Enspell II + Deathblossom testing.Follow

I'm glad to see some renewed interest in this entire thread from other people, honestly. I abstained from replying a few days ago because while something didn't sit right with me, I really didn't know where else to go with my ideas.

That being said, my idea of using enspells to test against Barspells to determine if they truly are just an addition to an "M.Eva" stat are out the window at this point. I've had statistics courses, and I can follow it for the most part (ie: Kaeko and Glamrawkus' posts make sense to me), but I'm poor at performing the analysis on my own. So while I wasn't able to run the numbers to determine if there was any suspect data in RDDs tests, it makes sense to me from the analysis others have provided.

That being said, I would like some input from others' points of view on how to go about testing Elemental Resistances' effects in a larger context. I'm almost positive a Ballista-type scenario would be appropriate, but I wouldn't want to waste time if the testing grounds' themselves induced error into the test.

Edited, Apr 26th 2009 8:23pm by ChanchanXI

Do you know what I'd like to see Enspell/Deathblossom tested on?

Tanihwa.

But that Bugard resists Nukes and Physical damage so much it's frustrating.

I think what I'm aiming for is something like the following:

If my M.Acc% (ie: my chance of landing a completely unresisted spell) is X, what happens to it when I add Y Elemental Resistance to the target? In addition, does adding elemental resistance cause a change in the amount of half, quarter, sixteenth resists?

That being said, Bio might actually be a good way of doing this coupled with Dark Resist gear and/or Dark Carol. I don't know if Dark Magic works the same way as Elemental Magic with regards to M.Acc v. Skill, but I would be shocked if it didn't.

I still wonder if performing the test in Ballista would be appropriate, given the record of people not trusting (justifiably) data that comes out of testing inside Ballista.

He has a very good testing procedure (mainly because you can almost idle while collecting large amounts of data with a parse), so that alone is both useful and admirable. Glam really suggested the data being too perfect first, but I am inclined to at least be skeptical like he is - you would be dumb not to at least look twice at something so spot on perfect after 400 trials, twice. I don't know RDD and don't want to imply something as under-handed as fudging numbers - I'm just saying Glam was right and smart for noting the 'perfectness' of the data.

With regards to testing with Bio, this will work as long as you find a mob you can be confident has no innate dark resists. If you want to test using non-enspell sources, the issue is really how tedious it can get because you have to cast instead of just idling and attacking. So trials take longer to come by.

One way to sort of 'cheat' with this, though, is to just make a looping windower macro that casts bio, waits 1.5 seconds, and repeats itself. (am I allowed to talk about windower here? lol) This way you can just check back every minute to reapply DB (which you can really automatically do as well). DB would be annoying to test though because the MEVA down effect won't overwrite itself.. You must wait for it to wear off then reapply.

You can try a looping macro like press once and you get...

DB
/wait 2
Bio
/wait 5
Bio
/wait 5
Bio
...
...
...
Count number of Bios you can get off in a minute before DB wears off
stop
loop back to using DB

This way you can make sure no stray Bios go off w/o the MEVA effect on. I know DB MEVA down works even if the WS misses, so you don't have to worry about missing, just collecting 100 TP every minute (easy on lower stuff, just make sure you don't kill it).

Also, you obviously have to test using a control too w/o DB if you want this.

EDIT: also there is a level difference factor that plays into resists, so if you pick a mob just really low leveled, the resist rate will go down a lot. How exactly level difference affects it is unknown, just that it exists.

Edited, Apr 27th 2009 5:25pm by Kaeko

Yes. Contrary to BG belief, we hate people who bitch about stupid stuff here too. We have an entire thread in the sticky I think from when spellcast came out, though I don't really care for spellcast's "on the edge" ness myself. The most you will get is lolratedowned.

Edited, Apr 28th 2009 2:51pm by LordMnementh

Edited, Apr 28th 2009 2:52pm by LordMnementh

Not really. Whatever point you were trying to make, you still fell for the gambler's fallacy. Each test (flip, melee swing, element resist) is independent and not influenced by previous test.

Basically, that first flip only has two outcomes, H or T. After that flip, assuming H, you still only have 2 outcomes, HH or HT. After the third, assuming another H as you did, you still only have 2 outcomes, HHH or HHT. At the start, assuming 3 consecutive flips, you have 8 possible outcomes, hence a 12.5% possibility of HHH, but each flip is still independent.

Same with melee swings. 3 consecutive swings at capped accuracy, there's a roughly 87% chance of Hit/Hit/Hit. But on the third swing, assuming the first two are hits, there's a 95% chance of a Hit, since the probability between Hit/Hit/Hit and Hit/Hit/Miss is still 95 to 5.

I was ignoring the thread up until now, and that's all I really had to comment on. Anything else I can say has been said.

I just thought of an even faster way to test MACC.

1)Get a LOT of MACC gear (lol wise)
2)Find a Chigoe that can hit you
3)Cast Blaze Spikes and survive

Best part is that you can do this on pretty much any job also meaning that you can use different jobs base enhancing skill to help determine accuracy as well.

The only factors you would have to be careful at that point is what levels the chigoes are, what level are your jobs, what day is it, and what's your enhancing skill (also try to keep your attributes the same while switching around MACC and jobs).

The best part is that it wouldn't be that difficult at all to get a large sample size. Those things swing so fast that testing with a lot of different set-ups and jobs wouldn't be that difficult at all.

Variables to check should be.

