101 posts

Trying to find some data on Epona's vs other ring options for WSs. Interested in any of the tops WSs on any of Epona's jobs, but any info on any WSs would be appreciated. Thanks!

Sage

265 posts

On top of my head I'd think it wouldn't do any good on Asuran Fists because of the maximum round hits. I'm not too good with math stuff but I use it full time other than that.

Monk. Need I say more?

[ffxivsig]1529088[/ffxivsig]

Guru

2,236 posts

Dustin, your numbers don't really make much sense.

Accuracy has no relevance on the value of DA/TA. They are independant variables.

Not sure how you're getting 35% DA. As mnk (given this forum), you're only going to hit 15% if /war, 20% if you using VV atma (generally a weaker atma choice for us). Maybe you're considering dnc's Saber Dance.

If you're including other jobs, then on the TA side you have to consider that thf can hit over 30% (trait + merits + atma + Raider+2 head + Triplus Dagger, maybe an extra 1% from Raider Earring).

I have absolutely no idea where you're getting the 'extra hits' numbers you listed (especially if you factored acc into them, which is incorrect methodology).

However, a few quick checks indicate that Epona's is worth about:

Mnk: 2.7% (Smite, /nin) to 2.5% (Smite, /war)

Pup: 2.1% (Pummel, /nin) to 1.9% (Pummel, /war)

Thf: 1.6% (Evis, /nin) to 1.5% (Evis, /war)

And in general, values under 2% can probably be matched by other rings (Strigoi, etc). Weaponskills of 5 hits or less are worth using Epona's for, and 6-hits can be a wash depending on job. 8-hit weaponskills (ie: Asuran) will never benefit from DA/TA in any way whatsoever.

Accuracy has no relevance on the value of DA/TA. They are independant variables.

Not sure how you're getting 35% DA. As mnk (given this forum), you're only going to hit 15% if /war, 20% if you using VV atma (generally a weaker atma choice for us). Maybe you're considering dnc's Saber Dance.

If you're including other jobs, then on the TA side you have to consider that thf can hit over 30% (trait + merits + atma + Raider+2 head + Triplus Dagger, maybe an extra 1% from Raider Earring).

I have absolutely no idea where you're getting the 'extra hits' numbers you listed (especially if you factored acc into them, which is incorrect methodology).

However, a few quick checks indicate that Epona's is worth about:

Mnk: 2.7% (Smite, /nin) to 2.5% (Smite, /war)

Pup: 2.1% (Pummel, /nin) to 1.9% (Pummel, /war)

Thf: 1.6% (Evis, /nin) to 1.5% (Evis, /war)

And in general, values under 2% can probably be matched by other rings (Strigoi, etc). Weaponskills of 5 hits or less are worth using Epona's for, and 6-hits can be a wash depending on job. 8-hit weaponskills (ie: Asuran) will never benefit from DA/TA in any way whatsoever.

Scholar

1,270 posts

I am not sure how you are only get 15%/20% for mnk, i guess you have BB and dont use atheling?

10 from /war

5 from vv( I thought it was 10, so part of my error))

then 2% from belt, 3% from mantle, 5% from brutal, 3% from epona

Which is 28%, (33% with my bad on the VV)

I wasn't aware that TA/DA could only fire 2x per ws max, I was calculating the odds of it firing on any hit, then multipling by the acc to get adjusted avg values. So yeah my numbers are gonna be way off lol. I just assumed that it could fire on any of the 4 hits, not just the first 2, and I guess it make sense that if TA fires, DA cant.

I am not sure why you wouldnt count acc, do DA/TA have 100% acc or something?

although our conclusions are identical ...

10 from /war

5 from vv( I thought it was 10, so part of my error))

then 2% from belt, 3% from mantle, 5% from brutal, 3% from epona

Which is 28%, (33% with my bad on the VV)

I wasn't aware that TA/DA could only fire 2x per ws max, I was calculating the odds of it firing on any hit, then multipling by the acc to get adjusted avg values. So yeah my numbers are gonna be way off lol. I just assumed that it could fire on any of the 4 hits, not just the first 2, and I guess it make sense that if TA fires, DA cant.

I am not sure why you wouldnt count acc, do DA/TA have 100% acc or something?

Quote:

Weaponskills of 5 hits or less are worth using Epona's for, and 6-hits can be a wash depending on job. 8-hit weaponskills (ie: Asuran) will never benefit from DA/TA in any way whatsoever.

Quote:

you only benefit on 4 (Victory smite) and 5 hit ws from equiping epona.

although our conclusions are identical ...

Scholar

1,423 posts

Why would you use twilight for ANYTHING on mnk? It doesn't win anywhere.

Because acc has 0 to do with DA/TA.

Let's say I hit for 1 dmg/hit and swing 100 times.

