
haste 1% with 0% haste: 101 attacks in 100 unit time period. (dot gain 1%)

double attack 1% with 10% DA: 111 attacks in 100 unit time period. (net dot gain from 10%DA: 111/110 = 0.9%)
haste 1% with 10% haste: 112.35 attacks in 100 unit time period. (net dot gain from 10%haste: 112.35/111.11 = 1.1%

the numbers get more drastic the higher you go. at 24% haste (haste+march) adding 1% haste gives you: 1.3% at 35% haste (double march) adding 1% haste gives you: 1.5% and so on.

there's never a time when adding equivalent double attack is better dot gain than haste. when you have 0% of both only the first point is equal.

So should I be using blitz ring instead of the +4 attack +4 accuracy ring if I have 18% haste from my equips?

[Amele]

if your accuracy is already capped? probably, because -4 accuracy won't do jack, and -.011 PDIF ~4/340 is less than any gain in haste (.011 is like.. .55% lost dot at 2.0 pdif)

if not, 4 accuracy is 2% hit rate, which is (best case, 95->93 2.2% lost dot, and worst case: 7->5 28.6% lost dot.)

1% haste won't be worth more than 2.2% dot until somewhere around 55% haste.

[Quizno]

what do you mean not enough research?

What I mean is that everything that has been calculated has been done assuming optimal conditions. That is, they have been calculated out assuming constant attacks even though this only happens when you are smart enough to possition yourself for seemless movement between mobs after one has been defeated.

What a more exact explanation of the problem I have with this arguement is:

At what point does double/triple attack become better than any particular level of haste?

Is 2% double attack better than 1% haste? Obviously this relies on a great deal of data, and even when you *think* you have calculated out a point where you *think* they are equal, you are still probably forgetting some important variables.

2.5% triple attack is... 2.5% triple attack.

Assuming two body slots with 15 accuracy, but one has 2.5% triple attack, the one with triple attack should do 5% more dot, right?

But wait, this doesn't take into consideration that a mob could die during the first or second hit of a triple attack round.

In any case, based on probability you can calculate out the exact fraction which a particular percentage of double or triple attack will improve overall performance.

All this aside, in the case of the body slot the real question becomes:

which creates more DOT, attack+str+dex(critical hit rate) or a chance of triple attack?

Has work been done to determine how 15 attack will improve your DOT? or str? or DEX? I mean exactly, not just in general.

or is this (obviously) dependent on what you are fighting?

Atk/STR v Acc/DEX v Atk/STR + Acc/DEX v Haste v DA/TA simply isn't understood completely enough yet. And to be honest, it probably never will be because everyone would rather.



What happens if SE releases a R/X body that has 2% haste and 15 acc? Would that be better than E.Body? What if it was 1%? What if it was 5%?

Or is the decission going to be based off of what everyone decides is the cool thing to wear?

[Demitrius]

Unless I missed something massive in the whole haste calculation (which is possible), then haste stacks additively and your delay is multiplied by 1-(total haste). Then just look at it this way:

100% DA vs 100% haste.

One gives you a 100% increase in swings.
The other gives you an infinite increase in swings.

Each percent of DA is equal to the last. Each percent of haste ramps up until you have an infinite gain from 99% to 100%. So it would seem to me that any amount of Haste is better than an equal amount of DA (or, alternately, half as much TA, which is the same thing).

[Quizno]

Unless I missed something massive in the whole haste calculation...

I'm a bit past that... but in any case equipment haste caps at 25 while DA caps at... 100? guess there isn't enough data/available equipment to say this for sure yet, but in any case.

DA/TA v Haste are the easy ones to calculate out. The real problem comes when comparing them to accuracy and dot considering the changes.

Sure, people are going to keep coming here and explaining how easy it is, and yet the whole topic will continue to be filled with hearsay and e-peen.

[HadesBothari]

It's more along the lines of 1) post something new or interesting or make someone look stupid and you're a f'ing genius, or 2) disagree and either a) post no proof to support a baseless claim and be called an idiot, or b) post proof to support your now not-baseless claim, which in turn brings you back to 1).

Well K so.. I'm dumb so I might as well go ahead and do this:

@ 18% Haste, delay = 404 = 6.8s/swing
1000 swings in = 6800
@ 23% Haste, delay = 379 = 6.4s/swing
1062 swings in 6800s if you never lose the effect of haste from mob death.

6800s = 136 jumps
2.5% TA on 1136 hits = 28 TAs = 56 extra swings
1056 swings in 6800s

So, 2.5% TA results in about 6 swings less if haste 5% at 18% is given every advantage. Yes, I know it gets better with more haste, but 2.5% TA is still not insignificant.

And while I'm putting my foot in my mouth, I might as well compare Homam body to Askar Body (and see if TA is at all significant):

DRG/WAR with double min and berserk = ~770 ATK & ~85 STR
Avg Dmg on neutral mob = 185? with 25% dmg bonus = 231? (Close enough to what the parse showed)

DRG/WAR with double min and berserk + Askar = ~790 ATK & ~90 STR
Avg Dmg on neutral mob = 193? with 25% dmg bonus = 241?

Now, from that parse the max possible weaponskills were about 45 (assuming you went at exactly 100) out of which the DRG did 28 so, lets assume 17 feather tickles at 100 TP >.>. That's 17 times that the store TP on askar would have helped. i.e. got you to 100% TP 1 hit faster. Now say you always went at 100, then it would have netted you an extra 17 swings worth of tp or ~250 tp.

Now, Homam has 2.5TA. So about 15 swings (out of 318) were the result of TA = 210 TP and about (3450[230/hit] + 3000[WS]) = [6450] damage

Askar also has 2% double attack. So 6 extra hits (303 total hits w/o TA). That's 1446 Dmg and ~88 extra TP.

That is, 2940 [Extra melee dmg] + 4500 (WS) + 1446 (DA) = 8886 extra dmg

So, Giving Askar every advantage, it adds what 2436 out of 111000? I'm sorry I was too dumb to know it was better of the top of my head.