Magic Accuracy and INT

[Tricen]

Oh right, the Resist! thing is there. I forgot about that since NONE OF THE RESIST TRAITS EVER WORK, FUCK.

[Wrish]

I found pchan's math decent this time, for a change. The assumption that bind (and sleep) either land fully, land half-duration, or resist fully is very reasonable - the use of [1 - sqrt(observed resist %)] is a welcome change from the mess of errors I used to see. My remaining concerns are mostly obvious:

Induction. It still remains unproven that the change of slope occurs at the same dINT for black-magic-enfeebling and elemental nukes. It could very well be that dSTAT suffers diminishing returns past +10 for nukes and +15 for enfeebles, or it could be a non-round figure like the annoying 10/11 multiplier we see for accuracy and H2H base damage.

Applicability. There is a major reason BLMs enjoy a mix of INT and raw skill. INT boosts their damage. For pure accuracy, we have no such incentive, and this helps us only with the tiny questions... the fringe cases such as Morrigan Slops versus Nashira Seraweels. (I'm presuming INT never affects potency of a black magic enfeeble.)

Verification. The number of trials, the recast times, the admission to only doing 500-1k trials a night (8-9 hours for 2 months straight), and the hint of evasiveness leave me quizzical. Can you upload some parse files or other media to show us you did the work? Repeatability is, after all, a requirement in scientific experiments. This test here contradicts a previous test by concluding with a different dINT, so scrutiny and reconciliation are the topics of the day. That includes re-analyzing the INT of EM Hpemdes (and those elementals in the previous test), since that is critical to the offending conclusion.

Interpretation. Tests 6, 7, and 8 suggest mostly that M.Acc scales half as fast with landing rate at low accuracy. The previous test did not really analyze low accuracy situations, though this test scratches the surface. Did we learn something new?

[Yugl]

or other media to show us you did the work?

Alright guys, lets settle in for over 10hrs of half-nakedchan spamming bind.

[BerenTebogo]

I skimmed that quick and I saw 10 hours of half naked pchan and my eyes glazed over for a minute. Damn you Yugl.

[Argettio]

There seems to be 4 types of people posting in this thread:
* Pchan - Posting his figures, unwilling to discuss it. It's fact

* Haters - Those that want to call him a geek for spending the time doing (these people need to GFTO)

* Stupid people - People claiming that SE made magic accuracy Mob specific, while melee accuracy, damage, enmity etc are all consistent across the game (these people need to shut up as they are pissing off pchan thus screwing the next group)

* Sensible people - Kaeko etc who are trying to discuss the data and some implications

I wrote this back around the time the first big tests were done on m.acc and elemental magic, it is by no means perfect, but it explains the possible implications of these findings.

Basically most 'old school' RDMs have far too much skill in there builds and not enough INT or MND, moving the turning point to dINT +15 only increases the value of INT.

Sadly FFXIGEAR is dead, so the links are broken.

[pchan]

It's not like you have these little shit storms with 2-3 of the same people, it's everywhere you post BG, Alla, etc.).

I only post on BG, go away you nerd.

I do find pchan's tests interesting, although as pointed out, the method of procurement warrants further explanation. I've never seen pchan falsify data, so ima go /w the innocent until proven guilty thing here, but I would like to hear from him how the data was obtained. Also as Failure tested, it's possible the INT rating on the dolphins might be off?

My post #34 of this thread is self explanitory, I don't need to prove anything. With 500-1k overnight, and another 500 If I keep the test going the whole day, you easly get 5-7,000 per week. I started the tests before Chirstmas and ended them a few days ago.

You can check the current equipement on FFXIAH, from Lynnminmay on Remora :

http://i189.photobucket.com/albums/z...nn/testmac.jpg


This is the gear I used to test it on PUP/RDM

- vermi to keep MP refreshed
- PUP/RDM with merit is 142 enfeebling skill
-weapon is lady bell, it has 1 DMG and is used to hit mobs for zero
-errant cuffs, merman's earring with +1 int augment, snow ring,rse feet,
-altruistic cape
-augmented earring to mac+2

this gear gives : 142 skill, 8 mac, 81 INT.