100 dmg.

Now let's say I hit for 1 dmg/hit and swing 100 times w/ 10% DA. 110 hits total

110 dmg.

110/100 = 10% increase.

Now let's say I only have 50% acc in both scenarios. I hit half as much, do half the amount of damage.

50 dmg vs 55 dmg.

55/50 = 10% increase.

No matter what your hitrate is, you get the same relative increases to your damage, which is why looking at acc is 100% irrelevant when looking at increases from da/ta

*Edited, Apr 5th 2011 6:58pm by Veggeto *

Monk Quote:

I am not sure why you wouldnt count acc, do DA/TA have 100% acc or something?

Because acc has 0 to do with DA/TA.

Let's say I hit for 1 dmg/hit and swing 100 times.

100 dmg.

Now let's say I hit for 1 dmg/hit and swing 100 times w/ 10% DA. 110 hits total

110 dmg.

110/100 = 10% increase.

Now let's say I only have 50% acc in both scenarios. I hit half as much, do half the amount of damage.

50 dmg vs 55 dmg.

55/50 = 10% increase.

No matter what your hitrate is, you get the same relative increases to your damage, which is why looking at acc is 100% irrelevant when looking at increases from da/ta

#7**dustinfoley**,
**Posted:** Apr 05 2011 at 5:23 PM, **Rating:** Sub-Default, (**Expand Post**) Besides being a **** you missed the point veg.

Guru

2,236 posts

My mistake on Aethling; forgot about that (though it was included in my comparison percentages). Didn't consider Twilight for mnk. Those two together could put you at 20% in gear, 25% with VV atma (still 5% shy of your 35%, even with the incorrect VV amount). I specifically excluded Epona's from the base amount because Epona's is the gear piece in question, so trying to express its value in TA/DA when you've already added that amount to your total misrepresents what you're trying to compare.

You're conflating the terms in your argument, including your explanation of how you're using them.

If you swing 100 times and add 10% DA, you swing 110 times.

If you hit 100 times and add 10% DA, you hit 110 times.

To hit 100 times with 95% acc means you swung 105.263 times; add 10% DA and you swing 115.789 times; re-apply accuracy and you hit 110 times.

In your quote above, however, you say you 'hit' 100 times, then add 10% DA and gain an extra 9.5 hits because of accuracy. This is wrong. A swing and a hit are indeed different, but you don't seem to actually understand that difference.

Despite the offense you took at it, Veggeto's explanation of the matter is exactly correct. If you add 10% DA that means you swing 10% more often, and by swinging 10% more often you hit 10% more often no matter what your accuracy is. 95%, 88%, 79%, 50%, doesn't matter; it's all the same. Therefore accuracy is irrelevant in a comparison of the value of DA/TA, except insofar as your accuracy itself may change due to the different gear choices (eg: Sniper Ring vs Epona's when below the acc cap).

Also, our conclusions are not quite identical. Aside from thf, there is no particular benefit in using most other rings in place of Epona even at the 6-hit level. Therefore it's worthwhile (ie: you won't do worse to any notable degree with Epona's than you would with any other ring choice) to use Epona's on any weaponskill for anyone aside from mnk using Asuran, or a high triple attack thf. (Also possible: Dnc using Saber Dance; but then you need to account for its decay rate to determine the overall value, and I don't feel like working that out right now.)

And as a side note, your explanation for Asuran (8-hit) not benefiting from Epona's appears (from the way you wrote it) to make incorrect assumptions about how DA/TA hits interract with base number of swings. You get a max of 8 swings per round. Asuran is an 8 swing weaponskill. Therefore no DA/TA will *ever* apply to Asuran, even if you miss some of those 8 swings.

Aside to the side note: since you edited your original post, some of these comments may not make sense to anyone reading this after the fact.

*Edited, Apr 5th 2011 7:25pm by Kinematics *

Quote:

I assumed that if you hit 100 times, with 10% DA, thats 10 extra swings which also had a 95% chance to hit aka 9.5 extra hits. I wasn't aware a DA swing had 100% acc since i can find no evidence for that anywhere.

See the difference between a swing and a hit?

See the difference between a swing and a hit?

You're conflating the terms in your argument, including your explanation of how you're using them.

If you swing 100 times and add 10% DA, you swing 110 times.

If you hit 100 times and add 10% DA, you hit 110 times.

To hit 100 times with 95% acc means you swung 105.263 times; add 10% DA and you swing 115.789 times; re-apply accuracy and you hit 110 times.

In your quote above, however, you say you 'hit' 100 times, then add 10% DA and gain an extra 9.5 hits because of accuracy. This is wrong. A swing and a hit are indeed different, but you don't seem to actually understand that difference.

