Forum Settings
       
Reply To Thread

Enhances Avatar Attack ExplainedFollow

#1 Oct 11 2009 at 2:13 PM Rating: Excellent
Sage
***
1,181 posts
tl;dr
-The attack bonus scales based on monster defense.
-Formula is AttackBonus=7*(AVATAR_ATTACK(before boost)/MONSTER_DEFENSE)^2
-This means the attack bonus is rubbish on hard monsters, and good on weak monsters (figures).

======================


How to increase Attack:
-Use Pet:Attack+ gear
-Use "Enhances Avatar Attack" gear.

Just what does this "Enhancement" do though?

The attack granted by Enhances Avatar Attack

I tested this by doing 100 blood pacts on Greater Colibri. 50 were done without Summoner's Pigaches, 50 were done with. This 50 excluded Criticals and Misses, so I actually did more like 60+ each.

Test A
Target: Greater Colibri (VT++) - L81/L82, difference in DEF=5 (should be negligable, VIT is the same.)
Known: Enhances xx gear does not affect melee hits. Merits in Attack do affect melee hits.
Aims of test: How much Attack does "Enhances" gear add?
Stats to note: 1 merit in Attack (+2 Attack)

M=Miss, N=Normal, C=Critical
50 tests performed each. (excluding misses and criticals)
No Spinning Dives were performed in Water weather or Watersday or Thunder Weather or Lightningsday.

With SMN Feet:
Mean=761.84
Standard Deviation = 95

N|585,587,599
N|607,610,624,627,634
N|651,658,664,668,673,683,698
N|705,708,709,715
N|751,754,776,777,779,785,785,786,789,790,791,796
N|807,808,832,833,835,845,847,848
N|861,864,871,873,875,877,878,885,895,897,897
N|

C|1051,1240,1282,1334




Without SMN Feet:
Mean=775.94
Standard Deviation = 90


N|586,592,593
N|629,630
N|653,660,670,687,691
N|700,703,704,708,709,717,731,734,741
N|760,768,771,776,787,794,798,799
N|812,813,814,820,821,825,832,843,843,849
N|854,855,863,866,868.869,874,878,890,893,898
N|910,916

C|1026,1050,1122,1134,1201,1248,1301


Main observation: No difference here. The mean even suggests I deal more damage without the feet! Note the standard deviation for the "with" test is greater than that of the one "without". This could be hinting that I am doing higher damage than the mean more often, but then again it is not conclusive proof. Perhaps at higher levels (where DEF is much higher) the effect is less noticable. I shall repeat the test on Lesser Colibri this time, to hopefully outline a bigger increase in damage.

Test B
Target: Lesser Colibri (EP--/EP-) - L63-65, will not fight Colibri shown as EP- (L65), will only fight L63-64 (VIT is the same for those two, DEF difference is 4 which should be negligable.)
Known: Enhances xx gear does not affect melee hits. Merits in Attack do affect melee hits.
Aims of test: How much Attack does "Enhances" gear add?
Stats to note: 1 merit in Attack (+2 Attack)

M=Miss, N=Normal, C=Critical
40 tests performed (excluding misses and criticals)
No Spinning Dives were performed in Water weather or Watersday or Thunder Weather or Lightningsday.


With SMN Feet:
Mean=1200.2
N|980
N|1006,1035,1044,1048
N|1067,1071,1072,1072,1075,1089,1092,1097
N|1121,1121,1139
N|1170,1172,1181,1197
N|1218,1228,1231,1233
N|1263,1272,1274,1275,1279,1280,1281
N|1303,1309,1310,1316,1322
N|1359
N|1430
N|1488,1488

C|1725

=980+4133+8635+3381+4720+4910+8924+6560+1359+1430+2976


L63|980,1035,1044,1048,1067,1071,1089,1092,1121,1121,1139,1170,1197,1231,1233,1263,1274,1279,1280,1281,1309,1310,1488,C1725
L64|1006,1072,1075,1097,1172,1181,1218,1228,1272,1275,1303,1316,1322,1359,1430,1488


Without SMN Feet:
Mean=1130.775
Standard Deviation =
N|995
N|1001,1004,1010,1024,1031,1033,1035,1045,1049
N|1050,1052,1058,1063,1068,1078,1091,1097
N|1101,1103,1112,1119,1119,1126,1137
N|1150,1175,1186,1192
N|1205,1205,1213,1213,1219
N|1279,1285
N|1312,1314,1335,1347
N|
N|
N|