If you want I can run another test over night, and post the parse .. but I'm sure the idiots will still find an excuse.

I contest out of wanting clarity. Not to make you look bad. If I wanted to make you look bad I wouldn't have worded it in such a nice fashion. Saying "I don't care what you think" to what I feel are valid questions is just hurting your supposed cause.

If you wish to ignore my questions that is certainly your choice; however, I (and I feel others) will look at anything you post with a certain caution. Again, you will probably say you don't give a shit. I have my own thoughts, you have yours. So if you don't wish to engage in some sort of meaningful discussion and hide your methodology I won't interfere in this thread. People are smart enough to make their own judgments.

Really ? I read your first post, and all I see is "your number are too perfect", and other attempts at contesting the possibility to get such a sample size. Now you got your answer. Did you ask lodeguy how it is possible to get a 7k sample size on ice elemental ? Because I'm 100% certain that holding a tough mob while "spamming" spells on it can't be faster than 200 / hour. When I personally did my tests on qiqirn poulterers (much faster since they don't melee), it was 300/ hour "only". So a sample size of 7,000 is at least 35 hours on those elementals, and probably more. And he obviously doesn't do it while AFK.

[Argettio]

Really ? I read your first post, and all I see is "your number are too perfect", and other attempts at contesting the possibility to get such a sample size. Now you got your answer.

You read it as contesting, but I read it as genuine interest in the methodology.

He did the same when first batch of testing done in this field (about a year ago), your not getting special treatment.

We both know, the burden off proof is on you. If your theory is sound then it will withstand people (rightfully) scrutinising it, so stop being obtuse and let them do go over your numbers/methods.

You can't just post things like this and expect everyone to "oh... ok, nice find", that's not science works, and it is certainly not how the FFXI community work.

[pchan]

You read it as contesting, but I read it as genuine interest in the methodology.

look

I know I sound like a total asshole here because I'm essentially accusing you of faking data, but ...

[Argettio]

look

Yes, because you won't release your methodology or give any extra details

Give him what he wants, let him go over your numbers, let he verify your findings.

Why is that so hard?

You are acting like some one who has something to hide and that never looks good.

[pchan]

Yes, because you won't release your methodology or give any extra details

I did. In the OP.

[Argettio]

I did. In the OP.

You gave some very brief details in the OP, people have asked for more information, more clarification on the specific methods of data collection.

You seem to be cutting your nose of to spite your face; you may have found something interesting, but your not letting people question/validate/discuss it.

But I can't be assed to argue this any more;
Either accept that no one is going to adopt your theory or start engaging in a meaningful discussion.

[pchan]

You gave some very brief details in the OP, people have asked for more information, more clarification on the specific methods of data collection.

You seem to be cutting your nose of to spite your face; you may have found something interesting, but your not letting people question/validate/discuss it.

But I can't be assed to argue this any more;
Either accept that no one is going to adopt your theory or start engaging in a meaningful discussion.

I didn't see you posting anything interesting. If you don't want to discuss, don't post. Simple. I was asked for clarification yesterday at 1 AM... give me a break, it's 13 now and I gave more clarification, which is different from not revealing my testing method (=what you say) which is clearly stated on page 1. Troll less.

[Tricen]

YOU'RE 13.



















Faggot.

[Argettio]

I didn't see you posting anything interesting. If you don't want to discuss, don't post. Simple. I was asked for clarification yesterday at 1 AM... give me a break, it's 13 now and I gave more clarification, which is different from not revealing my testing method (=what you say) which is clearly stated on page 1. Troll less.

I am trying to get you to revel the information that people are asking for, that isn't discussing, because discussion can not take place until you are willing to answer peoples questions.

You have posted 4 times in this thread today, not one of them has tried to encourage positive discussion, most have been simple attempts to belittle or disregard those that have raised any doubt about your findings.

[xsack]

Klutz you should change your name to Cunt cause all you do is cry like a bitch next time i want your opinion i'll beat it out of you!

