Yuki/Gekko/Kasha Attack bonus Testing

[Nameless]

Let me try this assumption and plug in all my parse result and see if it fits.

[Nameless]

Hmm.... I plug that assumption in, and it shows that at bird I should average 920 SS, which is definitely not true. Either the test sample is too small, or there is other tricks to it. As of now, I still think that Ratio has a cap, which results in the low numbers. Not sure how to explain yours tho.

[WraithLakshmi]

So is it possible then, that in high attack situations, where your cratio is capped normally, that gekko could be inferior to rana since gekko will not improve from the increased attack, but rana will?

No. Even if both have capped pDIF, the WSC mod on Rana is so low, it can't really ever catch up to Gekko.

[WraithLakshmi]

I think the bonus being applied after level correction is correct. I calculated it for being applied before, and my average gekko was calculating to be like 150 higher than I typically see in parsing. When I applied it after level correction, it was right where I generally parse.

[Nameless]

Still wouldn't fit my parse, apply before or after. Since my cratio on bird would still be capped even if the bonus is applied after lvl correlation. Again, I stand by that Ratio has a cap; I haven't read anywhere saying that Ratio doesn't have a cap. It could be a blind spot that everyone thought of one thing but it turns out to be not right. Just like everyone used to automatically think that haste spell + double SV march would be 55% haste when it caps at 43.3% (43.75% to be exact). I just never seen my SS hit over 1k without DA at bird, when I should have done it often based on the assumption that Ratio doesn't have a cap.

[Tricen]

Can you reiterate to us what all of this means in an anology that most people would get?

[Nameless]

Ratio = att/mob's def
cRatio = att/mob's def after lvl correlation, 0.05 per mob lvl higher than yours
Y/K/G, SS, GS Multiplier = special attack bonus linking to those above WSs

The general belief is that your attack/mob's def has no cap, so if your attack is 900, mob's def is 300, you have a 3.0 Ratio. This number then get multiplied by the multiplier and then get reduced by lvl correlation. So for Gekko, if the mob is 10 lvls higher than you, you get 3.0 x 2.0 - 0.5 = 5.5 cRatio. However, we all know that cRatio has a cap, around 2.2 or 2.3, so that 5.5 will then get capped to 2.2 or 2.3. The cRatio then decide your pDIF max and min by multiplying the 2.2 or 2.3 by 1.25 or 1.2 so your max pDIF will be 2.75, and min pDIF will be 1.95 (max - 0.8 ).

My theory is that the Ratio itself is capped the same as cRatio, or in other words, it is that the Ratio that has a cap instead of cRatio (cRatio will always be either same or smaller than Ratio, so if Ratio has a cap, cRatio will be at cap or less). So in the above case, Ratio after applying the multiplier will be capped at 2.2 or 2.3, and then reduce by the lvl correlation of 0.5 to 1.7 or 1.8.

This model fits all the parse results I have for Spinning slash, which has a 1.5 multiplier. And I believe this is easy to test, because this theory basically means that against mob higher lvl than you, you will never be able to reach the 2.2 or 2.3 cap due to the fact that lvl correlation will be applied after the cap. So obviously against mob like kirin, regardless how high your attack is, your will never have higher than 1.2 or 1.3 cRatio against it (assuming that kirin is lvl 95). And that fits my personal experience very nicely.

[shaddix]

No. Even if both have capped pDIF, the WSC mod on Rana is so low, it can't really ever catch up to Gekko.

What about with oni instead?

[Deemos]

Talkin about Rana, when i have zerk and minuet up, i can remember doing 700-800 with Hagun and now aprox 1200-1500 with peaks of 1650 on dmg:88. But when atk is below 500, it sux alot, even with that base dmg.

[Yarko]

Parsed another SAM/NIN vs SW Serket and the
cRatio for Gekko = (Ratio - Correction)*2
fits pretty much exactly.

The point Nameless brought up has some truth though; I had a BRD+COR party on DRK lately, and my attack averaged at 650 at most times (but went as high as 850 with Last Resort) but Spinning Slash average was lower than expected, and I didnt get any 1000+ SS without DAs either, which would be possible if my cRatio was capped. cRatio for SS seemed to be in the 1.9~2.0 range instead.
So there is a cap somewhere that doesn't allow this hidden attack boost to let you have as high WSs as those on Too Weak mobs on high level monsters.... I just can't quite figure where it is in a formula like cRatio = (Ratio - Correction)*Multiplier.
Maybe it does a different check, and the value caps at "Capped cRatio - Correction", which would explain the 1.9~2.0 I got on birds. What if the Ratio itself was high enough to reach the cap without the Multiplier boost though? Would the value still get that cap? If this was tested the whole thing would be easier to figure.

[Nameless]

I am personally leaning towards the assumption that the multiplier is applied after cRatio, and that the cRatio is capped at 2.2 or 2.3 minus lvl Correction. This will pretty much fit both of our data. I will test it on paper tonight when I get home to see if this theory fits my parse result.

Good work btw to confirm my SS result, which shows that there is clearly a cap somewhere that won't allow the cratio to reach 2.2 or 2.3 against higher lvl mobs.

