Ruke's All-Purpose Thread of DA/Askar/Haste/etc and Win!

[RKenshin]

Basically.

As your damage per swing increases, Askar gets closer to matching and eventually passing Ares.

As it decreases, Ares gets further ahead of Askar.

However, outside of an extremely decked & merited DRK in a buffed out situation, you won't normally hit high enough damage/hit for Askar to surpass Ares. Although you probably could come very close to matching it without being a super DRK.

Bottom line is the same though, right? That Askar body pails in comparison to Ares body?

I wouldn't say that it pails in comparison, it just falls a small bit behind in most situations for most people.

The difference is incredibly small either way.

[Max™]

Course, one situation I can think of which doesn't involve 250 Dmg Avg hits where I'd go Askar > Are's, Bravura trying to get 6 hit, or if I was trying to keep gear for War, Drk, Sam, Pld, Drg in some combination which wouldn't be covered by a Hauberk, Hauby, or Are's body.

[RKenshin]

If Askar completes a 6-hit setup, it is always better than Ares regardless of the situation. It even appears better than/comparable to Adaman Hauberk for most situations.

A guesstimated 5% acc and 5 extra damage per hit < 2% DA and one less hit to 100 TP... By a total damage difference of a little under 2% (pretty good amount). 5% acc may be understated though, depending.

[Valyana]

and we have never really had good confidence for how to get a useful pdif_average in the area where either pdif_max or pdif_min are constrained to 1.0. (above and below this area we've always assumed that pdif_average is the midpoint between min and max which may also be untrue).

I'm quite sure that's untrue, which I brought up in ye olde Nagamaki thread. In the midpoint of the pDIF graph, pDIF = 1.0 has a higher probability than other values; it is easy to see that there is one number coming up more often than others, and calling that pDIF = 1.0 has worked with the other formulas.

Since low and high cRatio give a 0.8 pDIF range, while around cRatio of 1 the range is only 0.5, my hypothesis was that if the random number generator gives a number in the missing 0.3, it just gets mapped to pDIF 1.0.

[Amele]

I'm quite sure that's untrue, which I brought up in ye olde Nagamaki thread. In the midpoint of the pDIF graph, pDIF = 1.0 has a higher probability than other values; it is easy to see that there is one number coming up more often than others, and calling that pDIF = 1.0 has worked with the other formulas.

Since low and high cRatio give a 0.8 pDIF range, while around cRatio of 1 the range is only 0.5, my hypothesis was that if the random number generator gives a number in the missing 0.3, it just gets mapped to pDIF 1.0.

briefly; conflating mode and average is a mistake. while mode = average = median in a normal distribution, we have no reason to expect that this is true for the range in discussion; and moreover because the formula is constraining to a particular value regardless of where it falls in the range we can provably demonstrate that the mode isn't the average for portions of the equation.(it's trivial to demonstrate this at the upper and lower ends). Using 1.0 for average in the range of 0.5-1.5 is literally saying that attack has no value and that's patently untrue.

as for 'calling it' 1.0; we're not calling it anything. the value of the symbol is absolute in the current model (i.e. a 1.0 is actually used mathematically as a 1.0) it's not some rating scale, it's a numeric value.

your theory about how it's constrained may very well be correct; but there's alot of testing to do before we can say that with confidence.

[Valyana]

I must not have been clear enough, as we seem to be in violent agreement. I was arguing that mode = pDIF 1.0, which will usually be different from the average of max and min. And different from the average pDIF as well.

[Amele]

violent agreement is a good term for it; yes, that's what I was arguing too; lol

[Miraun]

Heya.

So yeah, I'm a little surprised at the Ares versus Askar discussion that I saw here, which, as I read it, kinda broke down to, "If Askar doesn't give you enough Store TP to cut down Wsing by 1 hit / 100 TP, then Ares will do better"

I also read about the discussion about the marginal diminishing returns of stacking +DA, as the rate of change from 10 to 12% is less significant than the change from 0 to 2%.

