Magic Accuracy - Why so serious?

[Spekkio]

Because melee skill/accuracy isn't a debate, it's a collection of facts. The dilemma with magic isn't in determining which of two known bonuses is better for a given situation, but rather in trying to make such a comparison when the exact effects are unclear. And it's really not reasonable, in my opinion, to use a melee analogy as a starting point, given that those calculations regarding other aspects of magic which are already understood are more often than not fundamentally different from their melee counterparts, and the limited data available regarding magic accuracy would suggest that it may be as well.

we need to understand the underlying mechanics of the system before we can make any statements of "a is better than b" universally. there are times to eat meat and there are times to eat fish. personally my favorite times are the times to eat pie, but that's me. regardless, as suriam says, we need to be able to quantify the effect before we can attempt to define when magic accuracy will outperform skill or stats, how much of it is enough, etc.

[pchan]

The only similarities I noticed with melee accuracy is that Magic accuracy depends on base skill. Those were some quick tests but at low skill (240) magic accuracy was less usefull than skill (like 1 mac=0.5 skill or something like that), at 280 skill they had similar effects and at very high skill magic accuracy is superior (1.2 1.3 skill ).

edit : nvm about total mac.

[Arkanna]

This topic has been discussed with my ls mates a good bit lately. One of our BLMs got an Alkalurops from Tinnin and she's obviously enthusiastic about it, but the rest of us are for the most part skeptics.

Maybe a good test to compare accuracy would be, find the highest-level Hpemde's you can and Blind them 300 times (100 naked, 100 with Pluto's, and 100 with Alkalurops). I've killed them solo on BLM many times and they don't resist enfeebles much at all, but maybe being naked would change that. If not, I have a 66 RDM I could try naked vs. Pluto's tests with at least.

300 casts may not be a large enough sample set to get definitive data from, but it's a start. I'll try to start on this as I can (lots of in-game events and RL stuff lately, and assuming she'd even let me touch the staff at all) unless someone points out why it's a bad test.

[pendragoon]

I played with an Alkalurops that I received from ZNM.

I didn't do any large scale testing... but I did the best thing I could think of... Try to sleep VT Imps on RDM. (Since previous to testing, I failed to sleep imps about 50% of the time after getting aggro on my way to Nashmau.)

Alkalurops vs Test Pluto's. Alkalurops was tested first, just so that if any type of resists built up, it would be more active on the Pluto's. (I figured pluto's was the beter before the testing)

- Alkalurops landed about 50% of my 50 sleeps.

- Pluto's in testing landed about 90% of my 50 sleeps (I swear this was higher than my previous run-ins with imps)

This was all done in the best enfeeble gear I had (not the best), against a single imp (didn't have seperate imps for testing to make sure there were no level changes).

Af2 head,
enf. torque
snow rings
sea cape
mahatma hands/legs
golliard feet
phtm earrings

Lets just say I sold it for about 10x more than I thought it would be worth (7mil). The only reason I'd pay more than a single HQ staff price is so that I could resell it again.

[Haroun]

based my above statement on the tests referenced by this individual. perhaps he can post a link?

I was mainly referring to these two threads:

http://killingifrit.com/forums/index...ic=171164&st=0
http://ffxi.allakhazam.com/db/jobs.h...3;num=7;page=1

I've seen similar in non-BRD related posts, but it'd take some searching for me to find links. From what I've seen, every person claiming to have actually tested Alkalurops vs HQ Staves (as opposed to speculation) has come to the conclusion that HQ Staves have a better success rate. It's a small sample, but I have yet to see any evidence to the contrary.

[Virtuous Saint]

I played with an Alkalurops that I received from ZNM.

I didn't do any large scale testing... but I did the best thing I could think of... Try to sleep VT Imps on RDM. (Since previous to testing, I failed to sleep imps about 50% of the time after getting aggro on my way to Nashmau.)

Alkalurops vs Test Pluto's. Alkalurops was tested first, just so that if any type of resists built up, it would be more active on the Pluto's. (I figured pluto's was the beter before the testing)

- Alkalurops landed about 50% of my 50 sleeps.

- Pluto's in testing landed about 90% of my 50 sleeps (I swear this was higher than my previous run-ins with imps)

This was all done in the best enfeeble gear I had (not the best), against a single imp (didn't have seperate imps for testing to make sure there were no level changes).

