Conclusive jug testingFollow

I've not yet crunched the numbers, so I'll post them here for review. I'm likely to miss something as I don't nuke that much.
The nuking was done by:
level 99 blm/rdm
INT: 91 + 44
MAB: 30 from gear
there was also a f4% magic crit on the body piece,which may have influenced some of the samples.
All the elemental spells were tier 3 nukes.
darkness had drain, and I also did some w/ dark orb on blu/whm

blu had capped skill w/ no mab on.
INT 78
mnd 99

I cast banish2 on many of the mobs for /whm. I'm not sure how useful this will be as the m.acc will be very low. but I'll provide the numbers I have anyway.

The elemental nuking spanned fireday and earthday. The dark orb and banish were on following days but ended before lightning day.

 
	Tier 3 Nukes								 
	Fire	Earth	Water	Wind	Ice 	Thunder	Banish2	Drain	DarkOrb 
Raphie	741	345	49	700	801	522	25	280	597 
2nd	741	180	196	700	748	500		311	 
3rd	770	90	99	700		500			 
4th	741	180	196			500			 
									 
									 
broncha	534	417	446	477	514	586		208	 
2nd								155	 
									 
Louise	599	496	289	564	655	681	52	79	451 
2nd			289					129	 
3rd			304					150	 
4th			289					15	 
									 
Fargann	580	514	275	544	355	663	52	198	453 
2nd	580	498	275	544	635	663			 
					635				 
Julio	590	486	519	555	645	673	53	187	465 
2nd								245	465 
Yuly	673	555	593	47	733	762	16	238	545 
									 
Merle	599	496	529	564	655	681	10	232	468 
							10		468 
Shasra	580	475	507	544	661	663	12	259	243 
2nd	580	514	507	544	635	663			 
									 
Hobs	735	594	631	669	766	792	15	181	562 
									 
									 
Sieghard	590	486	519	555	645	673	12	304	465 
									 
falcorr 	607	493	529	568	664	694	13	280	480 
2nd							13		 
3rd							59		 
Gooey	448	395	206	419	678	518	7	107	343 
2nd	448	363	206	419	493	518		268	 
3rd		395			493				 
									 
Lulush	590	486	519	555	363	673	23	145	461 
2nd					645			181	 
3rd					671				 
4th					645				 
									 
Cetas	590	262	519	555	645	647		313	 
2nd		486							 
									 
Toloi	698	583	346	658	755	731	128	236	555 
2nd			346				15		555 
									 
Silas	549	443	29	485	582	613			 
2nd			29						 
3rd			31						 
 

pssst... remember stout servant.

Edited, Nov 23rd 2012 9:41pm by Xilk

Hey, here's a blast from the past! :D

This testing method is way easier now with the new SP2... I got several hundred Head Butt samples versus Wild Rabbits while trying to see if Ferine Cabasset +2 affected the TP mod.

Regular Head Butt damage:
592~621

Head Butt damage with Ferine Cabasset +2:
592~621

And I ended up with a bunch of outliers, just like you did: 77, 148, 149, 153, 304, 305, 306, 310, etc. Even though it's a physical move, just like Wild Oats, it's got these funny resist states (half, quarter, eighth). Weird, eh?

Edited, Dec 5th 2012 8:05pm by Bookmarku

Tried out some Gauge results with respect to the new SP2 ability. Target was VT tiger in Kuftal Tunnel basement:

Without SP2: It would be difficult to charm the Machairodus.
With SP2: Falkirk should be able to charm the Machairodus.

I mis-charmed once while trying this effect out, so it probably raises your Charm accuracy to whatever the cap is (95%?).

Edit: Alright, here's another Ferine Cabasset +2 test... I spent several hours down in Kuftal today trying to see if the BST AF3+2 head had any effect on the accuracy of Soporific. Used Soporific 600 times, half with the Cabasset, and half without. I removed all Beast Affinity levels, so it was a level 93 Presto Julio vs level 103 Machairodus for the entire duration:

                              Soporific Accuracy 
Without Ferine Cabasset +2          41.33% 
With Ferine Cabasset +2             43.00%

Such a minute difference could point to a tiny accuracy boost (and maybe other small hidden bonuses), or it could mean diddly-squat and be a coincidence.