Despite the offense you took at it, Veggeto's explanation of the matter is exactly correct. If you add 10% DA that means you swing 10% more often, and by swinging 10% more often you hit 10% more often no matter what your accuracy is. 95%, 88%, 79%, 50%, doesn't matter; it's all the same. Therefore accuracy is irrelevant in a comparison of the value of DA/TA, except insofar as your accuracy itself may change due to the different gear choices (eg: Sniper Ring vs Epona's when below the acc cap).

Also, our conclusions are not quite identical. Aside from thf, there is no particular benefit in using most other rings in place of Epona even at the 6-hit level. Therefore it's worthwhile (ie: you won't do worse to any notable degree with Epona's than you would with any other ring choice) to use Epona's on any weaponskill for anyone aside from mnk using Asuran, or a high triple attack thf. (Also possible: Dnc using Saber Dance; but then you need to account for its decay rate to determine the overall value, and I don't feel like working that out right now.)

And as a side note, your explanation for Asuran (8-hit) not benefiting from Epona's appears (from the way you wrote it) to make incorrect assumptions about how DA/TA hits interract with base number of swings. You get a max of 8 swings per round. Asuran is an 8 swing weaponskill. Therefore no DA/TA will *ever* apply to Asuran, even if you miss some of those 8 swings.

Aside to the side note: since you edited your original post, some of these comments may not make sense to anyone reading this after the fact.

#9**dustinfoley**,
**Posted:** Apr 06 2011 at 12:26 PM, **Rating:** Sub-Default, (**Expand Post**) So the correct and polite answer would have been.

Scholar

1,270 posts

Double Attack: On multi-hit weapon skills, Double Attack can process on any hit of the weapon skill. Note, however, that Double Attack cannot process to induce a weapon skill to attempt to hit more than 8 times.

Triple Attack: On multi-hit weapon skills, (using Dual Wield) Triple Attack can occur twice (once on first hit and again on off hand hit).

Question:

These are both from the wiki. It says it wont attempt to hit more then 8 times fine i get that, so it will never fire on an 8 hit ws. (Thx kinematics for clarifying that)

For a 6 hit ws though, DA could fire on any of the 6 but as soon as its fired 2x, it wont attempt to fire for the remainder even if they miss and you only land say 5 hits (da da single)?

For TA on a 6 hit ws, it can only fire on hit 1 or 2, and once it fires it wont fire again correct?

I am trying to work on my calculations and I just want to make sure I have this correct.

Triple Attack: On multi-hit weapon skills, (using Dual Wield) Triple Attack can occur twice (once on first hit and again on off hand hit).

Question:

These are both from the wiki. It says it wont attempt to hit more then 8 times fine i get that, so it will never fire on an 8 hit ws. (Thx kinematics for clarifying that)

For a 6 hit ws though, DA could fire on any of the 6 but as soon as its fired 2x, it wont attempt to fire for the remainder even if they miss and you only land say 5 hits (da da single)?

For TA on a 6 hit ws, it can only fire on hit 1 or 2, and once it fires it wont fire again correct?

I am trying to work on my calculations and I just want to make sure I have this correct.

Scholar

1,423 posts

Quote:

who at least take the time to explain my mistakes/misconceptions to me.

Quote:

Because acc has 0 to do with DA/TA.

Let's say I hit for 1 dmg/hit and swing 100 times.

100 dmg.

Now let's say I hit for 1 dmg/hit and swing 100 times w/ 10% DA. 110 hits total

110 dmg.

110/100 = 10% increase.

Now let's say I only have 50% acc in both scenarios. I hit half as much, do half the amount of damage.

50 dmg vs 55 dmg.

55/50 = 10% increase.

No matter what your hitrate is, you get the same relative increases to your damage, which is why looking at acc is 100% irrelevant when looking at increases from da/ta

Yea, I didn't explain anything here.

and as a matter of fact, I don't even recall saying anything to be a **** here. I asked you why you would use a twilight belt, since you were counting it for DA.

That was a serious question (one that you haven't answered yet).

Also, who gets mad over an internet forum, especially a simple question, I wasn't even openly flaming you or anything.

http://cdn3.knowyourmeme.com/i/000/105/129/original/1299855073171.gif?1299857687

Scholar

1,423 posts

Quote:

For a 6 hit ws though, DA could fire on any of the 6 but as soon as its fired 2x, it wont attempt to fire for the remainder even if they miss and you only land say 5 hits (da da single)?

Its eight attempts, not 8 landed hits.

DW kraken and anything else (octave, ridill, etc.) Hit or miss, you'll never swing more than 8 times between the two in a single round.

Guru

2,236 posts

Quote:

So the correct and polite answer would have been.