[WraithLakshmi]

Hopefully someone else in here can make some sense of these results I just got. I was fighting Steelshells, trying to pin down what cRatio is multiplied by to get pDIF(max) and pDIF(min). I meleed in 85 str and 390 attack, with soboro.

LEVEL 76: cRatio = 1.004
High - 53, pDIF = 1.359
Low - 28, pDIF = .718

LEVEL 75: cRatio = 1.065
High - 55, pDIF = 1.411
Low - 31, pDIF = .795

LEVEL 74: cRatio = 1.083
High - 56, pDIF = 1.400
Low - 32, pDIF = .800

LEVEL 73: cRatio = 1.101
High - 56, pDIF = 1.400
Low - 33, pDIF = .825



Highs: pDIF range, Multiplier range
76: 1.359 - 1.384, 1.353 - 1.378
75: 1.411 - 1.435, 1.325 - 1.347
74: 1.400 - 1.424, 1.293 - 1.315
73: 1.400 - 1.424, 1.272 - 1.293

Lows: pDIF range, Multiplier range
76: .718 - .743, .715 - .740
75: .795 - .820, .746 - .770
74: .800 - .824, .739 - .761
73: .825 - .849, .749 - .771

First thing I noticed: the multipliers for pDIF(max) aren't in the same ranges. However, the multipliers for pDIF(min) for levels 73, 74, and 75 are all around .75. In fact, the pDIF(max) values all appear to be equal to pDIF(min) + .6. So it would appear that pDIF(min) is calculated first, and pDIF(max) is based on that result, backwards from how it used to be done.

The only thing that doesn't fit in with this hypothesis are the values for level 76 crabs. The multipliers seem to be lower for these, which could imply that level correction is now calculated and applied after the calculation of pDIF(min) and pDIF(max).

Anyone have a clue how to piece this together?

[Nameless]

Nice, that's just IMO the last thing that we need to test out. I will do some data sorting to see if I can find a trend.

[Nameless]

This is my initial finding from reading through the Page 2 data:

(tier I) 1.8 < cRatio < 2.2
Max pDIF = cRatio x 1.25
Min pDIF = cRatio x 1.25 - 0.9

The reason for 1.8 here is that anything below 1.8, you will have a lower pdif average than your cRatio, which I don't think is SE's intention since they wanted to reward players for having higher cRatio. So I simply did this ((X x 1.25) + (X x 1.25 - 0.9))/2 = X. Solve for X and you get 1.8.

(tier II) 1.2 < cRatio < 1.8

Max pDIF = cRatio x 1.25
Min pDIF = cRatio x 0.75

At this tier, your pDIF average will always be same as your cRatio

(tier III) ? < cRatio < 1.2

Max pDIF = cRatio x 0.75 +0.6
Min pDIF = cRatio x 0.75

nvm, 1.2 and the formula looks good.

Either way, I think the question is if there is a tier IV, and we need more data for tier III.

[sleekmotorwurkz]

2 Questions that came up in the random question thread, but they are relevant to this topic so I guess having it here would be beneficial...

Im almost 69SAM and one thing ive been noticing is that when I 2hr I miss ALOT. Curious if anyone has ever tested if the acc/atk bonus from YGK is active during 2hr? Common sense would say it still works, but in my (admittedly limited) experience, it doesnt seem like it.
I would test it myself but I dont know how to do the calculations.

... and pretty dumb question but honestly never bothered to find out: during meikyo, what does each WS's tp count as? I've assumed they were 300, 200, and 100 like your tp shows, but also been told they're all 300 by someone...

[shaddix]

Talkin about Rana, when i have zerk and minuet up, i can remember doing 700-800 with Hagun and now aprox 1200-1500 with peaks of 1650 on dmg:88. But when atk is below 500, it sux alot, even with that base dmg.

Deemos if you are feeling bored would you mind seeing what you get out of rana using an oni+2xminuet+zerk?

[WraithLakshmi]

The accuracy bonus is still there during 2hr, there's no reason why it wouldn't be, and I've personally never missed a ws during Meikyo Shisui.

As far as the tp return, that's another matter. I'm not personally sure, but it could go one of two ways, either 100 for each, or 300, 200, 100. I think Hagun still works during 2hr, unlike sekkanoki, so with hagun that would be either 200 for each, or 300, 300, 200.

[Enedin]

Sooo testing? This is great stuff!

[Brenhen]

I actually did a similar post on the KI forums regarding this subject. My argument was that Y/G/K ignores a certain amount of defense rather than an attack bonus. The math came out similar to yarkos. I got flamed for it tho.

[Yarko]

Sooo testing? This is great stuff!

Most of the stuff is solved, these are the final results summed up (ignore first page!):

- Some single hit WSs appear to have a multiplier added after cRatio, this can be confused with an attack bonus or defense negation if the monster has a level inferior or equal to 75, but the effects are lower if the monster is higher level and a level correction function is applied.

- The multipliers values appear to be:

Tachi: Gekko        2.0
Tachi: Yukikaze    ~1.5
Tachi: Kasha      ~1.6-1.7

Spinning Slash      1.5
Ground Strike       1.6~1.75

- The multiplier does not allow the final cRatio value to break "2.2 - Lv Correction", how this cap is exactly applied is not clear at this point.