Edited, Dec 6th 2012 4:16pm by Bookmarku

First sample is a 41.3% hit rate +/- 5.6%
Second sample is a 43% hit rate +/- 5.6%

So they aren't significantly different. Also, does Soporific have resist states? Like, Sleep II lands for 90 seconds or 45 seconds. What does Soporific do?

The more states it has, the more exaggerated you'd expect differences in magic accuracy to be. So +2 Magic Accuracy would be +2% hit rate with only two resist states (full duration and not landing), but it would be more than that (~3%) with 3 resist states, and so on.

Edited, Dec 7th 2012 8:07am by Byrthnoth

Wow. This is really fantastic.
Thank you for the time you put into it.

guess I should help find the def and vit for those tigers.. thought I figured def out before...

Edited, Dec 7th 2012 3:30pm by Xilk

Ok, the Ferine Cabasset +2 parse is now essentially done. Maybe one more thing I can do is remove the AF3+2 head and try to replicate the attack boost with other gear. How much attack do you think the increased monster correlation was giving? If there was some Pet: Accuracy-, I would be able to better gauge how big the accuracy boost was. ; ;

Well, my ignorance is showing - even with Pet: Attack +44 in gear, Discreet was still hitting for only 99 damage minimum on level 102 Tigers... How big of an attack boost are we getting from this dang hat? o.o

Maybe I can check the accuracy bonus on that super evasive Fly NM in Aby-Mis.

Edit: Alrighty, here it is then, the accuracy test: Went to Aby-Mis, used DG, MC, and Vicissitude atmas. I had no Beast Affinity merits, so my AudaciousAnna was level 95. Without any accuracy bonuses, my pet's accuracy was floored (20.80% from a 25 minute parse) so I used some Pet: Accuracy gear to get the base.

                             Accuracy % 
Pet Accuracy +49               41.80% 
Pet Accuracy +49 
   and Ferine Cabasset+2       52.30% 

Approximately a 10% accuracy boost - so is part of the Enhanced Monster Correlation a +20 Accuracy bonus?

Edited, Dec 8th 2012 12:32am by Bookmarku

So Attack+72 is approximately between the two. I'd say that the bonuses are approximately +50/100 Attack?

That makes me wonder how large the accuracy bonuses are, considering that you're still capped with AF3+1 head against the 102 tigers.

Okay, so approximately +10 and +20.

I was not able to close the ranges for pet attack values in my parses.
that is the range when fighting greater colibri always put the attack between ~700 and 1100.

I don't know what I was missing. However, I also did not find a difference in damage range between no attack+ gear vs greater colibri and lotso f attack+ when using Hobs.

Which would make you think its capped already.. but the damage range was not the same as vs level 0 mobs obviously.

I suspect level may have a very direct impact.

but its also very clear that they have different attack values from 1 jug to the next.

most stark contrast between gooey and hobs methinks.

Edited, Dec 8th 2012 7:11pm by Xilk

The easiest way to determine pet attack would be reminiscent of these tests:
http://kegsay.livejournal.com/5510.html

Plan:
* Take your pet out and find the highest defense target that you have capped Ratio against, where increasing your pet's attack does not affect its damage (monsters have level ranges so this could be difficult). - This is capped Ratio.

* We don't actually know what "capped Ratio" is for monsters (I think), so we don't know what factor of the pet's attack this represents. It *is* known that pet pDIF caps at 4*(1~1.05 Randomizer), for an upper range of 4~4.2 times the base damage. This means that the lowest damage your pet does will always be a multiple of 4 when you're attack capped.
Example: If you add 5 attack and cap your damage on a level 30 mob, then you figure out that the mob has 250 defense, your pet's attack could be ~500 (2x), ~562 (2.25x), 750 (3x), or 1000 (4x). I expect it would be closer to 4x than 2x.

* You can test this. If you find out that a level 99 pet caps attack on a specific level 54 monster with +72 attack in gear and then find that it caps attack on a level 50 monster with +0 attack in gear, then you can figure out the real defenses of the level 54/50 monster and figure out the slope between them.
Example:
- Your pet caps attack against a Level 54 monster with 250 defense using +72 Attack in gear
- Your pet caps attack against a level 50 monster that has 232 defense using +0 attack in gear
- 72 Attack / 18 Defense = 4, so pet Ratio caps at 4.0.