1) Swings that miss can proc DA

2) DA doesnt have 100% acc

1) Swings that miss can proc DA

2) DA doesnt have 100% acc

No. Both of those are correct statments, but neither is an answer to the mistaken statements made, and each are so fundamental to the basics of how things work that there should be no reason to even bring them up when attempting to correct you.

Quote:

Veg may have been right but being a **** isn't needed over a video game

There are certainly plenty of opportunities to be annoyed at needless flaming on these and other message boards, but Veg's post (in this particular case, anyway) is not one of them. His explanation didn't delve deeply into the mechanics of how things work, but I would guess that he assumes he shouldn't need to for such a basic concept as DA procs.

Quote:

I go by the wiki and when there is no information or not enough, I just go by what I would assume makes the most sense

Well, this is a problem on your end, then (and wiki, to an extent), and it manifests in a large number of the posts you've made (I've had more than a few head-desk moments on reading your posts, but since they were usually mixed in with an entire thread of stupidity, I didn't feel like it was worth getting involved in). If you don't actually know how something works, you need to do more research, and/or ask questions for a better understanding. Your thread on exact haste values for gear was a simple and correct way to proceed.

wiki wrote:

Double Attack: On multi-hit weapon skills, Double Attack can process on any hit of the weapon skill. Note, however, that Double Attack cannot process to induce a weapon skill to attempt to hit more than 8 times.

Long ago there was speculation that DA could proc on any hit of a weaponskill. However a 3-DA TP return was never confirmed by anyone (testing with either Penta Thrust or Raging Rush).

Quote:

For a 6 hit ws though, DA could fire on any of the 6 but as soon as its fired 2x, it wont attempt to fire for the remainder even if they miss and you only land say 5 hits (da da single)?

Further, if the chance a DA could proc was maintained across all hits until 2 procs were realized, then weaponskills with a high number of hits (eg: Penta, DE, etc) would have abnormally high levels of DA procs compared to lower-hit weaponskills, which is not the case.

While difficult to be absolutely certain, the current evidence points strongly against the statement made on the wiki. It is likely, however, that noone ever updated the wiki (I'm not going to try to figure out when that particular comment was added).

Quote:

Triple Attack: On multi-hit weapon skills, (using Dual Wield) Triple Attack can occur twice (once on first hit and again on off hand hit).

This is also somewhat ambiguous (based on my knowledge). It's not certain that the second TA must come from the offhand weapon. It's known that you can only DA once on a 1-hit weaponskill (eg: Gekko), and that you can DA twice on a multi-hit weaponskill where you don't have an offhand weapon (eg: Raging Rush, Vorpal Blade, any H2H). However a single-hit weaponskill can DA twice if you're dual wielding (eg: Viper Bite).

As such I would say that you have one TA/DA attempt for the first hit, and one additional attempt if the weaponskill has any additional hits.

Scholar

1,270 posts

Veggeto wrote:

Quote:

For a 6 hit ws though, DA could fire on any of the 6 but as soon as its fired 2x, it wont attempt to fire for the remainder even if they miss and you only land say 5 hits (da da single)?

Its eight attempts, not 8 landed hits.

DW kraken and anything else (octave, ridill, etc.) Hit or miss, you'll never swing more than 8 times between the two in a single round.

Which on a 6 hit ws if DA fired 2 times, thats 8 attempted hits... not sure what your trying to correct...

Edit*

Thanks kine for clearing that up

Scholar

2,626 posts

Kine wrote:

It's known that you can only DA once on a 1-hit weaponskill (eg: Gekko)

Did you mean DA or TA? Otherwise, I don't see how anyone could ever think you can DA more than once for example on the same 1HIT ws (unless ppl actually thought somthing like that years ago lol).

Edit. Veg's post might of been sharp, but I didn't see any part where he was being a **** and there was certainly no flaming, and his question is valid.

Edit2. His post wasn't even sharp, he just started with a legitimate question, then carried on in his explanation.

Scholar

1,423 posts

dustinfoley wrote:

Veggeto wrote:

Quote:

Its eight attempts, not 8 landed hits.

DW kraken and anything else (octave, ridill, etc.) Hit or miss, you'll never swing more than 8 times between the two in a single round.

Which on a 6 hit ws if DA fired 2 times, thats 8 attempted hits... not sure what your trying to correct...

Edit*

Thanks kine for clearing that up

I was half asleep and thought you asked something completely different, lol.

Scholar

1,054 posts

Quote:

However, a few quick checks indicate that Epona's is worth about:

Mnk: 2.7% (Smite, /nin) to 2.5% (Smite, /war)

Pup: 2.1% (Pummel, /nin) to 1.9% (Pummel, /war)

Thf: 1.6% (Evis, /nin) to 1.5% (Evis, /war)

Mnk: 2.7% (Smite, /nin) to 2.5% (Smite, /war)

Pup: 2.1% (Pummel, /nin) to 1.9% (Pummel, /war)

Thf: 1.6% (Evis, /nin) to 1.5% (Evis, /war)

Normally I dont post here but I was interested in how you came up with Evis being worth 1.6% with Eponas.

