Pox Science - The Arcane Natures of Dragging Effects, 2x2 Relocation, and 2x2 Unit Proximity

Before anything else, a huge thank you to @xFeyBlade and @themacca for helping with this stuff. It was a couple hours in what for them must have been boring custom games, but this couldn't have happened without their aid. @CelticxCross1110 also helped with some damage testing. On to the data.

Terminology:

North is towards the top of the screen. As the grid is tilted, the actual directions that 1 AP of movement will take a champion are North East (NE), North West (NW), South East (SE), and South West (SW). For any complex coordinates in relation to another unit, those will be given as a combination of the two shortest linear moves between those points, ignoring terrain. e.g. 2NW+4NE is at range 6, positioned 2 spaces NW and then from that point 4 spaces NE.

Testing:

2x2 Relocation
We first tested how exactly 2x2 units were handled when being relocated by a 1x1, with various obstacles to restrict positioning. For this, a Deep Elf Priestess relocated a Fire Giant. We were able to hypothesize two rules from all of our testing, which are used in order:

1. The relocated 2x2 champion will try to be as far North as possible, meaning that it will try to have its highest square as close to the top of the screen as space allows.

2. If the champion has two positions that are equally far North, it will chose the left position over the right one.

This set of rules seems like the most complicated relocation set, but those two rules should (I hope, see "Further Testing" at the bottom) actually explain all relocate: X abilities on champions of all sizes. EDIT: These rules are confirmed to apply to a 2x2 unit relocating a 1x1 unit. All that is left is to see how a 2x2 unit relocating a 2x2 unit works. These should prove universal for all practical scenarios - the only extreme case is Mimic-ing a relocate ability onto a 2x2, or Cyclops Hauler relocating some of the 2x2 relics.

I've attached a couple visual examples to show how this functions with a single obstacle.



Single Square Dragging Effects
Unfortunately, single square drags such as from Dragging Attack do not follow the same rules as the knockback chart. This has repercussions on how Dragging attack plays out if a player is on the top or bottom of an engagement, much like relocate effects. For instance, notice that there is no way to pull SE unless the drag source is directly in-line. Additionally, a player in a NW relative position has an almost impossible task if they want the opponent to go anywhere other than NW. The easiest way to show this one is to look at the attached diagram, but there is one additional rule that governs how the pulled unit interacts with obstacles.

1. If a unit is getting dragged from a source that is not in a straight line, and the square that the dragged unit would move into is occupied, it will instead "slide" and move into the next most reasonable square. This applies no matter what the obstacle is.

The above rule is - apologies - strangely worded, but the gist is that if a 1x1 unit is getting pulled and has a partial obstacle, it will still move, just into a different space. If, however, the drag is blocked entirely, it won't occur.



Proximity and 2x2 Proximity
"The nearest unit" sounds easy enough, but can be more confusing than that. Fortunately for 1x1 units, this is pretty straight forward. Here is our hypothesis for how proximity to a 1x1 unit works.

1. When two units are equidistant from a 1x1 unit that has an associated "nearest" effect, the unit first in the deployment order will receive the effect.

In our testing, we used a 1x1 unit with Bloodtracker which on-deploy was adjacent to two opposing units. When first deployed, the Bloodtracker unit hit the upper left opposing champion that was deployed first. When the position of the two opponents was swapped, it hit the bottom right unit that was deployed first. CONFIRMED: Using Starshine, the first deployed unit was always the one considered most proximate. This was not the case, however, if it was a 2x2 unit: even if it was deployed first, a 2x2 with its nearest square at the same distance from another 1x1 would not gain majestic. A 2x2 has to have its farthest square at the same distance as the nearest 2x2 in order to gain the benefit. (Deploy order in that instance is still indeterminate.) [Further testing needed]

2x2 proximity is, unfortunately, a nightmare, and we still don't have it down. We did, however, get some useful data about a few cases. Here is a list of important or potentially important elements. This testing was done by using Repurpose on a Fire Giant.

1. There is a vertical preference; runes farther south are more likely to be tagged.

2. There is a horizontal preference of some sort. This may either be connected to proximity to the entirety of the champ, or a general left-right valuation as with relocate.

3. Most importantly, the square of the 2x2 unit upon which the spell is cast has a dramatic effect: it counts at the center of the proximity check.

Those rules give a frustratingly incomplete description of 2x2 and general proximity rules, but should help as a starting point to work from.

1x1 Manic
Easy stuff. Two rules:

1. Manic units follow the KB chart for movement. (courtesy of Linque)



2. When two opposing units are equidistant, the earlier deployment is counted as both nearer for movement and the attack.

To make the movement quick and easy, place the unit on the KB chart and move it with the flow of the arrows. Also, take note that manic doesn't slide like dragging does: if there is an obstruction on any space that a manic unit would enter, it stops. Also, like pull does, if one move of manic moves a unit into a space with a different directional command, the manic unit will turn.

Pull (Ability)
Pull follows the same directional rules as single-space drags, with a few additional scenarios.

A) A unit will not slide if a single square in its path is obstructed. It will be pulled over the obstruction if possible.
B) A unit will still slide if a final place of occupation is covered. e.g. If the third spot that a pulled unit would occupy is obstructed and the pull is not linear, then the pulled unit will slide.
C) If multiple squares along a unit's pull path are obstructed, it will slide along the length of all obstructions. This is strange to say, so reference the diagram.