* So if you establish what Ratio is, then you can back-calculate how much attack your pet has.
Example:
- (Pet Attack + 72)/250 = 4, so Pet Attack = 1000-72 = 928 Attack

Edited, Dec 8th 2012 7:05pm by Byrthnoth

Hmmm... HMMM!

I think a lot of what Byrth wrote went over my head, haha.

Would using the Carmine Eruca in Wajaom Woodlands (always level 70) be of any assistance?

Edit: It's nothing concrete, but to give an idea of where the Aymur (99) Attack Bonus V stands:

Attack Bonus V Testing vs Level 104 Machairodus:

                                  Melee Dmg     Crit Dmg    Crit. Hit Rate   Accuracy 
Without Headgear/Aymur             86~149       182~276         14.17%        84.51% 
With Aymur (99)/Ferine Earring    103~170       200~299         15.29%        86.24% 
With Ferine Cabasset +2           105~175       207~303         12.12%        94.85% 

It's slightly less than the Ferine Cabasset +2 bonus (though without the Monster Correlation restriction, which is nice) - probably in the 70~85 Attack range? Maybe we can better hammer this out when we know more about pet attack values...

Edited, Dec 9th 2012 10:49am by Bookmarku

I understand what Byrth is saying. I've just been dragging my feet on it. cuz its a pain separating parses and a pain finding the right mobs. YOu don't kno if you have it until after you look at the parse data. I've collected parse data vs djinn, troll pld, and pygmytoise. btw, pygmytoise and pld mobs are a bad IDea because they use protect or hardenn shell which changes their def... then you'd have to sort out all the hits between different def values.

And I was missing the equation about how to compare the attack/def ratios.

uhmmm where did that 72 number come from.. also, did you try aymur 99 and ferine together? did they stack? cuz if they DON"T stack (and that would suck). that sound like the attack bonus 5 trait which DRK has... and it would be from both..

then again I don't suppose that woudl be it, cuz warriors have attack bonus trait also, and it wouldn't be +72 you would lose 50 attack as the difference between attack bonus 2 and attack bonus 5

Edited, Dec 9th 2012 11:26pm by Xilk

I've done a few more parses to further examine the effects of Ferine Cabasset +1/2. I used my favorite mob as a target (the eternally level 70 Carmine Eruca - he is still unable to beat Crawler-Maat no matter how hard he tries) and for his lizard family opponent I selected the readily-available LizardFamiliar. Before starting, I took a stroll down memory lane and did a level cap check on this jug pet:

As a lvl 99 BST w/ 0 BA Merits: Healing Salve restored 714 HP - 1428 MaxHP (level 45?)

I wanted to be really sure that the jug pet info was accurate, so I edited my Main Job level and checked it out.

As a lvl 45 BST:
1: Healing Salve restored 655 HP - 1310 MaxHP (level 43)
2: Healing Salve restored 685 HP - 1370 MaxHP (level 44)
3: Healing Salve restored 714 HP - 1428 MaxHP (level 45)

Since it appears that LizardFamiliar gains ~59 HP per level, the level 60 version should theoretically have about 2313 MaxHP.

As a lvl 99 BST w/ 5 BA Merits and augmented Monster Gloves +2: Healing Salve restored 1153 HP - 2306 MaxHP (level 60)

Close enough! :) When the BST job was first released, did they unveil the caps for these early jugs? Kinda curious what the source is, haha. In any case, it's accurate!

LizardFamiliar (lvl 60) vs. Carmine Eruca (lvl 70)

                            Melee Dmg     Crit Dmg    Crit. Hit Rate   Accuracy 
Without Headgear/Aymur        27~61        70~120         9.57%         57.91% 
With Ferine Cabasset +1       31~61        75~121         9.80%         64.23% 
With Pet: Attack +25          31~61        76~122        11.35%         61.67% 
With Ferine Cabasset +2       35~64        83~130        10.47%         72.91% 
With Pet: Attack +50          35~64        85~128         9.52%         62.22% 
With Aymur (99)               41~71        91~139         7.84%         62.04% 
With Aymur (99)/Cabasset +2   52~87       102~150        10.10%         72.98% 

Edit: Used Astolfo (Pet: Attack +22), Beast Torque (Pet: Attack +3), Mirke Wardecors (Pet: Attack +15), Scorpion Mittens (Pet: Attack +3), and Desultor Tassets (Pet: Attack +7) to get a Pet: Attack +50 parse... and the damage range is looking nearly identical to the Ferine Cabasset +2 parse.