1) What effect does MACC have on the spikes hit rates (try to keep your attributes the same)?
2) What effect does increasing the level difference have on increasing the hit rate (careful, try to get the same Enhancing skill and same attributes)?
3) What effect does increasing your attributes do (probably the easiest to test)?
4) What effect does adding Enhancing skill do? (same as 1, keep attributes the same)?
5) What effect does correlating weather do (harder to test obviously(also try to use the same day and the opposing day. If the difference between your base is the same the relationship is +accuracy. If they results are different by a wide margin the relationship may be +%accuracy)?

Would do some of this right now if I wasn't swamped in schoolwork D:

Edited, Apr 30th 2009 5:02am by SlashAnonymous

I picked up Kparser, but for some reason it spit an error at me when I tried to start a parse and never really bothered to follow up on it. At the time, I'd just stepped outside WG to try poking Lessers with Enaero and then see if Death Blossom helped damage any. Without, I was mainly doing 1s. When DBs was in effect, it didn't really change. I'd maybe get a 2 or 5 every 1 out of 10 swings casting at 301, but the -MEVA wasn't at all enough to really hurt their innate Wind Resist. Preliminary opinion suggests it's about as mediocre as our Enspell IIs in that respect. I dunno, some people might like it more, but I've been disappointed with DB overall since I first unlocked it and it's not like my gear is sh*t.

On fresh ones I could usually get 100 TP before they do a TP move on me. Anytime after is hit and miss, but even with my Enspells gimped but their resist, I usually had one dead in under 90 seconds.

Think you're grasping at straws with that one, but I can take a few swings and see what happens.

Edit: Yeah, Lessers don't seem to resist Enstone very much. Since I'm swinging with Fencer's active, expected damage is 25. Most of the time it'd be that, 22 being a resist. If I was over 100 TP because I was holding WS, it'd be 20 or 18. After popping DB, this didn't really change.

Damage Summary 
Player               Total Dmg   Damage %   Melee Dmg   Range Dmg   Abil. Dmg  WSkill Dmg   Spell Dmg  Other Dmg 
Seriha                   41196   100.00 %       28482           0           0        4404           0       8310 
Total                    41196   100.00 %       28482           0           0        4404           0       8310 
 
 
Melee Damage 
Player            Melee Dmg   Melee %   Hit/Miss   M.Acc %  M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit% 
Seriha                28482   69.14 %     364/16   95.79 %     51/96    70.32     54   107/147  123.74   14.84 % 
 
 
Other Magical Damage  (Additional Effects and Spikes) 
Player            M.AE Dmg  # M.AE  M.AE Avg   R.AE Dmg  # R.AE  R.AE Avg   Spk.Dmg  # Spike  Spk.Avg 
Seriha                7818     334     23.41          0       0      0.00       492       82     6.00 
 
 
Weaponskill Damage 
Player                 WSkill Dmg   WSkill %  Hit/Miss   WS.Acc %   WS.Low/Hi   WS.Avg 
Seriha                       4404    10.69 %      12/0   100.00 %     253/421   367.00 
 - Death Blossom             4404   100.00 %      12/0   100.00 %     253/421   367.00

Small sample, yeah, but I didn't feel compelled to keep at trying Enstone with it doing nothing really out of the ordinary. The parsing program is lagging me a little, too. Not sure if there's anything I can do to cut that down, personally.

I'll do some Enaeros just so people believe me one the 1s.

Edited, May 1st 2009 1:03am by Seriha

Some Iceday smitings. I'm tired. z.z

Damage Summary 
Player               Total Dmg   Damage %   Melee Dmg   Range Dmg   Abil. Dmg  WSkill Dmg   Spell Dmg  Other Dmg 
Seriha                   82798   100.00 %       61720           0           0       16112         278       4688 
Total                    82798   100.00 %       61720           0           0       16112         278       4688 
 
 
Melee Damage 
Player            Melee Dmg   Melee %   Hit/Miss   M.Acc %  M.Low/Hi    M.Avg  #Crit  C.Low/Hi   C.Avg     Crit% 
Seriha                61720   74.54 %     816/46   94.66 %     50/97    70.09     78   106/151  128.10    9.56 % 
 
 
Other Magical Damage  (Additional Effects and Spikes) 
Player            M.AE Dmg  # M.AE  M.AE Avg   R.AE Dmg  # R.AE  R.AE Avg   Spk.Dmg  # Spike  Spk.Avg 
Seriha                3542     712      4.97          0       0      0.00      1146      196     5.85 
 
 
Weaponskill Damage 
Player                 WSkill Dmg   WSkill %  Hit/Miss   WS.Acc %   WS.Low/Hi   WS.Avg 
Seriha                      16112    19.46 %      42/0   100.00 %     124/555   383.62 
 - Death Blossom             8652    53.70 %      24/0   100.00 %     124/555   360.50 
 - Vorpal Blade              7460    46.30 %      18/0   100.00 %     349/522   414.44 
Seriha 
  Melee Additional Effects 
        0:   36 
+       1:  104 
        2:   16 
        3:   22 
 ^      4:    1 
        5:   34 
        6:   36 
        9:    3 
       10:   48 
       12:   56 
  Weaponskill 
    Death Blossom 
+     124:    1 
+     241:    1 
+     283:    1 
+     320:    1 
+     343:    1 
+     362:    1 
+^    380:    1 
+     398:    1 
+     402:    1 
+     456:    1 
+     462:    1 
+     555:    1 
    Vorpal Blade 
      349:    1 
      354:    1 
      390:    1 
+^    402:    2 
      426:    1 
      437:    1 
      448:    1 
      522:    1