[4.18 Ground Combat] Grenade louncher scatter is still wild

[gG-Unknown]

soldier miss 20 tiles, shoot under his own feet to explode himself

Chriss, would you  be so kind and  reveal mechanics of scatter  ?

table, formula whatever. For grenades, grenade lunchers, and reguler direct  fire.

Thanks

Edited July 21, 2024 by gG-Unknown

[Kouki]

On 7/21/2024 at 5:08 PM, gG-Unknown said:

soldier miss 20 tiles, shoot under his own feet to explode himself

Chriss, would you  be so kind and  reveal mechanics of scatter  ?

table, formula whatever. For grenades, grenade lunchers, and reguler direct  fire.

Thanks

Thanks for bringing this up! I'll bring this up to Chris

[Chris]

On 7/21/2024 at 10:08 AM, gG-Unknown said:

soldier miss 20 tiles, shoot under his own feet to explode himself

Chriss, would you  be so kind and  reveal mechanics of scatter  ?

table, formula whatever. For grenades, grenade lunchers, and reguler direct  fire.

Thanks

Thanks. I think I've already explained this once, but it's pretty straightforward - the game rolls to hit based on the hit chance shown on the cursor. If you hit, the shot goes where indicated. If you roll a miss, the shot will deviate so it doesn't hit the intended target. The maximum deviation is about 40 degrees in each direction, so it goes within an 80 degree arc from the front of the soldier. If that miss shot encounters a prop along the path, it rolls to see if it hits is based on the stopping chance of that prop - and if it hits it, then the grenade stops and explodes on the previous tile.

That's probably what's happened in the attached video, although I'll have to check a bit to see. I think it missed and scattered through the tree clump to the right, which it hit, so the grenade fell to the ground and exploded. If that's what happens, the bug is that the treeclump should have been discarded as adjacent cover and you shouldn't be able to hit it.

[gG-Unknown]

Another example of wild behaviour. No suspicious trees around, target about 15 tiles ahead, miss about  10 tiles.

 

[Bernie22405]

On 7/23/2024 at 8:30 AM, Chris said:

The maximum deviation is about 40 degrees in each direction, so it goes within an 80 degree arc from the front of the soldier.

An 80 degree arc is crazy wide IMHO.  I've trained people to use actual grenade launchers in the real world and a miss 40 degrees either side of an intended target would get someone retrained or disqualified from being issued that weapon.  Half that would be a lot more reasonable.

 

This is presuming that troops are "trained" on their weapons, which I realize the game does not portray.  

 

What if some minimum accuracy requirements were instituted for certain weapons?  Say 60% for HEVY or MG?  75% for sniper?  If the troop doesn't meet the threshold make the weapon greyed out and unselectable for them.

Edited July 26, 2024 by Bernie22405

[gG-Unknown]

another ridiculous  shot

 

 

Edited August 1, 2024 by gG-Unknown

[gG-Unknown]

and another one

 

[gG-Unknown]

On 7/23/2024 at 2:30 PM, Chris said:

The maximum deviation is about 40 degrees in each direction, so it goes within an 80 degree arc from the front of the soldier.

40 Degree is obviously wrong.

Would you  be so kind and   set 15 instead ?

To  prevent creating ultimate weapon, make grenades dmg fade based on range from central tile. Surprisingly dmg fade by range is very realistic,  there are numerous test on interet about safe range for various grenades. Due to explosion goes 360 degree, dmg fade by speed of volume of a sphere.

For game purpose dmg can fall off even   a bit slower like this :

central tile 45

range one tile 40 

range two ties 30

range three tiles 10

Same should be  done  with explosive throw-able grenades

Edited August 24, 2024 by gG-Unknown

[Xeferah]

I would like to add, perhaps it's already in the calculation, but to reduce the angle it misses with a bell (gauss) curve. So if you choose to do a 30 degree angle in both directions, and you split the 30 degree up in 6 sections, you would get a:
- 66% chance that the scatter is within 10 degrees both sides
- 29% chance that the scatter is within 10-20 degrees on both sides
- 5% chance that the scatter is more than 20 degrees on both sides

Or, written differently:
- 2,5% chance to scatter more than 20 degrees to the left
- 14,5% chance to scatter between 20 and 10 degrees to the left
- 33% chance to scatter between 10 and 0 degrees to the left
- 33% chance to scatter between 10 and 0 degrees to the right
- 14,5% chance to scatter between 20 and 10 degrees to the right
- 2,5% chance to scatter more than 20 degrees to the right

30 degrees here is chose and an example just because it's easy to divide 2 x 30 by 6.

