PSA regarding Cuchulain's Mantle

[Cream Soda]

Whether you see another parse or not has ZERO to do with calling that parse a legitimate test. There are WAY too many variables. For all you know, parse B could have had more flashes from puks and more warmups from mamools, where parse A could have had more knock backs from puks and more fire spit from mamools.

[shaddix]

Whether you see another parse or not has ZERO to do with calling that parse a legitimate test. There are WAY too many variables. For all you know, parse B could have had more flashes from puks and more warmups from mamools, where parse A could have had more knock backs from puks and more fire spit from mamools.

Both PTs were with one other extremely well-geared and merited relic DD, a very well-geared and merited non-relic DD, Minuetx2/Marchx2, and Mamool staging point, with exp/h valued at just over 32k/h over about 1.5-3 hours of exp. Killing about 180-190 mobs per hour.

Why would it be so terrible for cuchulain's to not be as amazing as this thread has previously made it out to be

[Cream Soda]

What does your quote have to do with what you quoted from me? It looks totally unrelated. Please explain the relation.

[Araelus]

Heh, looks like after getting new pDIF numbers I'mm have to fight 1,000 identical mobs that don't give blind/flash/etc. to put this whole "how much +% does 1 acc give?" issue to rest.

Re-doing some of my charts for el_tio, right now, though.

[shaddix]

Mobs dying too fast to use tp moves is what I was implying. Every 19 seconds a mob dying, and a second or two of that is spent /autotarget time. I am trying to emphasize that the parse has more credibility than it is being ascribed.

But, that is not the point! The point is there's no parse denying this one, just words!

Heh, looks like after getting new pDIF numbers I'mm have to fight 1,000 identical mobs that don't give blind/flash/etc. to put this whole "how much +% does 1 acc give?" issue to rest.

Just the man i'm looking for. VT onry

If I was on your server I would be helping you out all day

[Cream Soda]

It's still not a credible parse. Now, if he used direct parse, which does a mob breakdown
110 mobs total summary

Then
Mamool Ja Lurker x Number of fights
[insert parse for all lurkcers here]

Sea Puk x Number of fights.
[insert parse for all sea puks here]

and came up with the same results for each type of mob, it would be a much more credible source and would also be irrelivant to the number of each mob fought, but since it's a summary of all the mobs, then the number of each type of mob fought IS relevant

[Mayareira]

Anyway, yea, Cuchulain's Mantle is good, I don't deny it one bit, but it is a minor upgrade for a pretty high price, and unless your gear for all your jobs is perfect elsewhere, there are a lot better things you can buy.

There is point/line for every jobs where you have to start investing a lot for less impressive returns. For mages, that'd be post HQ elemental staves, required torques, necessary hMP set, etc. Prism Cape and HQ staves cost the same but we all know which one a mage should prioritize.

After that point, Cuchulain's Mantle is one of the best upgrades a melee or a versatile player can do. Comparing an upgrade from Chivalrious Chain > PCC to Amemet > Cuchulain's is a weak argument simply because it is an example to the lack of priority sense.

A better example would be an upgrade from Errant Houppelande > Mahatma Houppelande vs. Amemet > Cuchulain's for a versatile player that favors his melee jobs more. I personally don't have a Cuchulain's Mantle for obvious monetary reasons but I wouldn't blame a melee investing in equipment like HQ acc. rings and Cuchulain's mantle.

[FailureMidgard]

There is point/line for every jobs where you have to start investing a lot for less impressive returns. For mages, that'd be post HQ elemental staves, required torques, necessary hMP set, etc.

After that point, Cuchulain's Mantle is one of the best upgrades a melee or a versatile player can do. Comparing an upgrade from Chivalrious Chain > PCC to Amemet > Cuchulain's is a weak argument simply because it is an example to the lack of priority sense.

A better example would be an upgrade from Errant Houppelande > Mahatma Houppelande vs. Amemet > Cuchulain's for a versatile player that favors his melee jobs more. I personally don't have a Cuchulain's Mantle for obvious monetary reasons but I wouldn't blame a melee investing in equipment like HQ acc. rings and Cuchulain's mantle.