As /nin, I would guess a gorget adding 0.1 to first attack and Brutal+Aetheling with 8% DA, and natural 10% TA.

This would bring the avg to 6.644fTP. 6 (base) + 0.1 (gorget) + 2x(2x0.1) + 2x(0.9x0.08)

Adding Eponas would be 6.8114. 6(base) + 0.1 (gorget) + 2x(2x0.13) + 2x(0.87x0.11)

6.8114/6.644 = 2.52%

Please show me(step by step) how my thinking is flawed if 1.6% is correct.

Quote:

You can't transform numbers into other numbers like that. It'd just go on forever. That's like witchcraft.

There's one guy I know who

Quote:

is a big deal now.

Guru

2,236 posts

Evis: 5 hit, 1 offhand, 1.0 fTP

Base fTP: 6.0 + .1 (gorget)

TA: 10% trait/merits, 15% Apoc atma, 3% Triplus Dagger = 28%

DA: 5% Brutal + 3% Atheling = 8%

avg ftp: 6.1 + .28*2 + .72*.08 = 6.7176

Add Epona's

avg fTP: 6.1 + .31*2 + .69*.11 = 6.7959

6.7959 / 6.7176 = 1.17% increase

Leaving off Apoc atma and Triplus dagger:

6.1 + .1*2 + .9*.08 = 6.372

vs

6.1 + .13*2 + .87*.11 = 6.4557

6.4557 / 6.372 = 1.3%

Both of which are actually lower than my earlier quick estimate, so I may need to take another look at this.

In any case, your math is notably off in both calculations. You're giving 2x value for DAs when they should be 1x (1 additional hit; TA gets 2x for 2 additional hits). Even then, total for the case without Epona's should be 6.444 if you gave DAs 2x value, not the 6.644 you ended up with. Similar issue with the calculations with Epona's.

Base fTP: 6.0 + .1 (gorget)

TA: 10% trait/merits, 15% Apoc atma, 3% Triplus Dagger = 28%

DA: 5% Brutal + 3% Atheling = 8%

avg ftp: 6.1 + .28*2 + .72*.08 = 6.7176

Add Epona's

avg fTP: 6.1 + .31*2 + .69*.11 = 6.7959

6.7959 / 6.7176 = 1.17% increase

Leaving off Apoc atma and Triplus dagger:

6.1 + .1*2 + .9*.08 = 6.372

vs

6.1 + .13*2 + .87*.11 = 6.4557

6.4557 / 6.372 = 1.3%

Both of which are actually lower than my earlier quick estimate, so I may need to take another look at this.

In any case, your math is notably off in both calculations. You're giving 2x value for DAs when they should be 1x (1 additional hit; TA gets 2x for 2 additional hits). Even then, total for the case without Epona's should be 6.444 if you gave DAs 2x value, not the 6.644 you ended up with. Similar issue with the calculations with Epona's.

Scholar

1,054 posts

Kinematics wrote:

Evis: 5 hit, 1 offhand, 1.0 fTP

Base fTP: 6.0 + .1 (gorget)

TA: 10% trait/merits, 15% Apoc atma, 3% Triplus Dagger = 28%

DA: 5% Brutal + 3% Atheling = 8%

avg ftp: 6.1 + .28*2 + .72*.08 = 6.7176

Add Epona's

avg fTP: 6.1 + .31*2 + .69*.11 = 6.7959

6.7959 / 6.7176 = 1.17% increase

Leaving off Apoc atma and Triplus dagger:

6.1 + .1*2 + .9*.08 = 6.372

vs

6.1 + .13*2 + .87*.11 = 6.4557

6.4557 / 6.372 = 1.3%

Both of which are actually lower than my earlier quick estimate, so I may need to take another look at this.

In any case, your math is notably off in both calculations. You're giving 2x value for DAs when they should be 1x (1 additional hit; TA gets 2x for 2 additional hits). Even then, total for the case without Epona's should be 6.444 if you gave DAs 2x value, not the 6.644 you ended up with. Similar issue with the calculations with Epona's.

Base fTP: 6.0 + .1 (gorget)

TA: 10% trait/merits, 15% Apoc atma, 3% Triplus Dagger = 28%

DA: 5% Brutal + 3% Atheling = 8%

avg ftp: 6.1 + .28*2 + .72*.08 = 6.7176

Add Epona's

avg fTP: 6.1 + .31*2 + .69*.11 = 6.7959

6.7959 / 6.7176 = 1.17% increase

Leaving off Apoc atma and Triplus dagger:

6.1 + .1*2 + .9*.08 = 6.372

vs

6.1 + .13*2 + .87*.11 = 6.4557

6.4557 / 6.372 = 1.3%

Both of which are actually lower than my earlier quick estimate, so I may need to take another look at this.

