A) We all know that when one pulls a meleeing mob, it will run to within a prescribed range of the puller and begin to attack. If, once the mob is in range and attacking, however, the puller moves forward the mob will turn 180 degrees, walk backward, and turn to face the puller - this usually occurs when the puller has hit feet within the mobs selection circle. It should also be mentioned, that while the mob is shown visually as turning, it continues to execute frontal attacks as if it had never turned. Equally once a mob is in range and attacking, and the puller walks backward, the mob will move forward to maintain range. There is a small margin between of perhaps three yards before the range is modified, meaning that you can take one step forward, or one step back and there is no effect.

B) The size of the mob's model makes a difference, such that if u put your feet on Onyxia's selection circle and alternate strafe right and strafe left, Onyxia's model will flick left and right rapidly. If you're ten yards back from Onyxia and do the same, she'll only move slightly. This principle is seen most effectively on Lethon, where having your tank stand on Lethon's selection circle and run through him, turns him much more rapidly than having your tank stand yards back and run 180 degrees around him.

All of the above is based upon the puller and mob being on the same level; and by that i mean at the same elevation; on flat ground: no hills, no ramps, no steps. But my question is exactly that, what happens if puller and mob are not on the same elevation?

First question, where is range calculated from? From the middle of the model? From the base of the model? For the purposes of this thread, I'm going to assume from the base. Now lets assume the puller is stood 3 yards away the mob. Let's also assume that the puller, a Tauren warrior is 6 yards tall. The mob, a dwarf rogue is 2 yards tall. If range is calculated using the base of each model, then their range remains three yards, because if range was caluculated from the middle of their models then their effective range would be 3.5 yards, taking the height differential into account. Agree?

Second question, and assuming still that range is calculated from the base of the model. If the puller is stood on the top of a 45 degree ramp, and the mob at the bottom of that same ramp then despite the fact that there is a distance of three yards between the bases of their models 'as the crow flies', the elevation of the ramp means that there is in fact 3.5 yards between them. Agree?

Third question, if one accepts that the distance between puller and mob needs to exceed three yards before the mob will move forward (section A), and if one accepts that the z-axis affects range (question 2), do you agree that a puller, tanking a mob on a slope will cause the mob to move closer to the puller, and thus be more reactive to the movement of the puller (section B).

Why I am asking? Mainly out of curiosity, but also in relation to a discussion that sparked up over Instructor Razuvious and tanking Raz on the ramp. It's not that important, but if anyone knows, or is interested, would appreicate your thoughts.

Just to clarify, do you mean the vertical? As it's usually referred to as the Y axis.

I believe it does count, but my only testing for this is jumping off the cliff near the Gold Mine in Arathi Basin on my warrior and charging as I 'fall down' onto a target running by, I generally have to be fairly far down to be able to charge / intercept. This could also be the effect of my jump being forward and down rather than down. I have jumped direclty over targets, and had to charge 'backwards' on myself, from the button again only lights up half way down.

Could probably test it fairly easily, with a co-operative player on the other team, or finding a different cliff to leap from over somthing chargeable :)

Edit:

I believe the 'distance' is calculated from the edge of the models hit box, rather than the center of the model itself.

If you are working in two-dimensional space, otherwise, Z is accurate to indicate the vertical in relationship to position within three-dimensional space.

As far as I know, range is straight 3-D pythagorean. One example is MCing stuff outside UBRS. You can jump it off the ledge and MC will break even though the X,Y is directly below you.

From my experiences, I'd say that the Z axis most definitely affects range. There have been many times when, standing up on a hill, I have been out of the range of something that would have been pretty much right near me if we had been on the same X-Y plane. Think back to the ZF temple stairs, for example.

As to distance calculations, I would imagine that they are done from some set point within the bounding box of the character, probably the "center of gravity" as that seems like the most logical place.

Well, I know that I've jumped off of very high cliffs and then used Charge to avoid taking fall damage. Charge doesn't become available until I'm nearly at the ground, and my descent movement tends to be more vertical than horizontal.

I would very much expect that range is simply the distance of the line between your anchor point and the mob's anchor point.

I'm going to go off on a limb here and theorize that this would tend to be supported by the observation that extremely large mobs tend to have anchor points lowish to the ground, which puts them into melee range (To observe this, pull up Ragnaros, or Onyxia, or any other extremely large mob in a model viewer. The camera anchor point - which is also used for distance calculations and LOS calculations - is not on their head, as you might expect.)

I'm definitely sure Z-axis affects range, but the interaction between melee range and the Z is strange. I can think of many instances of me meleeing a mob down a sheer cliff or something, or mobs doing the same.

Stand on the outer side of one of the ramp 'rails' in the first DM north courtyard and aggro some dogs or Mol'dar for a good example.

What does that have to do with anything? And even then, a measly boar will turn around at the same rate as Lethon will if you run around a certain distance fromthe middle of its selection circle: its just Lethon's body is larger, so the angular velocity at the outside edge has to be larger (and thus he's perceived to flip 180 degrees around at blazing speed), but it still takes the same amount of time to turn him around.

So, the question is: is our 'range-box' a basic sphere or an elongated one. I would incline towards a sphere and as long as your 'range-sphere' for the ability/spell collides with the target's 'hit-sphere' then you're considered within range.

Hit boxes are cylinders, as with most games, and range is calculated from the surface of these cylinders. It's the easiest way to do semi-realistic hit detection without over-taxing the system.

Any Horde who has shot at critters while waiting for a zep in Durotar can attest to the z-axis counting for range.

For anecdotal proof of hit boxes being cylinders, when we do Roman'khan for the quest, I tank him from standing on the highest point of the bones near him, which places me above even his center of mass. A hunter below pulls him to me and I taunt him off the hunter.

For proof that distance is calculated from the surface of the cylinder on the x-y plane, you can take any large mob (Onyxia for example) and see if a warrior can intercept while in the hitbox and well outside minimum range from the center of mass.

On the z-axes, you can use the Horde zep tower in Durotar, southern edge. You can shoot the boars from a bit of an angle, but adders have to be right below you to be in range.

Vertical distance affects range ... for players.

Getting Impaled by Anub in no way hinders the attacks of Anub or Corpse Scarabs (which has caused quite a few wipes)