C|1615

=995+9232+8557+7817+4703+6055+2564+5308


L63|995,1031,1035,1045,1049,1063,1068,1091,1101,1112,1126,1137,1150,1175,1205,1213,1314,1347
L64|1001,1004,1010,1024,1033,1050,1052,1058,1078,1097,1103,1119,1119,1186,1192,1213,1219,1279,1285,1335,C1615

Well well, this is surprising! There is a noticeable increase in damage, giving on average 70 more damage with the feet equipped. Why wasn't this noticed on the Greater Colibri test? Just how much attack is added? Is this scaling?? Let's try to answer some of these questions.

First off, there was no noticeable increase in damage on the Greater Colibri test. This would indicate one of two things:
A - The attack bonus granted is small (<=10)
B - There is a small multiplier but because the value it was multiplying by was small, there was no noticeable increase.

Explanation "A" seems unlikely with the giant (compared to the first test) leap of damage seen on the Lesser Colibri. This leaves explanation "B", a multiplier is in effect on the PDIF calculation. This begs the question, how much and where? This is further backed up by an earlier test with Mountain Buster on LV0 monsters as shown below:

Mountain Buster (No Boots)
12x Multiplier
RANGE:1075-1612
CRIT:1747-2016

Test Min/Max:1174-1569
|1174,1186
|1203,1221,1227,1251,1298,1299
|1307,1315,1336,1386,1389
|1481
|1526,1532,1569

1841c,1966c,2008c

Mountain Buster (With Boots)
12x Multiplier
RANGE:1075-1612
CRIT:1747-2016

Test Min/Max:1265-1566
|
|1265
|1325,1328,1358,1366,1383,1390,1393
|1426,1450,1467,1473
|1530,1530,1536,1540,1566

1877c,1881c

The equation for Physical Blood Pacts is known to be:

(D+fSTR) * M * PDIF

If M was being increased, you would expect Greater Colibri to take a noticable amount more damage. This is not the case.
If D and/or fSTR was changing, you'd expect the range of damage dealt to be hard-shifted up (i.e. from 100-200 to 150-200, no low end values). This is not the case as the lowest scores in both tests were with the boots equipped.

This leaves PDIF.

PDIF is calculated using ATTACK/DEFENSE (commonly referred to as "Ratio"). This is level corrected for PCs (more like penalty for every level a PC is below the enemy) and then called cRatio. We know Avatars do not have this correction, making them ideal for Wyrms and the like. This ratio is subject to a cap of 2.0 for both Avatars and PCs.

An equation is then applied to this ratio depending on how big or small it is, the equation varies for both minimum and maximum. A random value is then chosen between this min and max and that becomes your PDIF.

Summoner's Pigaches seem to add in a multiplier to Ratio. If MinPDIF was affected, you wouldn't be seeing low numbers with the equipment on, but we are. If MaxPDIF was truly affected, you should see higher values all the time, but we aren't (Greater Colibri Test). This strongly suggests that Ratio is being manipulated, but how?

Figuring Out Avatar Attack

Using the control tests, we should be able to devise how much Attack an avatar has.

Greater Colibri Average Defense = 324.5
Lesser Colibri Average Defense = 233

Greater Colibri (Without) Lo/Hi BP Damage = 586-916
Calculated Lo/Hi (using D+fSTR*M*PDIF) = 576-921 (PDIFMin=1, PDIFMax=1.6, fSTR=-1, D=49)
PDIFMin = 1 when:-
1.25<=Ratio<=1.5
PDIFMax = 1.6 when:-
1.3<=Ratio<=1.4

Lesser Colibri (Without) Lo/Hi BP Damage = 995-1347
Calculated Lo/Hi (using D+fSTR*M*PDIF) = 940-1411 (PDIFMin=1.6, PDIFMax=2.4, fSTR=0, D=49)
PDIFMin = 1.6 when:-
Ratio==2
PDIFMax = 2.4 when:-
Ratio==2
Note: Ratio caps at 2.0

This immediately tells us that Avatars have at least double that of Lesser Colibris defense. Or in other words, they have at least 466 attack. It should also be noted that this value gives GColibri a Ratio of 1.436, in other words, a max SD damage of 992. Bear in mind the PDIF calculations are not crystal clear, so the value for attack can only be taken as an estimate.