In any case, your math is notably off in both calculations. You're giving 2x value for DAs when they should be 1x (1 additional hit; TA gets 2x for 2 additional hits). Even then, total for the case without Epona's should be 6.444 if you gave DAs 2x value, not the 6.644 you ended up with. Similar issue with the calculations with Epona's.

I didnt give DA 2x value, the x2 (in bold on quote below)was showing it could proc on hit #1 or hit #2. You are only showing the +TA +DA as only procing on the first hit.

Quote:

Adding Eponas would be 6.8114. 6(base) + 0.1 (gorget) + __2x__(2x0.13) + __2x__(0.87x0.11)

Am I mistaken it thinking that DA and TA has a chance to proc on first 2 hits?

Or rather am I missing the overflow possiblity(max 8 hits) that eliminates the chance of proc on the second hit? If so please demonstrate mathematically so I can follow.

Quote:

You can't transform numbers into other numbers like that. It'd just go on forever. That's like witchcraft.

There's one guy I know who

Quote:

is a big deal now.

Guru

2,236 posts

Quote:

I didnt give DA 2x value, the x2 (in bold on quote below)was showing it could proc on hit #1 or hit #2. You are only showing the +TA +DA as only procing on the first hit.

Ack, nope, you're right. I've really been slipping the last few posts, haven't I?

Baseline (no Apoc atma, no Triplus):

No Epona: 6.0 + 0.1 + 2 * (0.1*2 + 0.08*0.9) = 6.644

Epona: 6.0 + 0.1 + 2 * (0.13*2 + 0.11*0.87) = 6.8114

Gain: 6.8114 / 6.644 = 2.5%

So, yes, you're correct on the raw percentage increase. But also yes, you're missing how much is lost from 8+ hit overflow.

The total number of hits across all hit distributions (but not full combinatorials) is 1850 with an 8 hit cap, and 1920 with no cap. Average number of hits (including combinatorials, but no Epona's) with an 8 hit cap is 6.45; average number with no cap is 6.49.

With an 8 hit cap there's a ~1.94% increase with Epona's; with no cap there's a ~2.26% increase in overall damage.

There's some extra lossiness that I'm apparently missing or forgetting (don't have time for a thorough examination right now). Add Apoc atma and the gain with Epona's drops to 1.5% with an 8 hit cap, while it would be 2.0% with no cap.

415 posts

For the purpose of accurately modeling the ws damage, are you certain that TA is checked prior to DA. Mdk just told me that when using assassin's charge, you can get the TP for a double attack (=if DA proc), and he told me this was shown by CDF long time ago on BG, when people on that site actually knew sh*t and didn't spend their time commenting official forums.

*Edited, Apr 25th 2011 8:02am by lynnminmay *

Guru

2,236 posts

Based on parsed data (not my own), I'm pretty sure I remember the margin of error to be sufficiently small that you could conclusively (95%) say that TA proc'd before DA for melee. I don't have any reason to expect that it would be in a different order on weaponskills, however if there's real counterevidence to the TA>DA order, I'd love to see it.

415 posts

I'm doing an om'hpende test on THF/WAR level 76 with 5/5 triple attack merit and brutal earring (15% DA and 10% TA). Correct me if I'm wrong but

* If DA is checked first, the rate of 2 swings- rounds is 15% and the rate of 3 swings rounds is (1-0.15)*0.10=8.5%. Also the rate of multiple swings per round is 23.5%.

*If TA is checked first, the rate of 3 swings- rounds is 10% and the rate of 2 swings rounds is (1-0.10)*0.15=13.5%. Also the rate of multiple swings per round is 23.5%.

* If DA is checked first, the rate of 2 swings- rounds is 15% and the rate of 3 swings rounds is (1-0.15)*0.10=8.5%. Also the rate of multiple swings per round is 23.5%.

*If TA is checked first, the rate of 3 swings- rounds is 10% and the rate of 2 swings rounds is (1-0.10)*0.15=13.5%. Also the rate of multiple swings per round is 23.5%.

415 posts

Ok results of the above test

margins of error :

TA rate= 8.31% +/- 0.57%

DA rate= 13.38% +/- 0.71%

multi-attack%=21.71% +/- 0.86%

At best TA rate is 8.88% so this excludes TA being checked first

At best DA rate is 14.09% so this excludes DA being checked first

At best multi attack rate is 22.57% so this excludes both models alltogether

The only explanation I have found so far is that if DA and TA "proc" at the same time, none of them appears. With this model :

DA rate = 0.15*(1-0.10) = 13.5% lies within the above confident interval

TA rate = 0.10*(1-0.15) = 8.5% lies within the above CI

multiattack rate : 0.15*(1-0.10) + 0.10*(1-0.15)= 22% lies within the CI

You heard it here first : 100% TA rate and 100% DA rate means 0% multiattack rate. Suck less SE.

