Units Heights

[Danial]

The height of an X-COM soldier is 22, right?

 

Well, if you use trigonometry, the distance between the centre of two adjacent tiles equals 22.63.

 

Now, as we all know, UFO rounds down to integers, so that the 22.63 would become 22.

 

Are we seeing the link here?

 

Anyways, when you double the distance between two objects, their visual size halves.

 

So... (finally I hear you all say)

 

My idea is that UFO uses the ratio of 22 height to 22 distance as its 100% accuracy reference and works all calculations out from there!

 

So if a 22-tall unit stands 2 squares away, their "distance accuracy" value would be 50%. Alternatively, if a 10-tall unit (Silacoid) was standing 1 square away, they would have a DA value of 45%... and so on and so forth...

 

Sound reasonable?

 

Thoughts anyone?

[Danial]

On furthur thought, it could also be a distance of 22 (or 11) squares it works off...

 

I suppose the ultimate test would be to give a soldier a base-chance-to-hit of 100% and see at what distance he starts missing

[Bomb Bloke]

All I know is that I've seen enough crazy half-way-across-the-map-and-through-two-windows shots hit their targets.

 

If the ratio was as direct as you seem to suggest, the odds of hitting a unit more then, say, six tiles away would be minute. Unless I'm not understanding you at all, which is also a possibility.

 

But whatever you're talking about, I'm pretty sure you're onto something.

 

... It's been too long since I did trig.

[Danial]

All I know is that I've seen enough crazy half-way-across-the-map-and-through-two-windows shots hit their targets.   

Any accuracy has some chance of hitting its mark halfway across the map

 

Unless I'm not understanding you at all, which is also a possibility. 

All I'm saying is that I think they might have used 22-height @ 22-distance as their 100% accuracy reference and then scaled everything from there.

[Zombie]

Well, if you use trigonometry, the distance between the centre of two adjacent tiles equals 22.63.

 

Now, as we all know, UFO rounds down to integers, so that the 22.63 would become 22.

How did you arrive at the value of 22.63 Danial? I'm guessing that the game doesn't list how wide a tile is, so that means you had to make some assumptions. What are they?

 

Anyway, the theory sounds about right. For comparison, this is what the Official Strategy Guide has to say about ranged weapon modifiers (pg. 221):

 

Two more factors involved in determining firing accuracy are less easy to quantify. These factors are distance to the target and the target's size. Several random factors are involved in determining the effects of distance and target size on a soldier's firing accuracy, and all are figured into the final algorithm that determines whether a shot hits the target. Because these factors change from shot to shot, it is impossible to judge these effects accurately. To make a rough approximation, however, you can get relatively close by figuring that targets at twice the distance away are twice as hard to hit. In other words, a target 20 squares away has a distance modifier twice that of a target 10 squares away. Target size works the same way.

If you notice, I underlined the word "random". Taken in context (or read literally), the modifiers are only approximate as randomness plays a role somewhere. However, if anyone has the OSG for TFTD, randomness is not mentioned anywhere. This leads me to believe that the distance/size relationship is used as the primary method.

 

But wait, look closer at what the OSG for TFTD says:

 

Because these factors can change from shot to shot due to the target's posture and position, it is impossible to pinpoint their exact effects.

Hmmm. It seams that posture (crouching/standing) as well as position (aka orientation = greater chance to hit in a face-to-face encounter, lesser chance to hit if the target is facing sideways), play a role too. It makes sense though, as most of us already assumed these factors were involved anyway.

 

(Btw, the reason I am quoting the OSG for TFTD is because the wording is similar but explains some things better. In addition, the programmers probably kept the exact game physics which X-COM: EU used).

 

- Zombie

[NKF]

The way the tiles are drawn (I know, how it's drawn isn't necessarily how it's actually represented as actual game data), my guess is that the tiles are something like 32 x 32 in pixels in size if we were to take the aspect ratio away and look at them as if they were a flat tile.

 

No, actually, I think they're actually less than 32x32. They are indeed 32 pixels across - the hypotenuse. As it stands, we see it across in one direction at 32 pixels, and the other at 15.

 

Bah, I was never good with triangles...

 

- NKF

[Bomb Bloke]

Every now and then someone comes along and states that the tiles are most likely a square meter at the base, based on the idea that the tiles were six feet in height. I think one of those people who worked it out was me. Can't remember.