Af2 head,
enf. torque
snow rings
sea cape
mahatma hands/legs
golliard feet
phtm earrings

Lets just say I sold it for about 10x more than I thought it would be worth (7mil). The only reason I'd pay more than a single HQ staff price is so that I could resell it again.

That is the smallest test I have ever seen, if anyone believes this is acc account of alkalurops vs HQ staff you are smoking crack.

[Haroun]

That is the smallest test I have ever seen, if anyone believes this is acc account of alkalurops vs HQ staff you are smoking crack.

Do you know of any actual tests of Alkalurops vs HQ Staves that have shown results in favor of Alkalurops? I have yet to see one, with or without supporting data.

[Vail]

Do you know of any actual tests of Alkalurops vs HQ Staves that have shown results in favor of Alkalurops? I have yet to see one, with or without supporting data.

cause all you see are unbelievably small sample sizes using only a spell that more often then not lands fully potent (full duration) or out right resist (sleep/sleepII) seeing a test done with spells that can actually cover the different resist ranges on a more regular biases would give considerably more accurate results.

[pchan]

I

- Alkalurops landed about 50% of my 50 sleeps.

- Pluto's in testing landed about 90% of my 50 sleeps (I swear this was higher than my previous run-ins with imps)

I don't belive it's possible to get 90% accuracy on imps with sleep. I mean that would mean the bonus of HQ staff is 25 magic accuracy + 40% no-resist rate. I ve never seen this kind of increase from HQ staff. And from my experience imps resist about 50% of the time with normal gear (inc. HQ staff).

[Arkanna]

This topic has been discussed with my ls mates a good bit lately. One of our BLMs got an Alkalurops from Tinnin and she's obviously enthusiastic about it, but the rest of us are for the most part skeptics.

Maybe a good test to compare accuracy would be, find the highest-level Hpemde's you can and Blind them 300 times (100 naked, 100 with Pluto's, and 100 with Alkalurops). I've killed them solo on BLM many times and they don't resist enfeebles much at all, but maybe being naked would change that. If not, I have a 66 RDM I could try naked vs. Pluto's tests with at least.

300 casts may not be a large enough sample set to get definitive data from, but it's a start. I'll try to start on this as I can (lots of in-game events and RL stuff lately, and assuming she'd even let me touch the staff at all) unless someone points out why it's a bad test.

Unfortunately this test didn't work out so well. The owner of the Alkalurops went out to Sea and tried it as I described, but apparently the Hpemde's were too easy to enfeeble, even naked.

[Haroun]

cause all you see are unbelievably small sample sizes using only a spell that more often then not lands fully potent (full duration) or out right resist (sleep/sleepII) seeing a test done with spells that can actually cover the different resist ranges on a more regular biases would give considerably more accurate results.

We may not know to what extent, but it's reasonable to assume that HQ Staves have a lower resist rate than Alkalurops when everyone that has Alkalurops agrees that HQ Staves have a lower resist rate.

[Virtuous Saint]

We may not know to what extent, but it's reasonable to assume that HQ Staves have a lower resist rate than Alkalurops when everyone that has Alkalurops agrees that HQ Staves have a lower resist rate.

No that is untrue, and based on jp tests *which actually have some good tests and results, and a lot of people sharing info* alkalurops is best possible for BRD and enfeebling based magic. Obviously HQ staves are best for nuking as they add 15% dmg along with a what (could be) slightly lower acc of landing the spell.

[Haroun]

No that is untrue, and based on jp tests *which actually have some good tests and results, and a lot of people sharing info* alkalurops is best possible for BRD and enfeebling based magic. Obviously HQ staves are best for nuking as they add 15% dmg along with a what (could be) slightly lower acc of landing the spell.

Do you have a link to these tests? If I'm wrong I'm wrong, but I'd like to see them.

[Spekkio]

it's entirely possible you're comparing apples to oranges here too. without a proper understanding of macc, caps on magic accuracy, and how the HQ staves interact we can't necessarily make generalizations about a is strictly better than b. consider a weapon w/ +25 MAB. in the hands of a corsair, this would be far superior to an ele staff for QD (assuming they could use it ofc, but this is about the math, not a specific example) since it would be close to 20% or so of a boost in damage when the other MAB were taken into account. meanwhile, it was established a while ago in a thread revolving around the dorje that +25 MAB lived in a zone between NQ and HQ staves for damage. i would too be interested to see the testing data, but i don't think we can draw any wide reaching conclusions until the test space is expanded to show that one consistently outperforms the other.