I wouldn't blame someone either, unless they are a red mage without any HQ staves like you said. My original (and main point), was I wouldn't blame a melee for not having a Cuchulain's. As for the Chiv>PCC thing, it's easier to make a comparison inside a job because I can offer direct proof by numbers that one is better. But I dunno, you can make a lot of comparisons, and really, none are more valid than any others when compared abstractly here. It's based on the person and their current gear. Like I said, if you have nothing else to spend money on, go for it, it is a solid upgrade for sure. I want one myself, but can't justify the gil at the moment because I have non-melee jobs that need stuff.

Also to Shaddix, 19 seconds is nowhere near too fast for a mob to use it's TP. Have you never merited?

[Sylvrdragon]

My personal reason for loving this thread is that it MIGHT just entice people to actually try to take La Vaule on Saturday Nights. I suggest it every week in LS but never have a single taker. As such, while this thread may indeed cause the price to raise for a bit, it will ultimately lower it in the long run (in theory) to the point where Cuch mantles are as common as Foragers.

[shaddix]

It's still not a credible parse. Now, if he used direct parse, which does a mob breakdown
110 mobs total summary

Then
Mamool Ja Lurker x Number of fights
[insert parse for all lurkcers here]

Sea Puk x Number of fights.
[insert parse for all sea puks here]

and came up with the same results for each type of mob, it would be a much more credible source and would also be irrelevant to the number of each mob fought, but since it's a summary of all the mobs, then the number of each type of mob fought IS relevant

That is directparse I think, I will see if i can get him on aim or something soon and maybe he can provide more specifics assuming he still has the parse.

Also to Shaddix, 19 seconds is nowhere near too fast for a mob to use it's TP. Have you never merited?

Yeah I have merit lol, I have been in a party or two where mobs melt like that, and from what I recall they often die without ever using a tp move, or they die right as they are using one. But the point was that there's no time for the DD to tickle the mobs tp up to where they can use that flash move or if they do its right as they are going down from a 1200dmg ws.

[TybudX]

Then I'd do the BC, sell the mantle, and buy more useful things. Don't get me started on the whole "if you farm it/get the drop, then the monetary value is irrelevant". That is the stupidest argument anyone can ever make.

That's the argument you're making yourself, right now. How is selling Cuchulain's to buy other stuff any different from selling other stuff to buy Cuchulain's? Ok, you value all the jobs you have leveled more than the one melee job you say you have, that's fine. Your entire argument that Cuchulain's isn't worth the cost is based on it being costly, and that you could use that gil to buy 'more useful things'... well, guess what? Cuchulain's Mantle is that 'more useful thing' for me, and for a lot of other people. My one melee job is looking down the road of two HQs and a relic. That 'minor upgrade' is bargain basement at 3.5 mil compared to what little else I can change.

[Cream Soda]

Also, Shad, you say we don't have a parse, but as far as the credibility of 7 different mob types, the variety of TP moves, ect, you don't have one either (or at least a credible one)

[Suiram]

At some point in the Spring of 2007, in the course of testing things related to crit builds and sea gorgets, Nagamaki provided a fairly conclusive test demonstrating the now widely-accepted understanding about melee accuracy (Acc=Eva against EM is 75%, every change in 2 Accuracy changes 1%, every change in level difference changes 2%). I think the post itself got lost in some sort of forum issue. It looks like nobody embroiled in this current debate was active on these forums in April 2007, so maybe that's why nobody remembers it.

Here's the first result I found in google search trying to see if I could locate the post. Accuracy <-> Hit Rate relationship found! It just summarizes what Nagamaki's test said. I went to the thread referenced, but the post is lost as I remembered.

Thread that previously contained the lost accuracy tests: http://www.bluegartr.com/forum/44084...iscussion.html
Discussion of the forum rollback and another summary of the lost discussion: http://www.bluegartr.com/forum/45252...st-thread.html

I hope this is sufficient in the absence of the actual data.

[shaddix]

Also, Shad, you say we don't have a parse, but as far as the credibility of 7 different mob types, the variety of TP moves, ect, you don't have one either (or at least a credible one)

Well this is just a difference of opinion then, and I am not one to under-appreciate chaos and variance, I do understand that it is possible those parses could be incorrect. But if I have no information, some information is better than none, and we make decisions based on the information we have, whether it is complete or not.

I have no anecdotal evidence that 1acc=.5hitrate, and it is likely you do, and surely you give at least some credit to ruke's parses. So we both must wait for ruke to come back, or wait for someone to dig something up out of bookmarks, or wait our friend Araelus to do some testings.

