ARR Attack Magic Damage Test (beta3) Int, dtr & acc

[Deli]

I don't usually start a new thread, when I do, I test something...


Determination (DTR) scan test:

The yellow highlight ones are I made some mistake due to careless + I don't care. I put little mind inside and forgot to remove. oh well, I used test 2 similar range's mind value to back calculated the damage value without it. there can be more error on these 2 points just keep that in mind. The R value did increase after I tune the damage value down tho.






raw data

Int/mnd scan test:

I don't think I'll spend time to do the test after data will be saved, so I might as well just do it now. I instantly regret it at the first set tho... each set takes ~40 mins to gather @@:

Job: WHM






Note:

• If you know before, the regen amount has steps, such as 333-335mnd regen = 34 336-338mnd regen = 35, then you won't feel my this regen test is that odd.
• Regen can have critical cure within the normal values. Each regen will give 7 x single value back. If you ask me, I won't use it due to the cure/mp ratio.
• The results are pretty linear to me.
• here is the weapon damage test. each +magic damage = +3.39 in stone II.

all results are 50ish value averages.
I don't think I need to explain the tests this time. It's pretty clear.
Test targets are lv49 Air Spirits.

add in the +% per mind row, so you know by change that int/mnd, per int/mnd contributed that much +- hp.



new test raw data

Old tests:
Careful with what I do for the standard deviation... it might not be that good to do that. but...whatever...

Please note, below tests are just an initial test, to get a general sense how things work. It's not aiming to tell you how to do in general situation. Discussion and more test results, welcome.

Method:

• Use Conjurer with Stone II: so there is no bonus with spells use, while base damage is big enough to amplify the influence of the stats. (Assume they are multiplication relationship)
• Use Cleric Stance (Note: Cleric Stance will switch int and mnd value, so int<-->mnd. above listed were my mnd value and will become my int value after I use this ability. To simplify the naming, I just use int so that ppl won't get confused. for more info view the game manual.)
• lv49 Wind Spirits as test objects: 10 to 12 for each set. however, numbers of mobs are not important.
• Results are average from 100+ stone2 damage excluded critical damage. (Note: I only sampled 100ish damage values, if including critical values, my whole experiment will add in extra variable. Thus, excluded.)

Test1:
Goal: basic test with determination(dtr) and Intelligence(int) contribution to attack magic damage

summary: Average (avg) Standard Deviation (Stdev)
Int 320 dtr 225, Stone2 avg: 274.05 stdev:7.92 (Note: this one has acc+10, see test3)
Int 320 dtr 231, Stone2 avg: 277.33 stdev:9.08
Int 326 (from item), dtr 225 , Stone2 avg: 281.13 stdev: 8.15
Int 326 (from bonus stats points),dtr 225, Stone2 avg: 281.94 stdev: 8.35



discussion:

1. add Int is more effective than add dtr. (at this lv, this test. not in general, yet! just these data already took me 4 hours and 1hour to post it =.=)
2. When I was choosing gears, I've never though "acc" now is for everything, pysical and magical (warning! just my observation)... so keep that in mind within the test data there will be tiny flaw. The additional error coming from acc will be around 0.2%, however. see below acc test. common.
3. add stats from item roughly = to add stats from bonus points. The reason I did this test was because I remember SE did said something adjust item's stats contribution vs stats from the bonus point to make bonus point more stand out... but looks like it's not. However, if they are exactly equal, you can see my test error between set's 100 Stone2 avg will be 0.2%. please keep this in mind tho... some test doesn't really change that much...

Test 2:
Goal: how does dtr works, compare with low int value and see it's an addition or multipication relationship.

Summary:
Int 257, dtr 202, Stone2 avg: 229.77 stdev: 7.14(Note: this one has acc+10)
Int 257, dtr 208 (same +6 as above) avg: 230.04 stdev: 7.32 (Note: this stdev value (vs +10acc value above) does not match the other test either... can it be same problem as dtr that when you are lacking base value, these won't show up that much?)

Discussion:

1. dtr works as more complex formula. I can't get any relationship out of this test...
2. dtr almost have no contributuion...at low stats lv and this test

Test 3: ACC Test
Goal: to test why my mage ring has +acc stats...
Summary:
Int 257, dtr 208, stone2 avg: 230.04 (same as above) Stdev: 7.37
Int 257, dtr 208, Acc + 18 (Note: str, dex also +12) Stone2 avg: 231.05 Stdev: 6.71

Discussion:

1. There are tiny resist of magic involved. from the 100+ avg and stdev values, you can see the damage group probably more toward the higher non-resisted side.
2. keep that in mind my test error of these 100+ avg is 0.2%. so this tiny increase in avg damage is almost nothing.
3. You can go back to see above test to see the match of Stdev to confirm the credibility of this test, just little more.

finally, if you are curious, here is my test data.
(Note: I just share my test results with public, so the method and data will be clear, no fake tests. This is MY way to organize. If you don't like it, keep the "your data is so messy" to yourself. I don't need to organize toward [kupo!] automatic data-logger style. I <3 my hand input method.)
thread layout, however, I tried to make it as clear as possible ^^; still need lots works to make it looks clear.

derail but useful: (DD skill charts)

pug/mnk


ARC/Brd


Gla/pld


Lnc/DRG


MRD/War

[Ouchmaker]

Well done, good read!