Edit2: Aymur (99)'s Attack Bonus V stacks up with Ferine Cabasset +2's Monster Correlation bonus.

Edit3: Got another set of data with Pet: Attack +25 in gear - nearly identical damage range compared to the Ferine Cabasset +1.

So here's a question: if you have something that increases your attack by x% (like Phorcys Korazin, Drachenhorn), does that get calculated before or after Attack Bonus traits? The reason I ask is because, when looking at this info, it appears that Aymur (99) is still giving about 87~90 attack (which is consistent with Byrth's estimation), but the bonus of the Cabasset +2 seems to vary between Non-Aymur/Aymur parses. In the Aymur test, the Cabasset +2 is giving roughly Pet: Attack +60~65 (which would make sense if the bonus to EMC is +10% attack and LizardFamiliar's Attack was around 600ish after the Aymur boost). Just wanting to clarify if that sounds like a reasonable explanation.

Edited, Dec 12th 2012 7:46am by Bookmarku

Nice! Thanks for clearing that up. :)

Heyyy, good idea. :D I'll help if I can~

What pet/prey are you planning on doing? Something DC? I was thinking of maybe doing some lvl 93 PrestoJulio vs 93/94 Hispid Rarabs... :o

Edit: Version Update: The issue wherein pet movement speed was unintentionally fast has been resolved, and the maximum movement speed of pets has been increased to twenty-five percent of the average speed. :O :O :D

Edited, Dec 12th 2012 12:51pm by Bookmarku

Oooh, new BST-only Axe:

Ravana's Axe Level 99
DMG:75 Delay:276 HP+40 Accuracy+10 Attack+10 Pet: Adds "Regen" effect

And:

Skadi's Bazubands +1
DEF:32 STR+8 AGI+8 Attack+12 Ranged Attack+12 Haste+4% Pet: Haste+3% Set: Increases rate of critical hits

Sigyn's Jambeaux
DEF:29 STR+7 DEX+7 VIT+7 AGI+7 Evasion+8 Haste+4% Pet: Evasion+5

Edit: That note about Pet Movement Speed... I took it to mean that all pets will go from regular speed (100%) to an increased speed across the board (125%). Maybe?

Edited, Dec 12th 2012 1:26pm by Bookmarku

I'm still out-running CrudeRaphie. But Merle is seemingly fast... Hard to tell.

Some of the Skadi/Sygin gear seems decent. Toci's/Ocelomeh+1 still seem king for non-Killer, though you can replace Phorcys Salade with Skadi's Visor +1 for a lil' improvement in your Killer Effects setup.

Skadi+1 head/Ferine+2 body/Skadi+1 hands/Phorcys legs/Ferine+2 feet seems quite nice. Rioter's Collar is basically equivalent (can be slightly better or worse depending on gear/buffs) to Rancor Collar if you want to avoid the DT+10% and can't get Portus Collar.

Edited, Dec 12th 2012 3:25pm by Bookmarku

Looks good, sir.

A neat piece of info - looking back through the thread, it seems like the topic of a Pet: Haste cap has been on the back burner for about 11 months, haha. Thanks for laying it to rest. :)

How would PVP make this easier? You could control the player Def better and adjust it, but collecting a large enough sample size would be very difficult in pvp I believe. You would need at least 3x ppl. A bst, a tank (probably pld...) and a whm to help keep pld alive. Pets do ALOT Of damage for players. In pvp its not hard to 2~3 shot a player w/ a bst pet.

Also, How large of a sample size do we need? I've been hunting around trying to find the right mobs... and its very tedious. I have some abberations in the numbers though.. I'm around 200~300 samples /mob level. I"ve also been trying to do it w/out using ferine cabasset or aymur, though I've kept that option open by going after lizard types w/ hobs. I've done seaboard vultures, peiste, and sand lizards. The first 2 were way off. Its easy to see where Pdif caps on crit hits, but much harder to get it there for non-crit hits.