So here's a new alternative gearpiece to discuss in comparison to Askar/Ares. For DRGs and DRKs, what about Homam Corazza, with the 2.5% TA rate.

For sake of comparison, lets differentiate the ACC value influence between a job that's getting 100% utilization of the +15 ACC, and a job that's only utilizing +10 ACC.


BTW: Ruke, I still owe you pizza.

[Nameless]

I am kinda curious on the new stuff.

As a GS drk, what combo would you use? Algol + Aurum Cuirass + Store TP +1 ring for tp for a 6 hit build with Ares as body piece for ws; or Naglering + Askar body for tp for a 6 hit build with same ws piece.

Accuracy is not capped for being realistic (let's assume 85% with capped DE). 292 skill on both set with homam leg feet glove, askar helm, swift belt and sea torque. Second ring slot for Naglering set can be either toreader or toau ring. WS choice is spinning slash, so acc during ws shouldn't be a big deal consider it is a one hit ws.

I personally think algol might be better, and I think the lower delay and 3% TA is more than enough to offset the 2% DA from askar body and 17 more attack from Naglering not to mention the fire damage. Aurum also got +8 Dex which will offset the +4 acc from Naglering and any addition acc from not using the store tp +1 ring.

What is your take on a mathatical stand point?

[Amele]

Heya.

So yeah, I'm a little surprised at the Ares versus Askar discussion that I saw here, which, as I read it, kinda broke down to, "If Askar doesn't give you enough Store TP to cut down Wsing by 1 hit / 100 TP, then Ares will do better"

I also read about the discussion about the marginal diminishing returns of stacking +DA, as the rate of change from 10 to 12% is less significant than the change from 0 to 2%.

So here's a new alternative gearpiece to discuss in comparison to Askar/Ares. For DRGs and DRKs, what about Homam Corazza, with the 2.5% TA rate.

For sake of comparison, lets differentiate the ACC value influence between a job that's getting 100% utilization of the +15 ACC, and a job that's only utilizing +10 ACC.


BTW: Ruke, I still owe you pizza.

you just want to hurt my brain! D:

very briefly: if you're getting substantial useful accuracy out of the body slot you should continue to use accuracy based bodies.

less briefly; pretending that warrior can wear homam (so I can use all the numbers up above) you'll see that the war in ares was getting about 6.2% more dps in TP out of wearing ares over the generic baseline.

accuracy +15 is going to be 8.5% dps boost (hit rate +7.5%, 95/87.5) and that's completely ignoring the TA.

accuracy +10 is going to be a 5.5% dps boost (hit rate +5%, 95/90) the balance of the 6.2% should be easily made up by the 2.5% triple attack (best case dps bonus of about 5%).

[RKenshin]

I still think that approximating 2 acc = 1% hit-rate is a little over-exaggerated, I think it's close to 3-4 acc = 1%.

Heya.

So yeah, I'm a little surprised at the Ares versus Askar discussion that I saw here, which, as I read it, kinda broke down to, "If Askar doesn't give you enough Store TP to cut down Wsing by 1 hit / 100 TP, then Ares will do better"

I also read about the discussion about the marginal diminishing returns of stacking +DA, as the rate of change from 10 to 12% is less significant than the change from 0 to 2%.

So here's a new alternative gearpiece to discuss in comparison to Askar/Ares. For DRGs and DRKs, what about Homam Corazza, with the 2.5% TA rate.

For sake of comparison, lets differentiate the ACC value influence between a job that's getting 100% utilization of the +15 ACC, and a job that's only utilizing +10 ACC.


BTW: Ruke, I still owe you pizza.

Plugging in guesstimated numbers on what the differences between those three bodies would be in damage/hit averages and acc averages, it looks like this:
http://img232.imageshack.us/img232/8707/ffxi361ap7.jpg

Homam wins, even if you assume that only 10 of the 15 acc is in use. Slight mathematical advantage given to homam though, due to the fact that in reality TA is slightly less effective since you could land the killing blow to a mob on a TA proc and never swing another 2 hits. So maybe take off .25% or so from the total of homam's damage.

