Accuracy mechanics

[Kaunitz]

Hi!

I've come upon another problem (surprise, surprise!). I'm trying to figure out how accuracy works. Many mods revert to "damage mechanics" when it comes to tweaking musket lethality. However, because I want "fatigue effetcs" to work, and they only influence "accuracy", not "damage", I solely want to stick to accuracy this time. Also for sound purposes (I'm tired of hearing soldiers scream when noone falls) it would be quite usefull. I've come up with a hypothesis:

Spoiler Alert, click show to read: 

Explanation: A bullet flies randomly any path within the cone which is determined by the factors shown. In this figure, I've ignored left-right-error. It only depicts up-down-errors, but left-right won't be that important if you use tighter formations as I do.

According to this hypothesis, you can set the importance of the distance between the units by altering the "missile distance for half chance hit", whereas the importance of units' accuracy would be hidden in the "calibration area": If you consider accuracy to be % (or rather absolute numbers, since there exist accuracies higher than 100) of base calibration target area, then the calibration area becomes smaller, thus the shot will be better aimed. However, I fear that this model is far from being 100% correct.

An alternative possibility: maybe units' accuracy simply increases the "missile distance for half chance hit" (the distance between the muzzle and the bullys eye), and doesnt affect the calibration area (Darthvader somewhere said that a the smaller the calibration area, the better a unit can make use of its base accuracy)?

Another variable, the range of the weapon, might also intervene somehow.

Another question is the scale of the values. E.g. does the height of the target or the calibration area refer to the same scale as weapons' range?

If you ignore units accuracy and left-right-errors, and if you take the same scale for all variables, you can even draw a sketch by means of which it is possible to calculate the hit-chances. E.g., with a calibration area of 20 at 60m plus an infantry-target height of "1", I've come up with:
25m: 12,5% (a unit of 130 men in three ranks, firing in two ranks: 11 hits)
46m: 6,5% (5-6 hits)
70m: 4,3% (3-4 hits)
90m: 3,5% (2-3 hits)


PS: The size of the target is determined by "unknown 5" and "height" in battle entities, not by "radious sphere". And of course, the target can become bigger/smaller due to uneven terrain.

It would be great if modders could tell us their experiences with "accuracy"! I'm very thankful for any help I can get! A good way is to slow down or pause the game, for then you can see the bullets' trajectory.

[p.jakub88]

Interesting theory Kaunitz, that sounds possible.

Yes, the accuracy issues is quite complicated in ETW.

I'm glad that someone else is interesting in this topic, because i personally think that accuracy levels should be changed in ETW, especially if you entered new projectiles ranges for weapons as i done...

For me this problem is a academical way of thinking - it should be simple - most hits on point blank range (to max 100 m) and almost nothing hit in 300 m.

Unfortunately it isn't that simple as it sounds....

Of course it is possible to change overall unit accuracy, but everyone knows that.....

Maybe weapons in ETW have different accuracy - i.e flintlock, matchlock ,etc.

Because i noticed that i.e line infantry is far more accurate than i.e citizenry ( of course they have different accuracy levels set in units_land tables....


I agree that, if anyone will find the answer for this question should post here his though....

[Kaunitz]

Well, I don't think that you can make weapons with different accuracies, since the respective variables (half chance hit, calibration in kv_rules) exist only once and apply to all projectiles. You can only differentiate between artillery (it has seperate lines) and projectiles.

As far as I know, you can only simulate the effect of different weapons by tweaking the damage (base damage and damage mode (ap or effective range based) in projectile-table, damage distance multiplier and long/extreme range attackpower in kv_rules) and the accuracy of the unit bearing the weapon (but like I said, I don't know how accuracy interacts with calibration and half chance hit). You could also try to increase/decrease the velocity of the projectile in the projectile-table. I guess that slower projectiles have a curved trajectory and will have difficulties to hit moving targets, but I've never tried it out. I've read somewhere that there's also a variable determining some kind of projectile-gravity.

PS: I've started some non-representative, not very systematically carried out tests lately.
calibration 20, half chance hit 45, accuracy 1, target height 1: hardly any hits, even at point blank range
calibration 20, half chance hit 45, accuracy 50, target height 1: 3-4 hits at maximum range and medium range (didn't come closer for my unit always routed *g*)
calibration 30, half chance hit 45, accuracy 50, target height 1: 1-3 hits at long range, about 6 hits at medium range (didn't come closer for the aforementioned reason *g*)

it's interesting to note that I feel that the first volley seems to be more effective (even though fatigue effects are turned off). Maybe it has to do with the fact that I increased the musket range and the game engine doesn't know what to do with the "surplus" range to the vanilla range? But it might be a placebo as well.

[Kaunitz]

Just, for fun, I've done a little video with my current settings, to test whether "shock effects" work.

Spoiler Alert, click show to read: 

(Make sure you watch at 480p)

Settings: calibration 35, half chance hit 70, target height 1, accuracy 50;
It also works okay at maximum range (two units shooting at maximum range for 13 minutes until they're out of ammunition (30 rounds) --> 1,3 hits per minute).

[Kaunitz]

After some further testing, I think that my hypothesis can be scrapped.
My new "impression" of how accuracy works:

The higher you set the calibration value, the stronger is the effect of accuracy (I found that out via fatigue effects on accuracy.)
The distance to half chance hit just seems to ehrm ..."calibrate" the accuracy over distance. E.g. with a very low distance for half chance hit, at maximum range, there won't be a difference between a high accuracy and a low accuracy unit, for the effect of accuracy decreases over distance from a certain spot (half chance hit) on. At maximum range with a low half chance to hit distance, the effect of units' accuracy would have already decreased to zero.

But of course, it's just an impression.

[p.jakub88]

No offense Kaunitz, but maybe ask Creative Assembly members how this accuracy mechanics works?

Otherwise you will spend your whole life to figure it out....