Do I need the low hit to cap, or the high? I have the high end capping on sand lizards. Below are my parse summaries.
(not sure the exact levels... wiki says 61~66, but thats off by 1. There are only 5 different levels, not 6).

 
Sand Lizard 	Def	Hits	M.low	M.high	C.Hits	C.low	C.high 
(20 xp)      	246	222	466	495	40	509	534 
(22 xp)      	251	244	449	495	36	508	534 
(24 xp)      	256	255	451	495	40	509	534 
(26 xp)      	261	188	445	491	30	505	530 
(28 xp)      	266	260	432	491	41	505	529 
							 
							 
							 
att +40							 
Sand Lizard 	Def	Hits	M.low	M.high	C.Hits	C.low	C.high 
(20 xp)      	246	219	459	495	40	509	534 
(22 xp)      	251	188	465	495	44	509	534 
(24 xp)      	256	269	454	495	46	510	534 
(26 xp)      	261	298	441	491	43	505	530 
(28 xp)      	266	251	435	491	37	505	530 
 

I would need to get 471 damage as the low melee hit to cap pdif on the 20~24 xp lizards. This is easy to find by using adjusting between crit values and non-crit values using a factor of 1.08 for the crit damage bonus. However, you can see that the high end of the melee hits is already capped in both samples.


... and, as you can see, I have some inconsistancies. The low hits for the attack +40 sample should not go lower than the naked sample.

adding ~40~60 attack doesn't make much difference. I did a short parse on the peiste after my first 2, adding both aymur and cabbasset.. and the jump in damage was very noticeable. I estimate that I added ~225 attack that way.

If I go much lower for mobs, then I won't get anymore xp, and it will be more tedious to separate the mob levels. I think level 56 is as low as you can go at level 99, correct? I hope there is a difference in xp from a 56 to 57 mob...

Edited, Dec 20th 2012 1:09pm by Xilk

Edited, Dec 21st 2012 1:04am by Xilk

IF anyone finds these useful, I'm posting my other parse summaries.

 
seaboard vulture 
              	Def	Hits	M.low	M.high	C.Hits	C.low	C.high 
Lvl 83        	355	254	287	401	35	461	515 
att + 62	355	272	301	403	28	455	515 
Lvl 84        	361	325	289	387	39	444	511 
att + 62	361	178	296	394	21	460	508 

 
Peiste 73~76	Def	Hits	M.low	M.high	C.Hits	C.low	C.high 
73 (44xp)	310	142	368	466	22	488	511 
74 (47 xp)	315	264	365	466	40	488	511 
75 (50 xp)	320	263	354	455	38	483	507 
76 (52 xp)	325	200	350	449	33	483	507 
							 
ATT +62							 
Peiste 73~76	Def	Hits	M.low	M.high	C.Hits	C.low	C.high 
73 (44xp)	310	247	371	474	26	489	511 
74 (47 xp)	315	255	360	465	38	489	511 
75 (50 xp)	320	196	355	455	32	483	507 
76 (52 xp)	325	252	349	445	40	483	507 

Edited, Dec 20th 2012 1:29pm by Xilk

Lizard defenses from /check:
Level 59 Lizard: 230 Defense
Level 61 Lizard: 240 Defense
Level 62 Lizard: 246 Defense
Level 63 Lizard: 251 Defense
Level 64 Lizard: 256 defense
Level 65 Lizard: 261 Defense
Level 66 Lizard: 266 Defense

I verified that the defense is the same between Terrigan and Kuftal.

The next steps would be:
1) Determine how much +Pet Attack you need to cap on level 61 Lizards.
2) Determine how much Attack you need to add to start capping on level 62 Lizards (if you can).

You know that +64 Pet Attack caps you on 61 and not on 62, and looking at the splits I'd say that you're fairly uncapped on 62 still. I'd take off ~5 pet attack and see if your minimum changes. Also, you should be able to use the above defenses to target specific levels using /check if you're careful about it (maybe use a weapon with sub job skill).

Player ratio caps at 2 for 1H weapons and 2.25 for 2H weapons. Kegsay assumed a similar cap for Avatars in half of his analysis, but looking back at his data that's probably not correct. I'd personally guess that monster Ratio caps around 3~4, so you'd be looking at +18~24 attack to reach between tiers.

