Gun accuracy nerfed?

[Alec_White]

when people discuss ‘aiming’ i feel like theres an ‘aim one turn’ option im missing. i know (f), (a), and ©. is there amd option to aim for one turn?

Do you mean dot (.) to aim for one turn?

[pisskop]

Probs. never knew.

[kevin.granade]

when people discuss ‘aiming’ i feel like theres an ‘aim one turn’ option im missing. i know (f), (a), and ©. is there amd option to aim for one turn?

As Alec pointed out, that’s ‘.’ by default. The overhaul I reversed disabled this feature, which might be why you didn’t notice it.

[busmumrik]

So why did gun accuracy get nerfed when guns were already mostly pointless?
Even if I have some skill and full steadiness I have a hard time hitting zombies that are a few tiles away.
This is both unfun and very unrealistic.

Will gun accuracy get changed back to how it was before? if not, can I do it myself with some mod or changing some number in the game files?

EDIT: I made a test on latest experimental, and this is much worse than expected.

I made a gangster (freeform points) with 10 markmanship and 10 pistol skill and 12 in all stats. With these skills he should’ve been one of the best gunslingers in the world.
The gangster starts with a glock19 with 9x19 JHPs. and for every shot I fired, I pressed . until steadiness was max.
But the hit change is extremely low if the target is some distance away.
Even at 4 tiles away I miss most of my shots, and good luck to hit any shots more than 10 tiles away.

This make no sense from a realistic viewpoint and it’s also very frustrating and unfun for the player

[leebash]

My current character, a mediaeval peasant could still kill a bunch of zombies with a battle rifle.

At the time, he was only level 1 in most of the gun related skills (which he learned from an NPC). He had 20 strength and 4 for every other stats.

But later I noticed he sometimes miss every shot even if he literally shoved the gun barrel in the face of the target, but could hit pretty good from a distance if I press 5 or . after each shot.

[pisskop]

im under the impression there is close quarters accuracy penalty for most non-shotguns

[PlasmaChroma]

Some really quick and dirty statistics I generated from running the in-game code.

Start @ Per 10 & w/Iron Sights in FN SCAR-H, attacking a standard zombie.

[Lesseps]

Could you please elaborate for dummies : what does that tell us?

[pisskop]

When aiming at max steadiness, your weapon skill means relatively little compared to improving your weapon and/or perception.
Even a rifling god cannot control dispersion.

where the lines intersect is the optimal firing zone. You draw an imaginary vertical line over the range of your target for whatever setup most closely matches you, and see how well you perform in A: (total hits) and B: (headshots/crits).

By itself its just a pretty graph, but you can see that reducing the Suck of your weapon will have the most effect on your performance.

[Miloch]

BRM (Basic Rifle Marksmenship) in the army trains on a 5 meter target. When qualifying you hit from 5 to 300 meter targets. The 300 meter target is difficult for a new soldier to hit but not impossible. Granted a basic training soldier has some training before they actually try to qualify. Also, these are from the prone supported position. I have no idea what it would require for a standing position. The most we do is kneeling. Which I would not try to hit a 300 meter target unless I’d been doing it a lot. 5 meter targets I don’t even bother using the sighting for with a rifle from any position. These targets are human sized. The closest ones only show shoulders and above.
How far would a tile be IRL?

[pisskop]

I recently heard 1.5meters; that seems to be a fairly constant guesstimation too.

This last measurement was based on road length and the idea that a humie is 1m x 1m wide.

P:

but, butbutbut, I hasten to add: distance in any grid game is going to be relative, because it is impractical to represent proper distances in a game of such large scales as 1 monster per tile. Tank drones are easily as large as beetles (car) in even the most conservative of guesses, for instance, and a road lane takes 30+ seconds to traverse.

Its kind of like DF in a lot of ways, since I find it easier to illustrate. An infinite number of critters fit onto one tile, but how much water is 5/7 of water?

5/7 water is 5 tiles worth of 1/7 water, which evaporates in a few hours. But dwarves can drown in it, and fish will air-drown in 2/7 water, which is theoretically enough to hold anything dwarf-sized and swimming.

[Miloch]

Sounds good. Well I’m going to use your 1.5m for a realism estimate then. Three tiles should be an easy as hell shot from a stationary/steady position. Three tiles being 4.5m. My general thoughts would be anything under this would receive a massive bonus to hit until the first melee attack is received and then that can throw things off. Also, steadying your aim should be to a greater affect at first with it reducing as your get steadier and steadier. It’s easier to attain a basic position then it is to slowly concentrate your aim and calm your breathing for a really steady shot.

[Rysith]

Sounds good. Well I’m going to use your 1.5m for a realism estimate then. Three tiles should be an easy as hell shot from a stationary/steady position. Three tiles being 4.5m. My general thoughts would be anything under this would receive a massive bonus to hit until the first melee attack is received and then that can throw things off. Also, steadying your aim should be to a greater affect at first with it reducing as your get steadier and steadier. It’s easier to attain a basic position then it is to slowly concentrate your aim and calm your breathing for a really steady shot.