[Haroun]

While on the subject of JP tests, I'd like to add: http://wiki.ffo.jp/html/805.html. Which suggests that the hidden +M.Acc on HQ Staves would be roughly +25-30.

[rrynn]

might be fun to play with for slow II/para II in situations where you are not using eseal but the mob is still somewhat resistant. you loose the mnd from your shield slot, but the gain in macc would probably result in the mob spending more time actually slowed/paraed.

[Sylvrdragon]

Have a 75 BRD/WHM go out and use WHM enfeebs on Goldfish. They should have a significantly higher resist rate with gimped Enfeebling skill, but not outright.

Also, 100 casts is nowhere near enough. 1000 would be good, but not very practical for a single person. Maybe if we could pool together several naked BRDs to all go out and collect data, much like the TH Tests currently undergoing, we could get a respectable sample size. It will be a bit more complicated than the TH tests as you would have to keep track of base stats (i.e. Race), and merits, whereas the only known modifiers for TH are +1 items or Tier.

Have them take some Enfeebling gear, stat gear, HQ and NQ staves, and some lesser M.Acc gear. Might as well test everything we can. Actually, someone without the Alkalurops can do the HQ vs NQ vs Skill vs lower M.Acc. Come to think of it, if we can pin down M.Acc, then we don't need anyone with Alkalurops at all, so long as we know the basic principles on which the formulas are based.

Edit to add: Actually any job with no natural Enfeebling skill can do the underlying tests (so long as base stats are stated). BRD or SMN only necessary for the Alkalurops test specifically.

[pchan]

While on the subject of JP tests, I'd like to add: http://wiki.ffo.jp/html/805.html. Which suggests that the hidden +M.Acc on HQ Staves would be roughly +25-30.

That would not surprise me knowing SE. Look @ djorge. It's a pretty good weapon except if you are a pretty good BLM. It has slightly less damage boost, and no accuracy. Alkalurops has no damage boost (compared to staff) and .. slightly less accuracy ?

[Spekkio]

That would not surprise me knowing SE. Look @ djorge. It's a pretty good weapon except if you are a pretty good BLM. It has slightly less damage boost, and no accuracy. Alkalurops has no damage boost (compared to staff) and .. slightly less accuracy ?

it's the astral signa rule. anything that's broader is almost exclusively not as powerful.

[Starr]

Do you know of any actual tests of Alkalurops vs HQ Staves that have shown results in favor of Alkalurops? I have yet to see one, with or without supporting data.

Ok. First, honestly, I don't know what people mean when they say +15% magic accuracy. I do believe that there is a hidden magic accuracy stat, and that magic skill, stats are converted into it but there is no way to measure it so idk what 15% mac means. What is easy to test however is the % increase from non resisted spells. Sleep 100 times w/o staff, 100 times with staff and observe.

Now here are some decent tests I made on pudding. They resist sleep a bit (and a lot if you don't use appropriate gear). Keep in mind that I count partial resists as not resisted.


No staff, base skill=322 enfeebling, minor stat bonus (+0 INT +7MND) => 47.9% (500+ sleeps I)
HQ staff, base skill=322 enfeebling,minor stats bonus => 63% (300 sleeps I)
No staff, base skill 322 , +22 INT => 54.4% (500+)
No staff, base skill 322 , +12 mac, no stats bonus => 56.5% (500+)
No staff, base skill 322 , +11 skill ( = 333 total), no stats => 52.33% (500+)


- Notice that the HQ staff adds 15% more success (63%=47.9%+15.1%)
- Notice that 1 macc adds 0.71% so 20 macc should add 14.3%, assuming everything grows linearly. And no I don't have +20 mac gear avilable to test it. This is extrapolation. Especially if I want to keep 322 base skill.
- Notice that 2 INT add 0.59% so 10 INT should add 2.95% (extrapolation => MND and CHR work the same for related spells ...)
- Notice that 1 skill adds 0.40%
- So basically 1 mac >2 INT > 1 skill

Note : Flans have 89 INT, and I'm always below this level even with the +22 INT, being elvaan RDM/BLM with 64 INT. Keep in mind that due to the low # of sleeps the margin of error should be rather high but this should give an idea ...

So if everything is linear, you can add the accuracy from 20 mac and 10 INT and you get 14.3% +2.95%=17.25% vs 15.1% from HQ staff. Now ... there is some margin of error so don't take this test as "Alkalurops beats HQ staff 100%".