Test Update 8/5/2014 - Pet Changes Round Two

Sorry for not being able to post parses for several days, I will not elaborate on the why beyond simply saying I have not been available to post.

I won't bother posting the necro/beastlord parses from the first round of changes on Test as the variables have changed. Below are extended parses (> 20 minutes for trash mob) of the earth pet at EM15 vs a "normal creature" in current content. Keep in mind the aforementioned quote from Aristo about target goals.

EM15 Earth pet with certitude, all relevant aa, and burnout haste buff rk 2. No debuffs were used on the mob. Pet hold was on, no activatable or passive abilities from pet or owner - just raw mitigation.

DI Spread:



Round by Round:



Total hits: 1737 "hits"
Number of hits for DI 1: 775 "hits" for minimum damage

In this case of pet vs a yellow con (lv 101-102 mob), earth pet is currently seeing minimum DI hits 44.6% of the time. Please feel free to reference Aristo's goal vision for baseline pet mitigation.

Relevant changes between v1 and v2 changes on test server? Average DI spread is now slightly better. Average incoming dps is down from 7579dps to 7234 dps (just over 300 less incoming DPS thanks to a more favorable DI spread).

EM15 Earth pet - same conditions vs Roon:

DI Spread:


RBR


Total hits: 1339 "hits"
Number of hits for DI 1: 380 "hits" for minimum damage

In this case of pet vs named, not "normal creature" mob, earth pet is currently seeing minimum DI hits 28.4% of the time.

It would seem that the present rendition of the tank pet is in line with the Developers' quoted target goals.

Great parses Bashan, thanks for the work and your contributed time.

One thing to keep in mind that nobody's parses pick up reliably is the same problem with head to head comparison of warrior vs knight tanking power. Warriors have no self healing capabilities. Our discs are stronger and our baseline is smoother and more beefy when compared to knights of comparable gear. On paper, warriors in a head to head comparison of raw mitigation will look amazing next to similar knight parses.

The one thing? Self healing.

Knights have it in spades.
Pet owners have no lack of it, and the heals are now heads and shoulders better than where they were months ago.
Warriors have none of it.

What kind of heals per second is a knight capable of? With a debuffed (or frankly even without a debuffed) mob even in group gear it's strong enough that an external heal source (be it mercenary or player) isn't required for your average mundane creatures. While I'd bet dollars to doughnuts that pet classes (well ... maybe beastlords can hang up there) don't have quite the same heal capabilities - they certainly have enough to shift that average dps line down considerably. A pet owner chaining heals can cover how much of that difference in incoming dps? Quite a lot of it I'd wager for a lot of content. Named mobs are different given their higher actual hits and higher relative attack.

Once again, thank you for parsing for the community. It is very much appreciated.

Don't wager, it, why not test it? Especially, I'd love to see pets in a position that most tanks see the majority of their time grouping which is killing multiples. I'd love to see some data or hear some feedback from a pet class using a group EM-focus to tank multiple mobs and spam healing the pet, how long he can keep it up and vs how many, what levels. Granted, it's an environment that doesn't favor pets but it's probably more common than tanking singles

I have no idea how the warrior vs knight debates went but I'm glad someone brought up the idea of active defenses and it wasn't summarily dismissed as not worth including in comparisons as it has been in past debates between players and pets here

Heals per second is easy to calculate.

What is the base level of the heal?
What is the cast time?
Recast time?
Modifying aas (base heal, crit rates etc)

Voila - heal per second potential.

Having the most available healing aa and the fastest casting heals, the beastlord obviously wins this contest in raw heals per seconds. Fiddling around with my Mage doing parses, the cast time on heal is mid 2 second range and I was getting crits in the 50-60k range. Even mage healing potential isn't something to scoff at. If you get nothing but non-crit heals (20k) every 5 seconds (i'm intentionally low-balling here both on expected cast time and heal amount) - that effectively will work out to 4k heals per second ... 4k less effective dps your healer has to cover.

Necromancers get the short end of the stick on paper, but having spent 100s of hours pet tanking with my necro over the last 2 years I can't be more emphatic when I say ... do not discount the use of the cascading rune line of spells. With runes there is no 'overhealing'. It was and will remain one of my favorite tools in dealing with those 'oh snap' moments.

Group geared beastlord pet heal casts in 1 second and is critting for 84011 points of health.
Pet heal over time (lasts 1 minute) is 3k per tick and gives a heal when it wears off.

