Are aimed shots underpowered?

[Talonos]

So, I was doing some math to try and figure out how to balance some of my mod's weapons against the original, and I found something kind of strange. The average damage dealt per time unit is far, far lower for aimed shots than it is for snap shots and auto shots.

I define "average damage dealt per time unit" as "accuracy * damage * number of shots / %time units." I understand that accuracy isn't really your chance of hitting, but some sort of modifier that defines the possible angle of your shot, but I figure that at some distance (particularly far away) the listed accuracy is pretty close to being your actual chances of hitting. Your *actual* chances of hitting improve the closer you get to your target. Correct?

Soldier skill is ignored, because it's just a multiplier on accuracy, and if you multiply both the snap and aimed shots by the same number, the ratio between them will stay the same.

So, with that in mind, some examples:

Code: [Select]

        Shot:   Damage: Acc:    # att   Avg.Dam/att  TUs        Av. Dam/100% TU
Rifle
Aimed 30 1.1 1 33      0.8 41.3
Snap 30 0.6 1 18      0.25 72.0
Laser Pistol
Aimed 46 0.68 1 31.28      0.55 56.9
Snap 46 0.4 1 18.4      0.2 92.0
Laser Rifle
Aimed 60 1 1 60      0.5 120.0
Snap 60 0.65 1 39      0.25 156.0
Heavy Plasma
Aimed 115 1.1 1 126.5      0.6 210.8
Snap 115 0.75 1 86.25      0.3 287.5
In every case, the snap shot outperforms the aimed shot in terms of average damage per TU spent. This holds true for every weapon in the game, not just the four examples I list above. And at closer ranges, where your actual chance of hitting is far higher than the accuracy listed on your weapon, the snap shot only gets better!

Why is this? Is the snap shot supposed to be balanced against the aimed shot because it uses fewer bullets? Then why are the ammoless weapons like the laser still lopsided? When making my mod, I'm considering balancing the snap shot against the aimed shot by buffing the aimed shot or nerfing the snap shot, but before I start making design decisions based on math, I want to make sure I fully understand why math works out the way it does.

I'm sure with all the analysis that takes place here, somebody has noticed this before. Could they explain it to me?

[Shoes]

I would like the think that the inceased accuracy offsets the TU. If I could spend 200% TUs for a 100% aimed shot, I wouldn't use it all the time, but I certainly would use it when I needed a good shot. Damage output isn't the only thing to take into consideration!

[yrizoud]

Aimed shots can be justified despite their mathematical inferiority:

• - When it's important to avoiding hitting a wrong target. This includes :
  
  • - Friendly fire on XCOM agents or precious alien loot
  • - avoid having your rocket explode in transit to a far target,
  • - avoid destroying lots of wooden walls, since they give shooting openings for the aliens inside
• - when an alien has opportunity for reaction fire : To be "sure" that your first shot hits and kills/cripples the target.

I've tested quite a few tu/precision settings for a sniper rifle. There are two factors that make precise-but-slow shots weaker:
- even a "successful" shot has a minor deviation that can make it miss, especially at long range. So there's nearly no difference between 97% final accuracy and 160%
- the 0-200%  randomization of damage can reduce your successful shot to zero damage (a glancing hit). Hitting multiple times evens out the random.

[Shoes]

It would be interesting to have an "aimed" show that reduces the range of the random damage. A sniper rifle that hits for 50-150% damage.

[Quixote]

Related topic: I find that often, I will try to hit with an aimed shot first. It will miss. Then I will use the rest of the TU for a snap shot. It will hit.

I wonder if the accuracy for aimed and snap shots are actually mixed up? Or is it just selection bias from me?

[Solarius Scorch]

Aimed shots also use less ammo than auto shots

[Talonos]

Yeah, my math puts auto shots roughly 1.5 to 2 times as good per TU. (Except for the auto cannon, which has an auto dealing exatly three times as damaging per TU than the snap shot.) This makes sense. You get higher damage yields, at a cost of more ammo. Except for laser weapons. Those, you might as well full auto all the time, because anything else is at least half as effective. I try to keep the damage/tu for auto shots twice as high as a snap shot when making weapons.

In my mod, should I just stick with convention and have aimed shots deal less damage per TU than a snap shot? Or should I try to "fix" it? I'm leaning towards the first, because I assume there's something else in the math I'm missing, but...

[yrizoud]

Your "Av. Dam/100% TU" assumes the remaining TU carry over from one turn to the next, which is not the case
A more accurate representation of total expected damage is by enumerating the different combinations of shots that you can actually take in a full turn spent shooting : For example with a weapon that has 30%snap and 60%aimed, compare:
scenario 1 : 3 snaps (10% leftover TU)
scenario 2 : 1 aimed + 1 snap (10% leftover TU)

[Incognito]

In one of those game editors I once adjusted the aimed shot TU cost down and I think the Auto shot TU up abit, I did start to use aimed alot more.