[DubstepNinja]

Test 2:
Goal: how does dtr works, compare with low int value and see it's an addition or multipication relationship.

Summary:
Int 257, dtr 202, Stone2 avg: 229.77 stdev: 7.14(Note: this one has acc+10)
Int 257, dtr 208 (same +6 as above) avg: 230.04 stdev: 7.32 (Note: this stdev value (vs +10acc value above) does not match the other test either... can it be same problem as dtr that when you are lacking base value, these won't show up that much?)

Discussion:

1. dtr works as more complex formula. I can't get any relationship out of this test...
2. dtr almost have no contributuion...

.

Is it possible that determination has something to do with spell interruption rates?

Its effect on DPS seems almost nil in the tests

[rina_inverse]

so, still stacking INT?
got it

[Catapult]

Nice testing. It will be interesting to see if Determination has a greater impact on the disciples of war or not.

[Fantasia]

Is it possible that determination has something to do with spell interruption rates?

Its effect on DPS seems almost nil in the tests

I don't think so; the manual specifically says that Determination affects attack power, attack magic potency, and healing magic potency.

[Deli]

I need some more test points to plot a complete curve. I want to know it's a linear formula or it's a curved line, both int and dtr.
it makes no sense at base stats lv it almost boosted nothing. That's why I need further investigation.

for the replies above:

• for now, my priority is stacking best gear for int (mnd for whm) > Determination ~ acc ~ spell speed (if you really pay attention to spell speed and how fast it increase... ~300 spell speed, cure takes 1.98s; ~400 spell speed, cure takes 1.92s. 3% faster but that's like +100 spell speed...)
  
• There is someone else posting a Str vs dtr test on "other side," you guys can go search for it. the result for dtr was upsetting as I remember.
• this version, the spell interrupted rate is very low. I almost thought it doesn't exist. During the whole test, even though I have stoneskin protect me but still hp will drop to 60%, I only got interrupted once during "naked test" which I was only wearing gathering outfit. So, another test can be the vit vs interrupted rate. however. the chance is too low that I'll ignore this part in the near future...
• extra common for above one is that the game seem to shifted from lv-matter game to gear-matter game. Initially, my naked test was in really naked form. however, I almost got killed by the lv49 air spirits several times. That's why I had to abandon the data and boost up my magic defense. As many of you know, it doesn't really matter in 1.0 cus def gear didn't change that much. You could kill lots things naked. However, during this test, my damage were 150-180 while naked, and 110-130 while in gathering outfit. (note: even gathering outfit gives very good magic def. I think it's closed to top-notch lv50 crafted battle gears.) however, it can be just they rebalanced the mob strength.

[Laevetine]

Very nice info i never got around to doing much on the mage side of things, but i can tell you the same thing about Determination can be experienced on Physical damage side of things.

I did roughly the same going from naked to fully geared in different stats on PLD fighting the 45 basilisks near the ceruleum plant.

Its was hard to separate DET out by itself, but i could not find any good correlation to damage. Generally DET comes with other stats on gear and using what i determined on other stats DET seems like a /100 boost. (e.g. 23 DET roughly a .23 boost)

I tend to stay away from the other side apart from leaving feedback so I'm not sure what is being said there.

Minor tidbits i did find is 3STR = 1 Base Damage and 1VIT = 17 HP In case anyone wanted to know.

I started with no gear stacking up STR on each piece of gear till i was fully geared.

I fought sets of 4 Basilisks @ level 45 taking the average of the high and low 100 pot attack.

Then taking that number by the STR increase gave me a ~.33 increase in damage (give or take a small %)

I forgot to write down my starting STR, but i increased from base PLD with no allocation to +124 STR

Taking DET on top of STR I had a hard time seeing any real impact. (At most what i was speculating being a small % increase.)

(I'm no number wiz, but it seemed pretty on the level. Hopefully it might be useful to someone)

[Mudd]

Awesome work. Good read.

[PSxpert2011]

Magic DMG(who knew)

INT+MND= Mage class

<ta-da!!>

(see ino magicz) =P

edit:

MND= magic attack +
INT= enfeeble/ elemental defense +
(off the top of my head)