Pet push

[Skanda]

Like Kherryn said above, no pet proc except for Wind has a push build into it. Our pet doesn't have much more push then we do. When I solo I stand on the oppsite side from my pet and the mob stays put. Stands to reason then that bst push ~ pet push. I lay blame squarily on the shoulders of those people who do not position their pets.

[Coprolith]

Looks like my hunch was correct. The push from melee is simply a fixed amount per hit, and its the same for PCs and NPC, regardless of level and class. The constant only differs between 1H weapons and 2H weapons, and special attacks like kicks and bash

I did some tests in the arena, duelling my warder (SoSorsha, dpoc focused) against some chars on my other accounts. I used /pet hold then /pet sit instead of guard so he wouldnt push back, or walk back to his guarding position. The push therefore comes only from the attacker. I aligned the name of the pet with my main chat box, then took /locs before and after the fights. The push distance then follows from the initial (X1,Y1) and final (X2,Y2) as

push = sqrt[ (X2-X1)^2 + (Y2-Y1)^2 ]


I started with a series of PC vs warder:

1) L43 monk, dualwielding and doubleattacking, bare fists, no special attacks
After 804 s and 321 hits the warder was pushed a distance of 68.5, or 0.213/hit

2) L43 monk, doubleattacking, 2HB, no special attacks
After 417s and 85 hits the warder was pushed a distance of 36.1, or 0.424/hit

3) L43 monk, Flying kick only
After 621s and 44 hits the warder was pushed a distance of 38.1, or 0.867/hit

4) L65 Druid, single wielding and single-attacking, 1HS
after 919s and 173 hits the warder was pushed a distance of 37.0, or 0.214/hit

5) L65 Druid, single wielding and single-attacking, 1HS, 50% hasted
after 801s and 226 hits the warder was pushed a distance of 48.1, or 0.213/hit

All 1H-results are the same, regardless of level. Hasted delay, dw and da will up the push rate, that is, push per unit of time, but not the push per hit
Two-handed weapons have a push to them that is exactly twice that of a one-hander. Special attacks also have their own push/hit and in the case of Flying kick, its 4 times as large as a normal 1H'ed hit.


Next are two pet vs warder parses. I used mage earth pets of lower levels so that innate procs wouldnt disturb the push:

6) L44 mage earth pet, dualwielding and doubleattacking
after 695s and 221 hits, 13 bashes and 18 kicks the warder was pushed a distance of 63.1, or an average push of 0.250/hit

7) L57 earth pet, dualwielding and doubleattacking
after 500s and 210 hits, 17 bashes and 8 kicks the warder was pushed a distance of 59.0, or an average push of 0.251/hit

Same result for both pets again. The average push per hit is slighty higher then PC 1H'ed push, but that's entirely due to the pets special attacks which you can't shut off.

Finally i wanted to see how much our own warder is pushing. Unfortunately i can't use the same method as above because my mage doesnt have pet discipline yet. However i can measure the warder's push relative to the mage pet by letting them fight each other (using my toons to heal the pets to prolong the fight).
I duelled the L57 earth pet again and this time i let the warder fight back. As the fight progressed push went back and forth, with our warder slowly outpushing the mage pet. This is entirely due to the level difference however, the warder simply hits more often in the same amount of time.
In 288s the mage pet had 80 hits, 14 bashes and 7 kicks. The warder scored 166 hits and 23 bashes. By the end of the fight the warder had move a distance of only 7.6


So what does all this mean? Well if the push/hit is constant then the melee push rate of a PC or pet is simply:

push rate = #hits_per_unit_time * Constant = #swings_per_unit_time * hit_ratio * Constant
The constant is 0.213 for 1H, 0.426 for 2H, and higher for special attacks

Let's compare our warder to a dw'ing, da'ing PC under raid conditions:
- My unhasted warder swings roughly 1x per sec on average, excluding bashes. With 85% haste and a hitratio of 0.6, the push per sec from normal hits only is 1.85*0.6*0.213 = 0.236/s
-A PC dw'ing 2 20 base delay weapons at max haste swings his MH weapon (1+da_rate)/hasted_delay ~ 1.6/1.0s and his OH weapon dw_rate*(1+da_rate)/hasted_delay ~ 1.3/1.0s for a total #swings/s of 2.9. Using the same hitratio we get a push rate of 2.9*0.6*0.213=0.371/s.

That's more then 50% higher then our warder's push. I've excluded both the PCs and warders special attacks, but that makes little difference to the ratio of the push rates.
As has been pointed out, with the exception of Spirit of Wind (which no one uses on raids anyway), none of our warder's procs have a knockback component to them. A PC with proccing weapons that also have a knockback component can outpush a pet by a factor 2.

So it seems my earlier analysis of the problem with pet push was correct. Its simply a matter of the direction of the push, not the actual size of the push. Pets cant correct the direction of their push themselves as we do, so we have to do it for them. That requires only some skill and practice.