This is also assuming all sets already have a 6-hit.

Now gimme mah pizza!

[Callisto]

Edit: Nvm missed you explain it.

[Khrno]

Where does Aurum body stand in this body ranking? And HQ Hauby?

Adaberk > Homam > Askar > Ares

(Askar would be over Homam if the stp is needed for the 6-hit right?)

[RKenshin]

To get really specific (and much more accurate) I'd actually need some real numbers to plug in, not my guesstimated ones.

Also keep in mind the averages I used were rather high, so as you see Askar takes a lead ahead of Ares at an average that high. But usually an average wouldn't be that high I'd think for most players.

Aurum would likely fall in behind Ares though, at least at those averages. I'll play with more numbers later, about to leave work!

[Nameless]

I think hit build is got to be one of the most if not the most important thing for 2 handers. It probably out weights a big amount of accuracy or attack. Tho I am interested in finding out a breakeven point for hit build vs other factors.

Tho thinking about it, aurum's body is probably only really good for GS drk as jobs like war/sam and drg/sam can reach 6 hit build fairly easily with multi hit WS. And in that case, aurum probably crushs ares consider both are big accuracy drop but ares gives you a 6 hit build.

[Khrno]

I think hit build is got to be one of the most if not the most important thing for 2 handers. It probably out weights a big amount of accuracy .

What does it matter if you can get 100% TP in 5 hits (or 4 or 6), if in the end you end up hitting those 5 and missing 2 on the way to 100%?

Most important factor is Acc, I understand that much better now, after that then throw all the stats you want, but if you don't hit, it doesn't matter how fast or how many times you need to hit, just remember that, and don't discard Acc so easily.

[Tan]

What does it matter if you can get 100% TP in 5 hits (or 4 or 6), if in the end you end up hitting those 5 and missing 2 on the way to 100%?

Most important factor is Zanshin, I understand that much better now, after that then throw all the stats you want, but if you don't hit you get another chance to hit, it doesn't matter how fast or how many times you need to hit, just remember that, and don't discard Zanshin so easily.

Fixed for Hachiryu set. But I've gotta agree with Khrno, it won't matter how fast you swing or how many hits it takes you to get 100% TP, if you can't hit the damn mob. However, I kinda feel that some SAM's and /SAM's overlook Zanshin as the "gimp double attack" that it is. Personally I'm curious about what would be the best Acc % to make the most use out of SAM's Ikitoshen if it was full merited, especially with the plethora of Zanshin gear that is out there. Anyone else feel the same way as me? Also anyone have any more data or testing on the "Enhances Zanshin" for various gear pieces and/or activation rate etc?

I know this was kinda off topic but decided to throw it in since on the elusive but ever sexy Hachiryu Body which could potentially outdamage those for TPing.

[Amele]

Ruke, I dunno really. - nagamaki proved the acc = hit rate relationship to a pretty absurdly accurate level.. I mean if you think it needs retested I don't exactly have standing to say 'no you're wrong!' lol.

[Max™]

I think he's referring to how at higher levels of accuracy, such as the ones we 2hers get into these days, that it feels more like 3~4=1%.

Just what I thought when I read that.

[Khrno]

Fixed for Hachiryu set.

I know this was kinda off topic but decided to throw it in since on the elusive but ever sexy Hachiryu Body which could potentially outdamage those for TPing.

I popped Hydra znm the last week and both Enkidu body and Hachiryu feet dropped, after looking at feet I was wondering too what would be the extend of relying on Zanshin, to benefit of the Hachiryu stats, however even after getting the full set and the Store TP bonus (which well could be 5 as most other bonus set magic numbers is 5) you will be missing from 10 to 13 Haste in hands/legs/feet. Even for a SAM/DRG it would still be a huge loss of Haste.

(Oh and to end up the little story of the znm, I took Enkidu body for my NIN since it is a super awesome Scorpion Hauby :D)





And as Disclaimer: Hachiryu Haramaki will never be a better body piece than Adaman Hauberk ...






...for DRK (and BST lol).