Necromancers - a buffed pet's melee toughness

This article is Redh_armor's VN posting (with a follow-up) of how a Necro pet's melee effectiveness improves as it's buffed, including what stats have what/how-much effect.

I will attempt to pull together the results of several recent tests and posts into a single, cleaned-up post.

Here goes

-----------

Some conclusions can be drawn:
1) The abomination can be modelled as a cloth caster with very high armor factor and no base absorption
2) Damage is modified by con buff, dex buff, AF buff, melee resist, and ABS buff
3) Each of these result in factors that are combined by multiplication to yield the total damage result
4) Con and dex buffs currently have identical effect on reducing damage (very odd)
5) Abomination damage scales with 1/Armor_factor
6) Melee resist works as for players
7) ABS_buff works as for players and receives the 25% bonus of other spec spells
8) Correlation between the test data and the model below is quite good, given crude measurements and small samples

Test cases

Case 1) Abomination summoned by naked necro
Case 2) +baseline buff (delve 36)
Case 3) +baseline buff (delve 44)
Case 4) +baseline buff (delve 44) + spec buff (delve 45)
Case 5) +baseline buff (delve 44) + spec buff (delve 57)
Case 6) +baseline buff (delve 44) + spec buff (delve 57) +40HP, + 26% crush
Case 7) +baseline buff (delve 44) + spec buff (delve 57) +40HP, + 26% crush + 9% ABS buff
Case 8) +baseline buff (delve 44) + spec buff (delve 57) +40HP, + 26% crush + 9% ABS buff +22% ToA buff bonus +22%ToA heal bonus
Case 9) Abomination summoned by necro with nearly maxed gear except for 0% crush resistance
Case 10) Case 9 +baseline AF buff (delve 52) + spec AF buff (delve 55) + 22% ToA bonus
Case 11) Case 9 + baseline dex (48 delve) + spec dex/qui (63 delve) + 22% ToA bonus
Case 12) Case 10 + Case 11 + baseline con (delve 44) + spec str/con (delve 57) + ToA bonus
Case 13) Case 12 + 26% crush + 9% ABS buff

Observations
1) The method used to estimate HP is ok, but a bit crude.
There could easily be a +/- 1% error in the %life estimate, so at least a 20HP uncertainty
Considering this, the scatter in the HP estimates is quite good.

2) The cleric con buff has no impact on abomination HP

buff >>> apparent HP
0 >>>>> 1321
45 >>>> 1333
55 >>>> 1323
101 >>> 1325
126 >>> 1324

The total HP for a abomination would be about
1320 >>> base HP
150 >>>> due to 75 constitution on gear
200 >>>> due to 200 HP on gear
1670 >>> total

3) The cleric con buff certainly does reduce damage taken

buff >>> average >>> min
0 >>>>> 44.4 >>> 35
45 >>>> 39.6 >>> 29
55 >>>> 38.4 >>> 29
101 >>> 33.8 >>> 25
126 >>> 31.6 >>> 24

4) The cleric dex buff certainly does reduce damage taken

buff >>> average >>> min
0 >>>>> 44.6 >>> 34
168 >>> 25.5 >>> 20

5) The cleric AF buff certainly does reduce damage taken

buff >>> average >>> min
0 >>>>> 44.6 >>> 34
161 >>> 37.6 >>> 28

The AF bonus observed when Voodoodaddy was buffed by the AF buffs were
52 delve base >>> 83
55 delve spec >>> 78
base + spec >>>>> 161

This is interesting because it shows that at the current time AF is treated like a stat and:

AF buff is affected by the %buff bonus on ToA gear

From tank testing weapon damage scales with 1/AF

If we assume the abomination has high AF cloth, then its absorption is 0, and the
AF_buff = 161*(1+abs) = 161
so
44.6 = base_damage*base_AF/base_AF
37.6 = base_damage*base_AF/(base_AF + 161)
and
base_AF = 865

This is a very high AF, but who says this has to make sense relative to players?

The effect of AF buff on damage would then be

AF_abs_factor = 865/(865+AF_buff)

6) So how does this all add up?

As was show in an earlier post there are probably several forms of equations that can match the data.