/hugs

[rorith]

Awesome info! Thank you so much for doing all the work involved in this! I've bookmarked this thread to give out to people who claim our pets push too much. ;)

[Dummkopf]

Coprolith strikes again. Thanks for your work, that is exactly the info i need to show my fellow beastlords, mages and some raidleaders about pet push.

[Tzuka]

Hot damn, Copro. You arent human!  :wink:

[Naib]

All bow before Coprolith!!!

Well done sir!

[Skanda]

I have a question about that Coprolith. What was the Str of the Monk and the Druid when you did this test? A tank in my guild said he thinks Str effects push (after seeing your results I doubt it). For that to be true the the Monk and Druid's Str would have had to have been the same during that test otherwise there should have been a difference in the push per hit.

[Toln]

Excellent work on the parses Coprolith. I have a necro as my main and the many parses you have done that for this forum are very impressive and shed light on so many "myths" that have existed:)  
Really wish we had you over at the necro forum, it's kind of like the 'wild wild west' over there if you get my drift..

Pets are usually blamed in my guild, and in most I bet, when push is causing problems, but unlike most here, I tended to agree with those casting blame on them from my experience.

There are, at times, melees who are not pushing the correct way, but I honestly swear that i've seen many many occassions where 10 pets on one side of a mob were pushing it towards the 20 melees on the other side. (also in groups, Ie 1 melee on one side, 1 pet on the other.. pet does the pushing)
I know posts like "I have a feeling it's this way.." aren't generally well perceived. In fact I tend to disregard most of them myself. But here may be a few other causes that may contribute to the push situation of pets vs melees that I would like your opinion on if possible.

Have you done any parses regarding hit rates of PCs vs Pets?
If it's marginally higher for a pet than a PC on most raid encounters, the higher hit rate of pets could result in a much greater push by them.

Also regarding special attacks, such as backstabs from rogues, slam, bash, and kick.
When my pet bashes a monster, I can visibly see it move, which likely indicates that it pushes it a considerable amount. Have you done any parses on the amount of push generated by special attacks from melees, such as that?

And the issue with our pets being natural duel wielders/double attackers at higher levels of the game. Beastlords in particular among other classes do not receive double attack without AAs, hence the number of hits we would perform are significantly less than our pet assuming identical weapons and haste.
Also weapon delay which could play a part in reducing the number of actual hits a player is making on a monster even though the push generated by each individual hit may be similar for both pets and melees.

On a final note, I'm no math whiz, but you mentioned something about push direction.. where as in a raid setting, a pet sent on a monster will push precisely the same direction unless repositioned while a melee character will generally do that on his own.
I understand that on 'paper' of your parse that the push rates were the same, but in the actual raid environment, could this factor contribute to pets actually causing more push than melees?

So basically..  :)  I'm wondering if you think that these factors could contribute to causing pet push to be greater than the average melee push in certain situations, even though your parse showed that in the tests you did that it wasn't the case.

[Coprolith]

I have a question about that Coprolith. What was the Str of the Monk and the Druid when you did this test? A tank in my guild said he thinks Str effects push (after seeing your results I doubt it).

Unless your guildmate has some parses to back up that claim I'd file that one right behind 'Cha affects the chance of Fear and Root breaking'.
If its just his 'feeling', that is his in-game perception, then his statement is worth jack crap, minus jack.
(Oh and if you must know, monk str was ~170, dru is ~120)

Have you done any parses regarding hit rates of PCs vs Pets?
If it's marginally higher for a pet than a PC on most raid encounters, the higher hit rate of pets could result in a much greater push by them.

Don't have to parse hit rates, they can be readily calculated from base delay, haste, dw rate, da rate and hitratio. All of these are known within a few %.
As i said in my prev post when the push/hit is constant then the melee push rate of a PC or pet is simply: push rate = #hits_per_unit_time * Constant. hit rate = #hits_per_unit_time
A class that has innate da and dw will outpush a pet by 50% easily. That means that a beastlord, who does not have innate DA but only an AA skill that gives him/her a 15% da rate instead of the innate 65%, pushes more or less the same amount as his warder, which is exactly what we observe when we are soloing and lining up the mob in-between ourself and the pet.

When my pet bashes a monster, I can visibly see it move, which likely indicates that it pushes it a considerable amount. Have you done any parses on the amount of push generated by special attacks from melees, such as that?

Yes, see test #3 in my prev post. Monks' Flying Kick generates an additional push that easily matches bashes from our pets. I haven't parsed all special attacks, there may be minor differences between kick, bash, backstab, slam and flying kick, but the point is all melees have such a special attack, so this negates the pets bashes

Also, don't give too much weight to when you see the mob move. UI updates are seldomly in synch, especially the animations. In my Flying kick test i saw the pets visibly move after a kick as well, but usually 1 to 2 seconds after the actual kick. Lag has a large influence on this.