[Danial]

The way the tiles are drawn (I know, how it's drawn isn't necessarily how it's actually represented as actual game data), my guess is that the tiles are something like 32 x 32 in pixels in size if we were to take the aspect ratio away and look at them as if they were a flat tile.

 

No, actually, I think they're actually less than 32x32. They are indeed 32 pixels across - the hypotenuse. As it stands, we see it across in one direction at 32 pixels, and the other at 15.

 

Bah, I was never good with triangles...

 

- NKF

<{POST_SNAPBACK}>

Well the hypotenuse is 32px, which makes the sides 22.6... No assumptins there, just math

[Bomb Bloke]

Ah, that's right. So to state it plainly, the units used are pixils.

 

Now, a soldiers' height by my memory is indeed 22 'units'. But, if my earlier note about a tile being a meter square is correct, and if a tile has a width of 22 units, if those units are of an equal base value then... The soldiers are a meter tall.

 

Which sounds ridiculous, but if you follow the line of logic that these units are used purely for generating percentages, then I suppose there could be a point there.

[Danial]

There's never been a conclusive argument that 1 tile is 1m... They could be 6ft wide for all we know

 

The reason I thought it might have something to do with 22 tiles, is because someone mentioned in another thread that that was the limit of a unit's sight...

 

All these 22's surely have to have some logical link... Don't they?

[Bomb Bloke]

What the measurements are in real-life terms doesn't really matter. I've just measured a tile to be about 25 pixils high, I'm fairly sure that supports your arguements in some way, but I'm not even sure I'm disagreeing with you about anything and I don't feel like doing much maths right now.

 

If 22 squares is the distance you can see, then the accuracy stats you listed before can't be right. 100% at one tile away, 50% at two... It'd drop down way too quickly at that rate.

 

Regardless, a unit can't have 0% accuaracy at 22 tiles, or else you wouldn't be able to hit anything on the other side of the map. 22 tiles isn't much more then an Avenger and a half. There must be a base accuracy or something... I dunno. You've hit on something, but my brain has turned into a sort of mental paste...

 

I need to program more often, or I'll start to miss the mental exercise school gave me. Argh! I'm missing school! Someone, get me my head doctor!

 

Where was I... Ah, yes, accuracy tests. Bump a unit up to 100 accuracy points, then give him a gun with the same accuracy, and see what happens. Those gun modifiers are important, I've seen troops with 255 accuracy miss.

[Danial]

You're definitely right. The accuracy modifier at 22 tiles would be 0.0001% (1 in 1,000,000) which is just plain stupid

 

Hmmm...

[Hobbes]

There's never been a conclusive argument that 1 tile is 1m... They could be 6ft wide for all we know

 

The reason I thought it might have something to do with 22 tiles, is because someone mentioned in another thread that that was the limit of a unit's sight...

 

All these 22's surely have to have some logical link... Don't they?

<{POST_SNAPBACK}>

The vision limit is 21 tiles (22 if you count the tile where the soldier is standing). And there's another thing I remembered: what about the aliens units, if there's a relationship between height and accuracy? I have a table listing Sectoids with a height of 16, Cyberdisks with 15, and their accuracy doesn't seem to be that limited with long range shots.

 

Another thing I remembered: how big would be the difference between firing at targets in the vision arc (<21 tiles) and those outside it?

[Bomb Bloke]

Incedently, probably should have said this earlier, but those height stats aren't just used to calculate firing accuracy. They are also used to work out where the bullet is actually supposed to be aimed. Sure, you could potentialy use them to calculate it...

 

If a unit is kneeling, enemies have to aim lower to hit it, otherwise their shots will simply pass over the shorter units' head. Try it with your tanks, you can shoot right over them and hit taller enemies on the other side.

 

The shorter a unit is, the more likely the shot will hit the ground as opposed to just flying through the air until it hits an object. This is because the units aim from a high point to a relatively low point.

 

The chance to hit lowers because there is less target to hit.

 

Then you've got to throw in the terrain between the shooter and the target. Windows, for example, are 3D objects as far as the game is concerned. Thus, a shot which goes through a tile containing a window sill must miss the sill itself and go through the hole. A shot isn't aimed at a tile, it's aimed at a location within a tile!

 

Calculating a generic formula for a shot hitting, given the amount of different shapes of tiles and the different ways units can be placed, would be difficult to say the least. But, you may be close to working out how distance factors into the equations.