This is one of several interrelated testing threads I am starting in the upcoming weeks. Though they are done for Force, eventually some results or formulas derived from them should apply to VOOT. Things are just easier to test in Force. As I may slowly add data, please don't reply to any of these threads, possibly for a very long time or forever.
I'm going to attempt to numerically test and calculate VR body sizes. This particular test applies only to VRs in the standing animation. It is known that movement animations can change the effective hitbox of a VR's body in unexpected ways. But while standing, most VR hitboxes can roughly be imagined to conform to a vertical cylinder. Initial tests suggest that the hitbox is not obviously related to how the VR looks, for example the arms of Vox "Dan" do not appear to count as a physical part of his body, whether for collision detection with another VR's body, or for collision detection against an incoming projectile. Further body rubbing tests suggest that the cross-sectional area of this cylinder is not a perfect circle, but just some shape that is close to a circle.
The following tests take Fei-Yen VH and while facing them, rubs against various other VRs from the front, sides, and back, then checks the distance readouts after getting pushed out to the minimum stable distance. The resulting numbers should roughly be the sum of Fei-Yen's frontal length (distance from wherever the game considers to be the center of her body, to wherever the frontmost part of her body is) plus the frontal/left/right/back length of the other VR's cross-sectional area. Distance readouts in Force identically correspond to distance readouts in VOOT except that Force does not display a decimal point. The more data we gather, the more precisely we can reverse engineer the original VR sizes.
Fei-Yen VH into Vox "Dan": front 10, left 10, right 10, back 11
Fei-Yen VH into Temjin 747A: front 9, left 9, right 10, back 10
Fei-Yen VH into Fei-Yen VH: front 8~9, left 8~9, right 8~10, back 8~9
Fei-Yen into Fei-Yen is unstable because while standing still she wiggles her butt. The exact timing of the wiggle animation is random, based upon when she stops walking, so the distance changes depend on how in synch the two Fei-Yens are. The two Fei-Yens are not always facing the same direction here so let's test other orientations.
Fei-Yen VH into Fei-Yen VH (roughly facing the same direction): front 8~10, left 8~9, right 8~9, back 8~9
The front test should have been identical to the back test in this case, but it is not. My explanation is that the backward walking animation pushes out the other Fei-Yen more than the forward walking animation. One last test should be done.
Fei-Yen VH into Fei-Yen VH (facing opposite directions): front 8~9, left 8~9, right 8~10, back 9~11
My conclusion is that Fei-Yen has a true physical half-width of 4 for the front, sides, and back. However, the wiggling animation shifts her butt roughly a distance of 0.5 to the front and left, 1 to the right, and 1.5 to the back. This is reinforced by another test: take Fei-Yen VH and carefully check her close combat activation distance against another VR who stands still. It should be 99 exactly, but the shifting animation of her butt makes the double lock-on fluctuate on and off. (Note that the distance readout and yellow energy bars never fluctuate so those are apparently measured according to the position of her feet). At the outer limits of 100 the lock-on can become stable, which matches the idea that 0.5 is her butt's forward rocking distance. And at the inner limits of 98 the lock-on can become stable, which matches the idea that 1.5 is her butt's backward rocking distance.
I also did the following tests.
Vox "Dan" into Vox "Dan": front 11, left 12, right 12, back 12
Vox "Dan" into Fei-Yen VH: front 10, left 10, right 10, back 10