On a final note, I'm no math whiz, but you mentioned something about push direction.. where as in a raid setting, a pet sent on a monster will push precisely the same direction unless repositioned while a melee character will generally do that on his own.
I understand that on 'paper' of your parse that the push rates were the same, but in the actual raid environment, could this factor contribute to pets actually causing more push than melees?

My first post in this thread answers that. Pet's don't push more, it just seems that way because they just can't correct for the direction of the push themselves, they just keep pushes the mob in the direction they are facing. A PC would take a few steps sideways to counter the push, and knows almost intuitively how to position himself. However a PC that doesn't pay attention to push, that simply hits autoattack and follows the mob around, is worse then a pet (since he doing essentially the same thing as the pet, but with a bigger push).
If the pet's owner pays attention and has a little practice in this matter, he can also reposition his pet to behave just like any other raid member and not cause any trouble.

In a raid, its virtually impossible to get all the individual pushes to cancel each other out exactly. It has to be corrected for constantly. Even the smallest net push will grow into a large net push if left unchecked. But everyone has to do their part. Any member of the raid that doesnt pay attention to push, be it a PC or a pet, forces the rest of the raid to do additional corrections for them. As pet owner, the responsibility of your pets push falls on your shoulders tho.

It's been awhile since my last raid, so im prolly hopelessly outdated on positioning strategies, but i remember my old guild sometimes used a 'circle strafe' method to keep the mob confined to a small area inside the circle (I MT a few times with my pally back then). This requires wide open spaces tho.
And the pets that started out in the mobs back were facing the front of the mob after 180 degrees circle strafe, so it was more of an ellipse then a circle  :D

/hugs

[Chackra]

Hard to get a tank to understand that he needs to let the mob come to him fully before ingaging as I need to run past to position pet if he isnt the one turning mob around.

At least pets have an excuse to be stupid.  I'm happy to admit that pet push is hard to perfectly control, but I have seen cases where a MoB is inching toward danger when the puller could have easily stepped back another 20 feet if he wanted to (or a hundred feet for that matter), and the Raid leader screams:  "Get rid of all pets from now on!"

Pets can make a major contribution to dps even if you use them randomly and stupidly.  If they are used properly, they can be an infinitely more valuable tool as expendable and easily replaced shock troops.   If a raid leader doesn't know how to utilize one of the best tools the game gives him, he has no more business making decisions than if he didn't know what the term "tank" meant.

[Iustus]

I do not have any emperical testing, but I have to say this is not my experience. I think you are missing something with the higher level pets.

I have regularly done exp with a full group at the zoneline in kodtaz. This is with several melee in the group, often including both a rogue and a monk.

As a paladin, I am very focused on positioning. I use stuns with a large push to help me position.

Additionally, I bot a cleric and when we have a cleric in the group, I melee with my cleric. This means I need to keep the position of the mob close to my bot, unless I move him around (which I do not want to do).

I am able to do this no problem.

However, if a beastlord joins us, or if I bot a beastlord on the 3rd computer, the push is noticably more. Even when the only melee are me, the clericbot, and a beastlord bot, the pushing is too much for me to handle.

Beastlord pets do push more than any player (except someone intentionally casting knockback spells). I think you need to go back and test with high level pets and players.

-I

[Tastian]

"Beastlord pets do push more than any player (except someone intentionally casting knockback spells)."

That's just plain wrong.  Ask most beastlords that solo with their pet in rogue position how far across the zone their pet pushes mobs.  I personally out push my pet as it is now and that's with an uber 15% double attack.  In fact I can have my pet in rogue position + cleric hammer and still hold the mob very still.  Only time I run into an issue (issue meaning mob actually moves enough so I reposition a little bit) is when the cleric hammer is hitting from the side or at an angle.

[Coprolith]

Beastlord pets do push more than any player

No, they don't.

A beastlord + his pet + your cleric bot meleeing combined are what's causing the problems for your paladin. Botted toons don't move unless you tell them to, but pets will always move towards and face the mob once he's pushed. If you replaced the BST with any other melee class, you'd still have more trouble countering the push, but at least the PCs won't continue pushing once the mob gets out of range, nor will they change the direction they are facing which is essential in understanding how push works.

But thanks for blaming it all on the pet. Your mentality is exactly the cause of the frustation of the pet classes. You quote an example involving some botted toons and this is proof that pets push more? That's supposed to completely negate the results of my parses?

I'll repeat it one more time: pets push less then double-attacking PCs, they just can't control the direction of their push.

And if you don't believe that because of your "experience", then parse it yourself instead of coming here to insult me. The ball is in your park, not mine.

[Eatbugs]

I do not have any emperical testing

As a paladin

So you have a) no evidence to support your assertion and b) no direct experience with the topic at hand.   I haven't a clue why you think this would contribute anything to the discussion.  (Not to mention that you obviously didn't read Coprolith's parsing post.)

[tkyn]

This is awesome Coprolith, thanks so much for your work on this.

Any chance you could parse out the amount of push that a proc from spirit of rellic does or does not produce? It would be nice to be able to once and for all squash the pushing myths completely.