Suiram: ty sir


srsly bedtime now X_X I will try to test this out myself when I get a chance as well, birds if that's okay with you guys.

[Araelus]

Why would you go for ATT instead of ACC on higher DEF mobs? Don't those mobs tend to have higher EVAs as well, which may require even more ACC?

Earlier in that same post, however, I pointed out the higher level monsters have more evasion, which would change the player's starting accuracy. A lower starting accuracy increases the value of +X accuracy.

Holding player ATT and ACC constant, though, +X Attack becomes more useful as Def or MobLV increase. We can see that, comparing Chart 4 and Chart 5, say the player has 500att and 85% acc.

On monsters with less defense and of lower level (Chart 5) the pink square for 85% acc is farther right than the blue diamond for 500att. Thus the player would want to add +7acc and not +15att.

On a higher level, higher DEF monster (Chart 4) the situation is reversed.


Then again, we turn to the problem I overlooked with this math:

I wish people would stop acting as if accuracy doesn't give TP, or that TP doesn't equate to damage.

Even people who are normally spot on.

It kind of invalidates most of the numbers in this thread so far (altho it mostly furthers the conclusion of Cuchulain > atk backpiece).

Exactly. That was a poor oversight on my part.

Let's go back and add WS into the mix! We expect to see the return from ATT decrease and the return from ACC increase. Let me do that real quick, using the example FM provides:

Samurai

• 60% damage from WS.
• TPing Attack+ affects 40% of damage, not 100%, by how much it raises pDIF(avg)
• TPing Acc+x increases damage by [(0.5*x)/baseACC]*.9 based on the 10% of damage coming directly from Meditate

Pink squares: Base Accuracy %
Blue diamonds Base Attack vs. Greater Colibri
Yellow Triangles: Base Attack vs. lv85 500def mob
Teal Circles: Base Attack vs. lv92 700def mob

Spoiler: show

http://www.bluegartr.com/forum/attac...1&d=1239002773

Shaddix, sourcing from some of Ruke's posts (and some charts I've seen him make here on BG) comments that for 2-handed weapons the returns to accuracy are not linear approaching 95% hitrate.

Suppose this occurs above 90% hitrate. That would push those pink squares labeled higher than 90% left, decreasing the utility of +1 acc when over 90% base Accuracy.

But since this would not change the utility of +1 acc at 90% accuracy, it should still be clear on merit mobs players should have at least that much hitrate before even getting to 200 attack.

Here's four basic AH-melees from my last G.Colbri merit party:

          Overall Dmg%	Melee %	WS %	M.Acc%
SAM(Soboro)	26.19%	40.13%	59.87%	88.32%
SAM(Hagun1)	22.16%	47.65%	52.35%	78.57%
WAR(Perdu)	19.48%	57.26%	42.74%	88.67%
SAM(Hagun2)	28.25%	35.45%	64.55%	88.11%

I am sure those melee had over 200 attack. So... they need better gear in general, but most importantly more Accuracy.

My personal reason for loving this thread is that it MIGHT just entice people to actually try to take La Vaule on Saturday Nights. I suggest it every week in LS but never have a single taker. As such, while this thread may indeed cause the price to raise for a bit, it will ultimately lower it in the long run (in theory) to the point where Cuch mantles are as common as Foragers.

I'd really like to see this happen - be even better if they removed some of the crap drops from Shell/Wing and just threw Cuch Mantle in all three Stronghold Ops.

[FailureMidgard]

That's the argument you're making yourself, right now. How is selling Cuchulain's to buy other stuff any different from selling other stuff to buy Cuchulain's? Ok, you value all the jobs you have leveled more than the one melee job you say you have, that's fine. Your entire argument that Cuchulain's isn't worth the cost is based on it being costly, and that you could use that gil to buy 'more useful things'... well, guess what? Cuchulain's Mantle is that 'more useful thing' for me, and for a lot of other people. My one melee job is looking down the road of two HQs and a relic. That 'minor upgrade' is bargain basement at 3.5 mil compared to what little else I can change.

Then there's nothing wrong with you buying one...which my first post clearly stated "if you have nothing else to buy.." Why are you arguing with me if nothing is in contradiction with what I have said?

And thanks Suiram for posting that, I'll never understand where all these internet conspiracy people come from...