[Byrthnoth]

I'd just like to point out that it'd be harder to land debuffs on HNMs if they had low MND and got the rest of their magical evasion from the mythical magical evasion trait than if they had high MND. However, due to us generally not knowing INT or MND values of HNMs it's a little hard to deal with this in reference to them.

Pchan, admit if you used a bot or not so people stop bothering you.

[Klutz]

Klutz you should change your name to Cunt cause all you do is cry like a bitch next time i want your opinion i'll beat it out of you!

Dude your internet G-ness is so cool I wish I was just like you.

I see you rollin', and I'm hatin, etc.

[pchan]

I found pchan's math decent this time, for a change. The assumption that bind (and sleep) either land fully, land half-duration, or resist fully is very reasonable - the use of [1 - sqrt(observed resist %)] is a welcome change from the mess of errors I used to see.

Possible, but then again the previous data had not a lot of sample size, and had incoherence due to this (lodeguy's data showed 3% increase between dint=10 and dINT=20, when you would expect 5% with previous theory and 7.5% with this thread). Whatever you do you are limited to your sample size, having a 1% error doesn't mean your resust is accurate, it means that if 20 dudes do the same test as you, 19 "should" find a value in the same interval as you +/- 1%. This also applies to my tests too obviously.

That includes re-analyzing the INT of EM Hpemdes (and those elementals in the previous test), since that is critical to the offending conclusion.

I've already double checked on last page with blizzard I and stone I, and rechecked earlier with blizzard III. They do have 58 INT. So unless bind is not checked against target's INT (why not ? why yes ?) ... I can't explain the difference. Would SE be evil enough to check bind against CHR ? w. That would be .. interesting.

[Taruking]

* Stupid people - People claiming that SE made magic accuracy Mob specific, while melee accuracy, damage, enmity etc are all consistent across the game (these people need to shut up as they are pissing off pchan thus screwing the next group)

They are not always consistent across the game, for example there are several NMs that don't follow the usual enmity rules (Chigre for example) and we know that because the result is obvious. None of the formulas for melee accuracy/damage, magic accuracy etc are derived from tests specifically run on HNMs. The current formula for melee acc is still somewhat useful because ppl use them to improve performance in merit party. For HNMs melees either pile accuracy or take advantage of BRD/COR's buffs/Feint so there is no need to determine wheter SE have implemented a different set of melee accuracy formula for HNMs. This does not apply to magic accuracy. We do know certain HNMs are highly resistant to enfeebling magic but we don't really understand the mechanism behind.

The biggest problem with magic accuracy compared to physical damage accuracy is we can't check mob for the information we need, which makes it difficult to determine the value and the combined effect of magic accuracy related spells/JA (Death Blossom, stutter step etc). To run the test on HNMs are made difficult by the fact that BLMs and RDMs can only cast limited numbers of spells in HNM battles and it may not create statistically significant sample.

[Argettio]

They are not always consistent across the game, for example there are several NMs that don't follow the usual enmity rules (Chigre for example) and we know that because the result is obvious. None of the formulas for melee accuracy/damage, magic accuracy etc are derived from tests specifically run on HNMs. The current formula for melee acc is still somewhat useful because ppl use them to improve performance in merit party. For HNMs melees either pile accuracy or take advantage of BRD/COR's buffs/Feint so there is no need to determine wheter SE have implemented a different set of melee accuracy formula for HNMs. This does not apply to magic accuracy. We do know certain HNMs are highly resistant to enfeebling magic but we don't really understand the mechanism behind.

The biggest problem with magic accuracy compared to physical damage accuracy is we can't check mob for the information we need, which makes it difficult to determine the value and the combined effect of magic accuracy related spells/JA (Death Blossom, stutter step etc). To run the test on HNMs are made difficult by the fact that BLMs and RDMs can only cast limited numbers of spells in HNM battles and it may not create statistically significant sample.

There is a difference between unknown evasion and different accuracy rules.

The only thing that is know to vary in NMs vs Normal mobs is enmity where some NMs (1-2) have different hate systems (read: accelerated decay) and some NMs have hate reseting moves.

Most NMs/HNMs are just Wild Rabbits on steriods as far as the game mechanics is concerned.