I wonder if part of it is not having the proper balance between the factors that affect hit calculation (which, ATM, seem to be skill, perception, steadying time, and gunsights)? It seems reasonable to me that they would interact as follows to produce ‘confidence’, the distribution of likely hits/misses:

• Steadying time has the highest direct impact on confidence: spend longer aiming, tracking, etc., have a better chance to hit. There could be a small penalty for each tile that the target moved[1]
• Skill affects steadiness gain per time, with only a minor direct confidence impact. Someone with pistols 0 will take longer to brace, get a sight picture, etc. , but even with pistols 10 a hip shot is still a hip shot. Someone with pistols 10 can quickly brace and aim, though.
• Perception, and maybe Dex (or the lowest of, or something) have a minor impact on confidence. Maybe at low steadying levels DX has a dominant role but as steadiness increases PER becomes dominant
• Gunsights (and gun/ammo range) could work as caps on maximum effective steadiness - a scoped gun can get more effective steadiness than an iron-sights gun.
• I like the existing generally-inverse relationship between steadiness gain per time and maximum effective steadiness - a laser sight acquires targets much faster than a scope, but the scope is much better for overall accuracy.

As proposed sanity checks:

• guns 0 without steadying effectively cannot hit. Wild, unskilled fire should be ineffective
• guns 10 without steadying shouldn’t be able to hit often outside of 5 tiles, regardless of weapon
• guns 5 should make shots ‘effective’, meaning a 75% chance of hitting at mid-steadiness, at around the following ranges (modulo gun differences):
• pistols: 5-7 tiles
• SMGs and shotguns: 10-15 tiles (shotguns have raw damage, SMGs have automatic fire and common ammunition)
• rifles: 20-25 tiles
• guns 10 with a single ‘.’ of aiming should be able to reliably hit within ‘close range’, 5-7 tiles
• guns 10 with mid-steadiness should give effective ranges of ~50% more than guns 5
• guns 10, with scope, a decent rifle, and max steadiness, should have a 75% hit range of at least 60 tiles.

[1] And it makes no sense right now that steadiness is retained between targets, particularly in it’s manifestations with low-recoil single shot weapons like bows. I can spend 30 seconds steadying on an approaching zed and then put perfectly-aimed arrows into things until I have to move or run out of targets.

[Coolthulhu]

I just tested it with 8/8/8/8 character with 10 in all skills and precisely shooting a 0 dodge target (debug monster) with glock 19 with 9mm JHP at 3 tiles of range dealt about as much damage as whacking it with said gun.
At point blank, the gun dealt about as much damage as fists, though a tiny bit faster.

Now, 10 in unarmed and melee means world champion in martial arts, but 10 in gun skills is supposed not be too far from Simo Häyhä.

[NuG]

I was wondering why I wasn’t hitting anything with my crossbow unless it was 3 tiles away, or less, recently. I was thinking about the size of things just the other day, it’s hard to guesstimate size in a game like this to some extent.

On one hand the tank drones being are on one tile, like has been said, but the regular military tanks are bigger than some houses. I just try to use my imagination a bit.

I have been lvl’ing up melee and cutting along with my crossbow skills. I was hoping to use some guns eventually, something with a scope for long range. I hope it will still be worthwhile to go for it, if not I’ll just have to use something else.

[Azrad]

Hard to tell with my current survivor, who already has rifles in 16 and marksmanship in 18 prior to the change. Using a Barrett rifle.

Tested it out by spawning 9 hulks from 25 tiles away, then 50, then 60.

25 it seems normal. Hulks dead before they could even try to hit my survivor. Further distances is the same, but missing shots happen, or only 10 damage happens, but generally shots do 60-120 damage at 50+ range, unlike 30 or less where 200 damage happens, likelier to happen as the hulks close in.

For some reason though, turrets are super evasive. Took like 20 misses at 60 tiles away before a bullet hit.

Oddly, removing my survivor’s skills didn’t seem to change things much. Tested it on one hulk with 0 marksmanship and rifle. Damage seems to be the same, and the only noticeable change was doing precise aim took longer.

[Zilenan91]

Marksmanship vastly decreases the time it takes to aim as well as the accuracy a bit. Try taking the mods off your barrett as well as removing any ranged hit chance increasing bionics as a test.

[Firestorm_01]

Do we have issue on github about it?

[TooDAMNMuch]

5/7 water is 5 tiles worth of 1/7 water, which evaporates in a few hours. But dwarves can drown in it, and fish will air-drown in 2/7 water, which is theoretically enough to hold anything dwarf-sized and swimming.

not trying to off-topic, but DF water is spectacularly retarded, personally i assume it’s (number) to the power of (amount of water on the tile).

unless the dwarves are microscopic and gigantic at the same time, it’s the only way that makes sense and it still only barely does and at times, not consistently.

[Firestorm_01]

Ok, now we have:

Issue: Gun accuracy calculation broken