Testing him vs mobs without a merc now

Edit - just got a pet heal over 90k

That's what you seem to think is the only baseline to compare a pet.

1000 mana from a spell bar, 1.5 second recast, 1 reagent needed.

Survivability running 100% of the time... So, you zone into a neriak HA and pop Last Stand for the first 2 mob pull?

That's the best possible mitigation for a tank in the game. It's not something a pet should be compared to at all.

Just cause a defensive warrior can tank it, or make the parse look like nothing, doesn't mean a pet should.

**Fixing** pets will fix their broken unintended tanking ability.

Defend it with whatever cherry coating you want, the big daddy devs have made it known they also know pets were borked.

Saying it didn't hurt anything or whatever you guys wanna use to justify it doesn't change the fact that it was broken.

Broken things get fixed. Look at my epic. /points. I didn't post or get mad about that change one bit. It was broken. It got fixed. Time for you guys to adapt and move along.

Bst pet DI:



Bst pet RBR



Healing Stats


Spells Cast



You wanted a more realistic scenario. The mob was slowed with aa Sha's reprisal. No mercenary was summoned. I almost fell asleep healing him. I told pet to hold at 15% mob hp to extend the parse. I finished the fight at 98% mana.

Pet: EM15 Beastlord warder with certitude, Plakt's protection, Extraordinary Velocity, Spirit of Nak Rk. II

Pet equipped with basic mage summoned gear.


Keep in mind that under the old pet tanking model from months ago, I solo'd Roon over a 29 minute marathon fight. Also keep in mind that this was done prior to the massive increase in pet healing strength. Back then, the single target pet heal was reasonable, but nowhere near where it is now. This was only possible because pets never took a max hit and had very steady predictable incoming damage.

Salve of Blezon is now incredible - noncrit heals in the upper 30s to 45k range, crits highest I saw was just over 90k just now. Pair that with promised and the heal over time ... and it was quite boring. I had lots of down time during which casting dots, nukes, swarm pets would have had the fight over pretty quickly. Without the slow I would have cast more heals - but still with down time.

Would be nice to see decent heals, I wonder what healing-per-second the cotf mage reflect-heal averages out to?

I gotta say that it does baffle me that people think that the baseline to compare pets against is the best mitigating class in the game WITH raid gear ... and a class that has no self healing capabilities. At best, sorry ... hate to say it ... but at best pets should be compared to knights for raw mitigation and incomming dps in that pet owners are likewise capable of "self healing" in this sense.

But ... i'm just one guy and represent only one guys opinion

Any melee DPS class with sufficient healing yes, but not with a merc. I tried moloing HA#1 partisan last night (I was bored and no one else seemed inclined to do it), but inevitably died on the 3rd mob (a blue con). This is despite full raid buffs and using a shield, attempting to proc slows, and the healer on reactive. Alas reactive just isn't too good for the start of a new fight and that's the dangerous bit. There is no hope in hell of me getting remotely close on Roon (and neither do I care).

As for the effectiveness of the different types of pet I was just going on the data I saw here: https://forums.station.sony.com/eq/...-changes-round-two.212698/page-2#post-3111897 but clearly there is conflicting data if he was able to molo roon with water and you were not. I guess it's just marginal and needs luck.

I'm not a mage of course, but IMO the dps pets should trade more dps for less AC. The choice should be stark. Ideally there would be a nice gradual tradeoff between abilities, be they dps vs defense or vs utility. I'm not arguing against the earth pet being a good tank provided it doesn't beat real tanks at their job.

And tanks gearing requirements have always been too steep. And pet classes too easy. And it would be awesome if pet classes could customize/gear their pets with non-summoned customizeable items a la mercenary equipment and not one silly earring.

