Page 1 of 1

Movement optimization

Posted: Thu Jan 26, 2017 5:19 am
by cryptc
Does anyone know how the movement skill divides into blue and yellow moves?

Seems to me that it's halved and divided into the two, but that means when it's an odd number it must either be rounded to blue or yellow. So it would be useful to know which way it rounds, since I consider one extra blue movement to be superior to one extra of yellow (I hardly ever sprint to the edge), I could remove one piece of gear if that nets me an extra blue square (or add a piece of gear if I realize that the bonus movement I get is just a yellow one).

Does anyone know?

Re: Movement optimization

Posted: Thu Jan 26, 2017 6:05 am
by dodger
should be easy enough to test.

Re: Movement optimization

Posted: Thu Jan 26, 2017 5:57 pm
by cryptc
So I tested a bit, and it seems there's something I don't understand with the calculation, because it's not behaving in a way I can fully predict

I had a 15 movement char be 8+8 fully loaded, and a 17 be 9+9, which seems it's halved rounded up for both movements

But then 17 with one less equipment turned into 10+10, and with two less it turned into 10+11

16 with smg was 10+11, and with smg and remove one piece of gear it was 11+11

So seems like I'll just have to note characters movement and experiment with each one unless someone knows how it's calculated.

Re: Movement optimization

Posted: Thu Jan 26, 2017 8:51 pm
by fowlJ
Mobility is equal to the distance a soldier can move in a single action expressed in metres, with a single tile in the game being 1.5 metres to a side. So a soldier with 12 Mobility (15 Base - 3 Utility Items) can move 12 metres, or 12 / 1.5 = 8 tiles.

A character can move up to twice their mobility at once when you dash, which occasionally results in an extra tile that would be lost to rounding if you had made two separate moves.

Re: Movement optimization

Posted: Fri Jan 27, 2017 1:11 am
by dodger
Yeah just tried this myself with a 17 mobility and an 11 mobility on a train map where I had a nice long open area I could run in a straight line. 11+11 or 22 dash for the 17 and 7+7 or a 14 dash for the 11 mobility. Matches with fowlJ's formula and indicates it takes more than .66666 to round up to the extra tile on the dash.

Re: Movement optimization

Posted: Fri Jan 27, 2017 6:31 am
by cryptc
That seems to work out correctly, so with modified movement (+3 from smg, -1 for each piece of gear, not sure about any other specials like cannons or grenade launcher), the squares will work out to be:

10: 6/7
11: 7/7
12: 8/8
13: 8/9
14: 9/9
15: 10/10
16: 10/11
17: 11/11
18: 12/12
19: 12/13
20: 13/13
21: 14/14
22: 14/15
23: 15/15

So if blue movements are most important, the movement numbers to aim for are: 12,14,15,17,18,20 and 23

Re: Movement optimization

Posted: Fri Jan 27, 2017 6:55 am
by Garthor
Soldiers can (and, in fact, often do) move diagonally, so looking solely at "how far can this soldier run in a cardinal direction" is going to give you a warped sense of the worth of mobility.

Re: Movement optimization

Posted: Fri Jan 27, 2017 7:05 am
by cryptc
isn't the diagonal movement just calculated using your movement as a circle though? there might be certain artifacts to the circle shape obviously, but that should be pretty minor, so I would assume the diagonal movement is at least comparable to the straight movement.

Re: Movement optimization

Posted: Fri Jan 27, 2017 7:29 am
by trihero
cryptc wrote:isn't the diagonal movement just calculated using your movement as a circle though? there might be certain artifacts to the circle shape obviously, but that should be pretty minor, so I would assume the diagonal movement is at least comparable to the straight movement.
Not sure what you mean by "as a circle," but diagonally moving is coded to use the pythagorean theorem, therefore 1 tile diagonal is really 1.41 movements.

Re: Movement optimization

Posted: Fri Jan 27, 2017 7:51 am
by cryptc
on a square by square basis yeah, but the total length becomes a diagonal of a triangle that isn't always of two equal lines, and the resulting paths form a circle of possible movements where the straight movement is the radius of the circle. It might be that other thresholds matter for specific points on the circle as it grows with increased movement, but I find it reasonable to think that the circle might be growing a lot just when your straight movement has increased by one.

Only way to test that would be having a lot of different movement soldiers on a very flat map and compare all directions in a 90 degree angle.

Re: Movement optimization

Posted: Sun Jan 29, 2017 9:08 am
by BookofAeons
I worked this out for EW here: http://www.reddit.com/r/Xcom/comments/3 ... nsensical/

It's identical in XCOM2 except the formula is now greater-than-or-equal, so the chart is inaccurate for straight-line movement of mobilities divisible by 3.

Re: Movement optimization

Posted: Mon Jan 30, 2017 5:51 pm
by dodger
BookofAeons wrote:I worked this out for EW here: http://www.reddit.com/r/Xcom/comments/3 ... nsensical/

It's identical in XCOM2 except the formula is now greater-than-or-equal, so the chart is inaccurate for straight-line movement of mobilities divisible by 3.
Can you de-math that just a bit? Are there certain mobility numbers that move more efficiently/effectively that we afford shot for ascryptc tried to list above?

Re: Movement optimization

Posted: Mon Jan 30, 2017 8:32 pm
by BookofAeons
dodger wrote:Can you de-math that just a bit? Are there certain mobility numbers that move more efficiently/effectively that we afford shot for ascryptc tried to list above?
TL;DR Every point of mobility provides greater blue movement in some direction. The chart at the bottom is accurate except for straight-line movement; that's the clearest way I could think of to convey the information but I'm open to suggestions.