So Gorefang is capped vs. a 230 Defense monster at +0 Pet Attack and a 240 Defense monster with +62 Pet Attack, but not against a 246 Defense monster with +62 Pet Attack. Assuming all of those statements are true, this indicates that Pet/Monster Ratio cap is greater than 3.875. Assuming that we are almost at the Ratio cap for 61 and 59 Lizards, it would be approximately 6.

When Monsters fight Players, they get +0.05 Ratio per level. So when Ig-Alima (Level 120) attacks you (Level 99), he gets +1.05 Ratio. It is not known whether this also applies to monsters that are attacking other monsters, so that would be another factor that we might have to control for in the future. For instance, Hobs might be getting (40*0.05 =) +2 Ratio just because of level difference against the 59 Lizards. He would only get +1.9 Ratio against the level 61 monsters, which would make the attack needed to move between the two levels more dramatically different.

Edited, Dec 21st 2012 9:32am by Byrthnoth

Edited, Dec 21st 2012 9:46am by Byrthnoth

I would try the test you suggested first to verify that attack isn't actually having any effect. For additional potency, you could pull the monsters with Dia II.

However, it's possible that monster Ratio is capped and then their level correction term is stacked on top and can surpass the cap. So something like:
Attack/Defense = Ratio
(Attacker's Level - Defender's Level)*0.05 = Level Correction
Ratio + Level Correction = Effective Ratio

Ratio could be capped at, say, 2, and then Level Correction could be the only thing you're actually varying in these tests. This would also explain why BST pets wreck monsters that are a lower level than them so efficiently. It would also explain why, unless I'm mistaken, Kegsay never convincingly showed a difference between Attack+ cases and cases without +Attack because he used a very similar method to us. Also similarly, he was targeting monsters ~40 levels below his Avatar (Level 35 Bees at 75).



An interesting test would be to use a Level 99 character/MNK in Ballista and let a pet melee them with Counterstance up. That would show you if level correction is required to reach 4.0 pDIF.

Edited, Dec 21st 2012 4:20pm by Byrthnoth

Considering you got a lower Melee Low for 62 Lizards, I'm not very convinced that there was any change.

Okay, I was able to get a few ls mates to help check the mnk aspect.
I had full hp+ gear on, and NO damage mitigation gear (black belt, brawny adargas, etc...)

w/ counterstance up full time, I had 70 defense.

also 119 Vit

These are gorefang's hits against my 99 mnk/war

 
Hit/Miss    M.low/Hi    M.avg    #crit    C.low/Hi    C.avg      C.rate 
121/24      102/238     198.76     16     257/328     288.63   13.22 % 

as you can see, its a small sample size.. unfortunately, I forgot that brenner has 'official' matches which boot you out... so we were locked out from getting a larger sample.

However, it is clear he is not getting near the 4.0 ratio we see with pets vs lower level mobs.

Edited, Dec 24th 2012 10:15pm by Xilk

Yeah, sorrry, I'm pretty sure thats a guard proc also. I was facing him at the beginning, but wasn't getting many hits and then later decided to turn around. Below are the offense details for gorefanghobs. the next lowest hit is 154.

I can probably do that charmed counterstance test. on a non-mnk job /mnk so we don't get guard procs. However, I think getting an nm which reliably charms might be easier. I don't have brd or know any brd who have /mnk but I can /mnk on blu w/ 1 damage clubs and get charmed by raskovnik in aby konschtat. just need to make sure there are no significant dd atma on. and plenty of regen.

Spoiler
GorefangHobs
Melee
102: 1
154: 1
156: 1
160: 2
161: 1
163: 1
164: 1
165: 2
166: 1
167: 1
171: 2
173: 1
174: 2
176: 2
177: 1
178: 1
179: 1
180: 3
181: 1
182: 1
183: 4
184: 3
187: 3
188: 1
189: 1
190: 2
195: 2
196: 1
197: 1
198: 3
199: 1
200: 1
^ 201: 3
203: 2
205: 1
206: 2
+ 207: 5
208: 4
209: 3
210: 2
211: 1
212: 1
214: 2
215: 2
217: 3
218: 2
220: 1
221: 3
223: 3
227: 1
228: 1
229: 2
230: 2
234: 2
235: 1
236: 3
237: 1
238: 2