It would be really cool too if berzerkers got some kind of love prior to the next expansion. SK's aren't quite as useful as warriors/paladins on raids I hear

Separate issues though. All of them

The odd thing is necro pets on live seem be very close to the desired goals stated for this change.
EM20 Lev (& necro warrior pet max pet AA
NPC DPS 10787

Stix in 441s, 8435k @19128dps --- Soulhate 7059k @16008dps (83.69%) --- Siddar 1376k @3149dps (16.31%)

Tanking summary for: Siddar --- Total damage: 4713810 --- Avg hit: 8387 --- Swings: 733 --- Defended: 104 (14.2%) --- Hit: 562 (76.7%) --- Missed: 67 (9.1%) --- Accuracy: 89.3% --- Dodged: 18 (2.7%) --- Parried: 12 (1.7%) --- Blocked: 40 (5.5%) --- Riposted: 20 (2.9%) --- Absorbed: 14 (1.9%)

Siddar - vs - Stix
Attempts 687 100%
Missed 63 9.17%
Dodged 18 2.62%
Parried 11 1.6%
Blocked 40 5.82%
Riposted 19 2.77%
Absorbed 13 1.89%
3924 13 1.89%
4359 216 31.44%
4758 4 0.58%
5286 15 2.18%
5592 2 0.29%
6213 14 2.04%
6426 2 0.29%
7140 18 2.62%
7261 2 0.29%
8067 23 3.35%
8094 1 0.15%
8993 18 2.62%
9920 14 2.04%
10597 2 0.29%
10847 19 2.77%
11431 4 0.58%
11774 20 2.91%
12701 15 2.18%
13100 2 0.29%
13628 24 3.49%
13934 3 0.44%
14555 14 2.04%
15482 11 1.6%
15603 1 0.15%
16322 1 0.15%
16409 8 1.16%
17336 12 1.75%
18105 1 0.15%
18262 12 1.75%
18939 1 0.15%
19189 6 0.87%
19773 2 0.29%
20116 6 0.87%
21043 4 0.58%
21970 13 1.89%

Notice the DI 1 rate is 31.44% not the 80% to 90% while on Test DI 1 rate is roughly half that when DI 1 target is stated as 35-40%.

You need a "pure" parse with only 20 damage intervals in it. If your pet is doing various defensive abilities or you are mixing up multiple types of attack then it becomes very hard to know how many of those numbers need to be combined together to calculate the raw DI 1 figure.

That said, I do indeed feel sorry for necros this time, who seem to have had their pet thrown under the bus due to the overpowered AC of mage pets. However traditionally they always did have a weaker pet I guess (hence their old preference to fear kiting due to weak pets). I don't know how the two *should* stack up against each other, but I've never really viewed necro pets as overpowered!

Parses of a group geared SK tanking a trash mob without discs, compared to a raid focused (EM 20) mage earth pet:



Mage Earth pet receives 7536 incoming DPS.



Group geared SK (gear link) receives 4502 incoming DPS.

Mage pet was attacking and receiving rune procs. SK was self buffed only, no discs.

There is no perfect way to parse this having only 20 DI in the parse mean you'er purposely not using things that you would always have making your parse completely detached from in game facts.

I'll repeat you can not parse what is being discussed in this thread correctly every method is flawed based upon the fact that all 2-5 minute parses are flawed by there duration and few trash are named mob in group content lasts longer then that.

You put pet on guard here and back it off that is how you parse passive mitigation then the npc lasts 10 minutes or 4 hours .

Also, use /log to toggle your logs OFF prior to engaging the mob. Once you engage, click off any Vie spells then turn your log back on.

Not sure how to handle innate pet rune procs. Perhaps pet classes can chime in to assist.

This parse genuinely confuses me ... as in it does not compute whatsoever.

Here is an extended (close to 30 minute parse) of a group geared sk (magelo in my sig below). No disciplines were used, shield was equipped.

RBR


How is it that your parse is so catastrophically different than mine? Average dps was over 7500 with more spikes and larger spikes to boot. Are you certain you did not have any cleric vies, discs et al used? Was the mob debuffed? Was it slowed?

For the record, my shadowknight has capped mod2s, max aa, and 13 more raw ac than your shadowknight.

Like I said ... it literally makes no sense. Granted my parse was over 1800 seconds and yours was 1/6th the length ...

Why would i want to parse passive mitigation? Have you not heard repeatedly from people that the above is not a valid measure of pet or player tanking its only a small peace of a larger puzzle.

Do not buff your pet with the proc buffs.
Have spell on hold.
Put pet hold on.
Have pet briefly engage to set aggro on pet then tell pet to back off.
Take of cleric vies prior to engage.

Doing this you get a clean 20 DI spread and take the reflective block buff out of the equation.

Pro-tip ... for whatever reason healing a pet does not generate threat. Having your pet autoattack for 3 seconds ... you can afk for an hour or more letting mercs heal and they will never snag aggro.