[for reference, my thf gear was : signet staff, lamiabane, turban, rapparee, dusk gloves +1, morgana's choker, brutal, novia, hoard ring, toreador, acerbic sash, skadi's chaussee, wood F ledelsens, total STR = 62 ]

*Edited, May 1st 2011 9:46am by lynnminmay *

*Edited, May 1st 2011 10:23am by lynnminmay *

Melee DataPimpchan 11474 8819 1 2655Player # Melee Attacks # Melee Rounds Attacks/Round # Extra AttacksPimpchan 1180 733 0 1 1Player # +1 Rounds # +2 Rounds # +3 Rounds # +4 Rounds # >+4 RoundsPimpchan 1915 21,71 % 0 0Player # MultiAttack Rounds MultiAttack % Kills w/Min Attacks Kills w/<Min AttacksTreat As:Multi-attacks per attack (2x/3x):Pimpchan 1180 13,38 % 61,68 % 733 8,31 % 38,32 %Player # Double Attacks DA Rate Perc. DA # Triple Attacks TA Rate Perc. TAMulti-attacks per round (Kicks):Pimpchan Yes 1180 14,60 %Player Footwork? # Rounds w/Kicks Kick Attack RateMulti-attacks per attack (Zanshin):Pimpchan 7043 947 13,45 %Player # Missed First Attacks # DA w/Missed First Possible Zanshin %Pimpchan 20,14 % 18,69 %Player Acc. Rate of First Attacks Acc. Rate of Second Attacks

margins of error :

TA rate= 8.31% +/- 0.57%

DA rate= 13.38% +/- 0.71%

multi-attack%=21.71% +/- 0.86%

At best TA rate is 8.88% so this excludes TA being checked first

At best DA rate is 14.09% so this excludes DA being checked first

At best multi attack rate is 22.57% so this excludes both models alltogether

The only explanation I have found so far is that if DA and TA "proc" at the same time, none of them appears. With this model :

DA rate = 0.15*(1-0.10) = 13.5% lies within the above confident interval

TA rate = 0.10*(1-0.15) = 8.5% lies within the above CI

multiattack rate : 0.15*(1-0.10) + 0.10*(1-0.15)= 22% lies within the CI

You heard it here first : 100% TA rate and 100% DA rate means 0% multiattack rate. Suck less SE.

[for reference, my thf gear was : signet staff, lamiabane, turban, rapparee, dusk gloves +1, morgana's choker, brutal, novia, hoard ring, toreador, acerbic sash, skadi's chaussee, wood F ledelsens, total STR = 62 ]

Guru

2,236 posts

Quote:

I'm doing an om'hpende test on THF/WAR level 76 with 5/5 triple attack merit and brutal earring (15% DA and 10% TA). Correct me if I'm wrong but

* If DA is checked first, the rate of 2 swings- rounds is 15% and the rate of 3 swings rounds is (1-0.15)*0.10=8.5%. Also the rate of multiple swings per round is 23.5%.

*If TA is checked first, the rate of 3 swings- rounds is 10% and the rate of 2 swings rounds is (1-0.10)*0.15=13.5%. Also the rate of multiple swings per round is 23.5%.

* If DA is checked first, the rate of 2 swings- rounds is 15% and the rate of 3 swings rounds is (1-0.15)*0.10=8.5%. Also the rate of multiple swings per round is 23.5%.

*If TA is checked first, the rate of 3 swings- rounds is 10% and the rate of 2 swings rounds is (1-0.10)*0.15=13.5%. Also the rate of multiple swings per round is 23.5%.

The premise seems accurate.

Correcting for the +4 and >+4 rounds won't really impact the final values, so...

Just, wow. I would not have expected this result.

Yes, it looks very much like the chance that either of them will proc is exclusive of the chances for the other to proc. The problem is determining, programmatically, which one actually -did- proc.