Here is one of them:

Damage = Base_dmg * Con_abs_factor * Dex_abs_factor * AF_abs_factor * Resist_abs_factor * ABS_buff_abs_factor

Base_dmg will vary depending on attacker and weapon
Con_abs_factor = 100%-con_buff/4
Dex_abs_factor = 100%-dex_buff/4
AF_abs_factor = 895 / (895 + AF_buff)
Resist_abs_factor = 100% - melee_resist
ABS_buff_abs_factor = 100% - ABS_buff

Notes
con_buff
> the total value of the constitution gain from con and str/con buffs. It does not appear to have a cap
> this is affected by cleric spec level, MoA bonus, ToA buff bonus
dex_buff
> the total value of the dexterity gain from dex and dex/qui buffs. It does not appear to have a cap
> this is affected by cleric spec level (baseline), MoA bonus, ToA buff bonus
AF_buff
> the total gain in AF that a cloth caster receives from an armor factor buff
> this is affected by ToA buff bonus
> Not sure if it is affected by cleric spec level (baseline) and MoA bonus
melee_resist
> the necro's resistance to the attacker's damage type. It is capped base on necro level (I.e., 26 max)
ABS_buff
> the delve value of the buff times 1.25. That is (3.75%, 7.5%, and 11.25%) for the (3%, 6%, and 9%) buffs

7) Please show your results

case >>> Test dmg >>> Model dmg >>> error
1 >>>> 44.4 >>> 44.4 >>>> 0.0%
2 >>>> 39.6 >>> 39.4 >>>> 0.5%
3 >>>> 38.4 >>> 38.3 >>>> 0.2%
4 >>>> 33.8 >>> 33.2 >>>> 1.8%
5 >>>> 31.6 >>> 30.4 >>>> 3.7%
6 >>>> 21.8 >>> 22.5 >>> -3.3%
7 >>>> 19.8 >>> 20.0 >>> -0.9%
8 >>>> 17.1 >>> 18.0 >>> -5.3%
9 >>>> 44.6 >>> 44.4 >>>> 0.4%
10 >>> 37.6 >>> 37.4 >>>> 0.4%
11 >>> 25.5 >>> 25.8 >>> -1.0%
12 >>> 13.9 >>> 13.4 >>>> 3.5%
13 >>>> 8.8 >>>> 8.8 >>> -0.1%

Not a bad match

Method to estimate HP
Gladto the pounds on the abomination until it is badly injured.
(this is done from behind to avoid wasting time with parry)
She then cast 3 level 46 heals and I note the total HP healed.
Since I am using a modified ToA interface the pet's %life is reported in Gladto's status window.
I note the %life before and after the 3 heal spells

--------------
Additional data

Results

Case 1 +0 con
22 hits
44.4 average
53 max
35 min

1321 Apparent HP
1149 Healed at end of trial
10% Initial life
97% Final life

Case 2 +45 con
25 hits
39.6 average
49 max
29 min

1333 Apparent HP
800 Healed at end of trial
10% Initial life
70% Final life

Case 3 +55 con
25 hits
38.4 average
48 max
29 min

1323 Apparent HP
754 Healed at end of trial
10% Initial life
67% Final life

Case 4 +101 con
33 hits
33.8 average
40 max
25 min

1325 Apparent HP
742 Healed at end of trial
10% Initial life
66% Final life

Case 5 +126 con
31 hits
31.6 average
38 max
24 min

1324 Apparent HP
781 Healed at end of trial
11% Initial life
70% Final life

Case 6 +126 con, +40 hits, +26% crush resistance
41 hits
21.8 average
28 max
17 min

1337 Apparent HP
802 Healed at end of trial
10% Initial life
70% Final life

Case 7 +126 con, +40 hits, +26% crush resistance, 9% ABS buff
43 hits
19.8 average
24 max
15 min

1326 Apparent HP
809 Healed at end of trial
30% Initial life
91% Final life

Case 8 +153 con, +40 hits, +26% crush resistance, 9% ABS buff
55 hits
17.1 average
21 max
13 min

1335 Apparent HP
1001 Healed at end of trial
22% Initial life
97% Final life

Case 9 max gear, but no crush resistance
31 hits
44.6 average
54 max
34 min

Case 10 +161AF
28 hits
37.6 average
46 max
28 min

Case 11 +168 dex
32 hits
25.5 average
33 max
20 min

Case 12 161AF + 168dex +153 con,
44 hits
13.9 average
17 max
10 min

Case 13 161AF + 168dex +153 con, +26% crush resistance, 9% ABS buff
33 hits
8.8 average
11 max
7 min

[edit. a few mods in attempt to straight colums, and improve legibility]

Update:

Q: At first, you say "base_AF = 865". But then you say "AF_abs_factor = 895 / (895 + AF_buff)". Which one is accurate, 865 or 895?

A: 865 is correct
Sorry about that.