Now using the Lesser Colibri (with) test, we should be able to come up with an attack value correct for the pdif formulas employed.

Lesser Colibri (With) = 980-1488
Calculated Lo/Hi = 940-1411, fallen outside of bounds, raised to 2.55 modifier?
PDIF=1.6 when Ratio==2
PDIF=2.4

This suggests that the PDIF cap has been broken with Summoner's Pigaches. However, it could equally be due to the face that PDIF calculations are not 100% accurate. If the PDIF cap was broken, you would expect the Mountain Buster readings (with boots) to be off the charts. However, the Max there is very close to the Max without, suggesting the boots do not break the 2.4x PDIF cap (Normal Hit).

Instead, what we are seeing here is a shift in attack to the high end values. By using the mean of the Lesser Colibri tests, we should be able to see how much attack has been added:

Mean (Without) = 1130.775
Mean (With) = 1200.2

Knowing the PDIFMin and PDIFMax are 1.6 and 2.4, the AVERAGE PDIF should be 2.0 without the boots. This should rise with the boots, and the difference can tell us the attack boost.

PDIF necessary to create 1130.775 = 1.923086735
PDIF necessary to create 1200.2 = 2.041156463
Difference = 0.1180697283
Increase in Attack = 27.51

The Multiplier of Ratio

So for a Ratio of 2 (capped), a total of 27.51 attack was added.
For a Ratio of 1.4~, this attack bonus decreased rapidly until it was not noticeable. If the attack bonus was static at 27.51, the boots should have dealt on average 48.83 points of damage more. But they didn't. This leads me to believe that attack scales based on the value of ratio. The curve of best fit was as follows:

Attack Bonus = 7*Ratio^2

So for a ratio of 2, you will gain 28 attack.
For a ratio of 1, you will gain 7 attack.


Summary on "Enhances Avatar Attack" gear

- The effect only kicks in when using a blood pact. (which makes sense, since the attack bonus is dependant on having a defense value)
- The attack bonus changes based on the defense of the monster you are Blood Pacting.
- The overall attack boost can be calculated by: AttackBonus = 7 x (ATTACK/DEFENSE)^2
- The higher the monsters defense, the lower the attack bonus.
- The maximum boost this will give is 28.
____________________________
Kegsay - 85SMN/85BLU/85SCH
Goldsmithing-60, Cooking-61, Synergy-60
Shaded Specs Obtained 20 June 2005

http://kegsay.livejournal.com/
#2 Oct 11 2009 at 3:08 PM Rating: Default
Avatard
*****
11,630 posts
Someone inform SE of this bug! I don't want my attack boost to be shrunk because the other mob has high def, I want it it increase more so I can pass the defense to begin with!
#3 Oct 12 2009 at 11:35 AM Rating: Good
***
3,059 posts
kegsay wrote:
- The effect only kicks in when using a blood pact. (which makes sense, since the attack bonus is dependant on having a defense value)
I'd say that's poor wording. Normal hits still use the random values of pDIF made from the cRatio (avatars even crit too). There's a definate defense component. Whatever reason it only works on BPs, it most likely works the same way as skill and just processes when a BP is the current action.

I'm also wondering if it works the same way for Koschei Crackows and Evoker Gages. Probably does, but it wouldn't be the first time SE screwed up gear descriptions.
____________________________
"And I'm prepared to passionately argue this point until nothing makes sense anymore. If that doesn't work, then the hours upon hours of whining will." -Red Mage Statscowski

Saggo of Garuda Lakshmi
#4 Oct 19 2009 at 2:11 AM Rating: Decent
Scholar
**
697 posts
Enilanerda mentioned see'ing different attack values with different pieces of gear with 'Enhances Avatar Attack' although he never finished the test, I'm curious if you'll be able to use an alternate piece of gear to compare to your original and possible a third test see'ing if they stack?
____________________________
75 RDM BLM SCH SAM PUP
Reply To Thread

Colors Smileys Quote OriginalQuote Checked Help

 

Recent Visitors: 16 All times are in CDT
Anonymous Guests (16)