I have no idea if this kind of calculation is already used, but this one would make the most sense and feels the best in the game.

Edited August 21, 2024 by Xeferah

[Skitso]

If I imagine myself shooting with a grenade launcher, I'd probably manage to point it quite well in the right direction, but the vertical angle (distance) would probably provide to be a lot more difficult to estimate. So what I would propose would be to decrease the lateral deviation to maybe 10-15 degrees per direction and make it deviate more in way of distance (too short or too far)

Edited August 22, 2024 by Skitso

[gG-Unknown]

15 hours ago, Skitso said:

If I imagine myself shooting with a grenade launcher, I'd probably manage to point it quite well in the right direction, but the shooting angle (distance) would probably provide to be a lot more difficult to estimate. So what I would propose would be to decrease the lateral deviation to maybe 10-15 degrees per direction and make it deviate more in way of distance (too short or too far)

Based on Chriss iinfo, and collection  of misses, the miss scatter is created as trigonometry  function.

40 degree to the left OR to thee right are two triangles.

https://www.carbidedepot.com/formulas-trigright.asp

Range of fire is "side b" ; angle A is up to 40 ; then volume of miss  is eguals to "side a"

When you get  lenght of  "side a" you  random  generate an angle, then send grenade to the new location.

I think, that for game simplification we dont need to  apply  elliptic miss area, it is much more easier to work with an circle.

Anyway, if you try  to  push your  idea further,  it would be  wise allso add  some form of feedback how the miss are looks like. Perhaps add another UI element - possible miss area.

[gG-Unknown]

16 hours ago, Xeferah said:

 if you choose to do a 30 degree angle in both directions,

If  you  choose use a trigonometry, (I gave link in  my previous  post,  now waiting for approval  ), then calculate volume  of miss for max range for  30 degree, you will see it is nonsense.

Edited August 22, 2024 by gG-Unknown

[Skitso]

It also looks funny when a missed shot hits an obstacle. Currently the arc ends neatly and perfectly to the obstacle, looking like it was aimed there.

[gG-Unknown]

22 hours ago, Skitso said:

It also looks funny when a missed shot hits an obstacle. Currently the arc ends neatly and perfectly to the obstacle, looking like it was aimed there.

Yes, the information feedback what COULD be hit is missing. Unfortunately, I  have no iidea how to  make it, without complete UI mess.

This is a side reason, why  I would rather keep miss area low, therfore accidental hits caused by different trajectory is low (easier to guess)

Edited August 23, 2024 by gG-Unknown

[Chris]

I'm going to try reducing the possible miss arc from 80 degrees total (40 on each side) down to 50 degrees. I'm very reluctant to reduce it further than that because there needs to be enough scatter that miss shots actually miss most of the time, otherwise the HEVY is going to be hugely overpowered.

There are still some bugs with the HEVY we need to fix, but they're not particularly simple and they'll be done in Milestone 5 hopefully. Those include the grenade sometimes scattering behind the shooter at short range, and the bug where if you're corner-peeking with the HEVY that miss shots can sometimes hit the corner (which is why most of those "soldier blowing themselves up" situations occur). 

[gG-Unknown]

23 minutes ago, Chris said:

I'm going to try reducing the possible miss arc from 80 degrees total (40 on each side) down to 50 degrees. / at least something.

I'm very reluctant to reduce it further than that because there needs to be enough scatter that miss shots actually miss most of the time, otherwise the HEVY is going to be hugely overpowered. / I see your worries. I remember, that in some previous posts  you said, that  due to chances, you might hit the original targeted spot even if you roll the  miss chance. Perhaps change the formula for counting  miss position to always miss the central tile. e.i. if you  roll miss chance then you never hit  the target.

 

Over power feeling comes from fact, that you make dmg even if you miss. This can be solved by fade out dmg from central tile. Fading dmg for grenades (in general) also helps reduce  insane efficiency of throwable grenades at groups of opponents. Human player, tends to place soldier close together to get firepower advantage at one spot. In case AI  use grenade, and kill several soldiers at once it means re-load. In case  any grenades use fade out  dmg, then game is fluent, less reload, less frustration from "sudden wipe" or "Unfair AI" feeling.

Edited August 27, 2024 by gG-Unknown