[shaddix]

I'll never understand where all these people with facts and parses come from...

fyp :3

[Araelus]

Shaddix, in all fairness, "a little knowledge is a dangerous thing". That one parse from Mamools flies in the face what I understand to be accepted knowledge, and it does so on shaky grounds because we cannot see what went on to form those ACC% values.

That said, I am willing to test the accuracy thing in the future, but I have a lot of other projects going on right now. I am sure that either I, or you, or someone will sit down and parse it out.

For now I have to agree with FM. I am very uncomfortable treating the bonus from acc+ becoming asymptotic as one nears 95% based on one parse result that was designed to compare SAM/WAR and SAM/DRG overall, not to return data on the accuracy question at hand. However, I am not dogmatic about 1acc = 0.5% hitrate. If the data shows this is not a constant rate, then I accept what the data shows. Right now I am sticking with the normally accepted equation because it is the current inference to best explanation.

By the way, thank you for the links, Suiram.

[FailureMidgard]

Did you also notice that the difference between /war and /drg (after berserk/etc) on his melee hits was 8 total damage?

And I think we're all willing to accept 2acc:1hit MAY be wrong, but until someone PROVES it, it's retarded as fuck to think what you're saying.

BTW, the difference between Sam Polearm and GK skill is 240->276 (36 skill), or 36*.9 or ~32 accuracy. Do you really think a Polearm Sam parses only 6% lower on accuracy than a GK Sam? I can guarantee you that is not the case.

Anyway, if someone wants to prove it, then they should be attacking mobs with a near static evasion (and a 1 level difference at most really), and do enough tests to actually conclude it. If I get bored enough, I may actually go a test that you would find conclusive, but nobody else in their right mind would.

Shaddix, in all fairness, "a little knowledge is a dangerous thing". That one parse from Mamools flies in the face of accepted knowledge, and it does so on shaky grounds.

That said, I am willing to test the accuracy thing in the future, but I have a lot of other projects going on right now. I am sure that either I, or you, or someone will sit down and parse it out.

Right now I have to agree with FM; I am very uncomfortable making an assumption about acc+ becoming asymptotic as one nears 95% based on one parse result that was not designed to return data on the question at hand.

And I'll say another thank you to Suiram for those links.

Just quoting these 2 posts for Shaddix, since he seems to have comprehension issues. Only quoting Araelus' post to hopefully let Shaddix see that he is being unreasonable with his "proof".

Like I said before, we are all willing to believe the 2acc=1hit may be wrong. Nobody is saying it's set in stone, but it has been parsed and tested in a decently conclusive manner in the past, and 1 parse with no controlled variables is not proof that it is wrong. Furthermore, in no way is it even enough evidence to say that the burden is with the first assumption. Right now, the burden is with anyone saying it is other than 2acc=1hit. I may even go do some fast parses tomorrow if I'm bored enough, but you need to drop your internet conspiracy crap (BTW, linking 1 parse with no controlled variable is not proof as I said...and saying that none of us linked a parse, therefor it wasn't true, is definitely an "internet conspiracy" attitude).

Also, like I said in my first post, even if you want to use Ruke's parse for the accuracy, use it for the attack values also. It is completely realistic for solid DD to be capped on attack. Part of Ruke's premise for /drg beating out /war was that berserk did not up his melee damage in any noticeable fashion. He was attack capped. So going back to the initial attack versus accuracy debate...you didn't really make any ground there.

Finally, like I said in that post, I think any Samurai can tell you that the have more than a 5-6% difference in their accuracy between GK and Polearm, which is what your numbers are implying would happen (unless, in fact the A+/etc rating actually changes the accuracy curve).

Again, not saying 2acc:1hit is set in stone, but your linked parse proves nothing at all, and there has been data collection done to support these values that are widely assumed.

[Edit: OH YEA. I kinda forgot, but, a very very possible explanation for Ruke's parse, is he was sitting at 93% accuracy base...anything under that was due to various things that cause misses, and when he switched to /drg, he got capped at 95% for the majority of the parse, which only saw a ~2% increase in the actual parse. Basically, his /drg parse was the effective accuracy cap for that camp, due to shadows/flashes/etc.]

[Tsuko_Asura]

I admit to skimming, but I don't understand what's so difficult to grasp about the fact that MJSP has, what.. Six, seven different types of targets? Different types of targets with different, variable behaviors that can and will skew an accuracy parse.. I mean, really, you shouldn't be arguing with people who have proven their ability to manage such things if you can't handle such a blatantly simple concept.