My first thought was that SE (against reasonable expectations) set up the odds for both of them as (1-OtherN)*N%. So, 13.5% for DA and 8.5% for TA. Except that you still have to figure out which of them proc'd. I suppose it could be something like a stacked comparison:

oddsOfDA = (100% - (TAPercent + QAPercent)) * DAPercent; oddsOfTA = (100% - (DAPercent + QAPercent)) * TAPercent; oddsOfQA = (100% - (DAPercent + TAPercent)) * QAPercent; randomCheck = random(); if (randomCheck <= oddsOfDA) ProcDA(); else if (randomCheck <= (oddsOfDA + oddsOfTA)) ProcTA(); else if (randomCheck <= (oddsOfDA + oddsOfTA + oddsOfQA)) ProcQA();

Instead of simply using the expected sequential order check:

proc = false; if (QAPercent > 0) { randomCheck = random(); if (randomCheck <= QAPercent) { ProcQA(); proc = true; } } if ((proc == false) && (TAPercent > 0)) { randomCheck = random(); if (randomCheck <= TAPercent) { ProcTA(); proc = true; } } if ((proc == false) && (DAPercent > 0)) { randomCheck = random(); if (randomCheck <= DAPercent) { ProcDA(); proc = true; } }

While there are additional ways to optimize the code, essentially it looks like they want to minimize the number of random() calls made (potentially significantly more expensive than the local operations).

The first bit of code could be made to work properly if you redid the calculations to effectively give one of the proc types priority over the others. Or perhaps it was intentional; perhaps large amounts of DA+TA is *supposed* to be detrimental. Impossible to say, really.

Would be interesting to see what sort of response this would get on the official boards...

This is also annoying since it means I'll have to redo the calculations in the spreadsheets for figuring out the impact of DA and TA (which weren't exactly a picnic to begin with).

415 posts

Moar data. I added pole grip to the previous gear, which puts me at 17% DA and 10% TA

This completely confirms what is happening

expected rate if TA is checked first (*)

TA : 10%

DA : (1-0.10)*1.17=15.3%

expected rate if DA is checked first (**)

DA : 17%

TA : (1-0.17)*1.10=8.3%

Observed rates (with confident interval)

TA : 8.04% +/- 0.57%

DA : 15.62% +/- 0.76%

The observed TA rate excludes (*) by a large margin (~1.5% difference at least)

The observed DA rate excludes (**) by a decent margin (~0.6% difference at least)

Obviously the expected rates are the DA rate from (*) and the TA rate from (**) respectively which confirms what I concluded in the previous post, as in : when the game rolls a DA and a TA, none occur.**Problem**.

Note 1 : the above confident intervals are done at 95% likelyood.

Note 2 : My MONK/WAR with atheling/epona/brutal and apocalypse has 21% DA and 18% which means that I'm only getting 17.2% actual DArate and 14.22% TA rate. If I was single wielding my signet staff it's a drop of my avergae #hits/per round of about -3.3% compared to the previous model.

*Edited, May 2nd 2011 11:24am by lynnminmay *

Melee DataPimpchan 11435 8672 1 2763Player # Melee Attacks # Melee Rounds Attacks/Round # Extra AttacksPimpchan 1355 697 3 0 1Player # +1 Rounds # +2 Rounds # +3 Rounds # +4 Rounds # >+4 RoundsPimpchan 2056 23,71 % 0 0Player # MultiAttack Rounds MultiAttack % Kills w/Min Attacks Kills w/<Min AttacksTreat As:Multi-attacks per attack (2x/3x):Pimpchan 1355 15,63 % 66,03 % 697 8,04 % 33,97 %Player # Double Attacks DA Rate Perc. DA # Triple Attacks TA Rate Perc. TAMulti-attacks per round (Kicks):Pimpchan Yes 1355 17,00 %Player Footwork? # Rounds w/Kicks Kick Attack RateMulti-attacks per attack (Zanshin):Pimpchan 6915 1094 15,82 %Player # Missed First Attacks # DA w/Missed First Possible Zanshin %Pimpchan 20,26 % 17,00 %Player Acc. Rate of First Attacks Acc. Rate of Second Attacks

This completely confirms what is happening

expected rate if TA is checked first (*)

TA : 10%

DA : (1-0.10)*1.17=15.3%

expected rate if DA is checked first (**)

DA : 17%

TA : (1-0.17)*1.10=8.3%

Observed rates (with confident interval)

TA : 8.04% +/- 0.57%

DA : 15.62% +/- 0.76%

The observed TA rate excludes (*) by a large margin (~1.5% difference at least)

The observed DA rate excludes (**) by a decent margin (~0.6% difference at least)

Obviously the expected rates are the DA rate from (*) and the TA rate from (**) respectively which confirms what I concluded in the previous post, as in : when the game rolls a DA and a TA, none occur.

Note 1 : the above confident intervals are done at 95% likelyood.

Note 2 : My MONK/WAR with atheling/epona/brutal and apocalypse has 21% DA and 18% which means that I'm only getting 17.2% actual DArate and 14.22% TA rate. If I was single wielding my signet staff it's a drop of my avergae #hits/per round of about -3.3% compared to the previous model.

**B***i*__U__~~S~~
Smileys

**Post**

© 2018 ZAM Network LLC