Deep Wounds

Some damage testing was done around Deep Wounds using Grundlewulf. By using Berzerk Attack, Pounce, and basic attacks, we sussed out the ordering and rules on the damage calculation.

1. The added damage from Wounds is rounded down. If a unit is attacked by an opponent with 13 damage while afflicted by 1 stack of wounds, they will receive 19 damage, not 20.

2. Basic attacks are the only things that apply a stack of wounds. Berzerk Attack and Pounce do not.

3. Both Berzerk Attack and Pounce, however, benefit from stacks of Deep Wounds. e.g. a Grundlewulf with banner that Pounces on a 0 Def unit afflicted by 1 stack of wounds will inflict 21 damage.

4. The Deep Wounds damage multiplier is calculated before defense (and because of how ordering works, one would assume before Resistance and Tough as well). This means that if the above Grundlewulf attacks a 3 defense unit with a stack of wounds, the Wulf will deal 18 damage, not 16.

Fascinate

Fascinate follows the KB chart. The easiest way to track Fascinate is to think of it as providing Manic to all opposing units within the radius, sans the attack.

Further Testing:

NOTE: A bevy of incidental observations and additional commentary indicate that the top square of a 2x2 is in some cases considered the primary point, but not in all cases. Can confirm that it works at least for the targeting of manic by a 1x1 unit. Unfortunately, there are still some strange cases of 2x2 proximity showing up.

Need to test 2x2 manic.

Need to confirm wandering. CONFIRMED: KB chart with a chance to do otherwise (specifics on chances and on which of the movements it can happen still needed).

I hope people have found this helpful, will update it as I finish more work.

 

I would love a code cleaning for more logical interactions.

 

The stuff is actually pretty consistent - the real issue with 2x2 proximity is that there are several entangled preferences that are tricky to sort out. Once those rules are broken down, however, putting them back together should be a cinch.

That said, the dragging mechanics are bonkers.

 

I mean, more predictable by any player at any time.

I would start with a graphical AoE that indicates directions of knockback and dragging.

 

Oh, yeah, maybe that. This is certainly not that.

But hey, we know it now at least. Baby steps.

 

Devs should have made charts like there long time ago.
Gj any way

 

Over a thousand hours of gameplay and I still didn't know this. Thankyou.

The poor souls that are going to get swallowed whole

 

Non targeted locating and relocating (ping, bloodtracker, manic) defaults to the top square of the 2x2 champ. Didn't read it all but seemed there was some confusion that could have been solved by that.

 

Not specifically any confusion on that, just something I have suspected but not yet verified. Most of the other material here has no connection to that.

And actually, Bloodtracker only sort of works like that. The test units were 2x2s, and if the top square was all that mattered, the number 2 deploy would have been hit in the first trial.

 

What'd be really neat is if the GUI implemented a system whereby when you moused over the target after clicking the ability, if there's a relocation element to the ability a little arrow indicating the result would appear.

 

Post updated with more information on Pull, Manic, and a bit more on 2x2 relocation. A bit of testing done with Starshine, nothing much conclusive. Will post more as I finish various tests.

 

the problem i find with relocate 2x2 is trying to get it in front of the relocator perhaps they can make a position placement like DD
also what is wrong with ironfist glider reckless flight random movement isn't as effective as skeeczick sniper

 

Nice , i knew all the relocate rules , but manic was a little shady to me , as sometimes if enemy champion is 1 space away , but your manic champ is adjacent to a friendly champion , it will simply harmlessly bounce off the friendly rather then moving the 1 space towards the enemy

 

@kalasle Sorry about my ignorance, but where is is this Knockback Chart? Also, I wonder about wandering too.

 

@Agirgis1 Glad to help. Also, will update the first post that manic doesn't slide, unlike dragging effects.
@Lushiris I'll put wandering on my list. I would guess it functions like manic, but with the shrine as the target point. Still worth making sure. I've put the old KB chart up in the first post as well under the manic section.

 

The real offender is top/left player getting screwed with demon beckoned appearing on the left further from your opponent, while the bottom/right player gets the demon, appearing on the left again, closer to his/her foe.

 

Wait the last map is according to the g'hren throwers or to the voil bombers? If it the view of voil bombers the NW arrows should be pointing the opposite direction?
If the one in the Voil in middle has maniac it should move NE+ NW am i right? IF i understood the theory correctly

 

This is amazing, you're amazing

 

No, that is not the real offender of beckon demon.

 

The map there shows in what direction Voil Bomber (the one around whom the map is centered) would move if a knockback effect were placed on a certain square. So, for instance, if an Avalanche was placed directly on the unit, the Bomber would move 1 SE because there is a SE arrow on that position.

If, for example, G'Hern Punisher had Manic, its first set of movement would be 2SE. Its next movement would be 1SE+1NE. I can make another diagram if that would be more clear.

Also, although these mechanics would appear to provide a large advantage to one player or the other, much of it depends on context. For instance, a NW player has a big advantage when using relocate enemy, because they get 2 extra spaces of pull. Certainly different, but much of the importance of that difference depends on specific contexts.