Jump to content
Sign in to follow this  
ColonelSandersLite

realistic skill and precision values

Recommended Posts

I thought I'd start a thread with the goal bieng the most realistic skill and precision values we can get.

To find these, open your PlayerName.ArmAProfile file and find the following lines:

skillFriendly=xxxxxxxxx;

skillEnemy=xxxxxxxxx;

precisionFriendly=xxxxxxxxx;

precisionEnemy=xxxxxxxxx;

Note that there's a set of values for both dificulties.

Tried straight 1 values, and that's far from realistic. They spot you instantly from 500 yards away crawling in the bush at night without NVGs and put a single bullet through your head biggrin_o.gif.

Obviously, straight 0's will be wrong too.

The defaults:

skillFriendly=0.750000;

skillEnemy=0.600000;

precisionFriendly=0.750000;

precisionEnemy=0.600000;

skillFriendly=0.850000;

skillEnemy=0.850000;

precisionFriendly=0.850000;

precisionEnemy=0.850000;

These seem flat wrong to me in most situations.

I'm experimenting with the below and will be giving feedback later. Note that although the Northern democratic republic or whatever probably shouldn't have the same training level as the US army, I wish to keep it more generic so it will be fair no matter what's fighting what.

skillFriendly=0.50000;

skillEnemy=0.500000;

precisionFriendly=0.450000;

precisionEnemy=0.450000;

So, what values are y'all using and how do you find them?

I'm also curious as to how these numbers interact with the skill levels set for the guys in the map editor.

Edit: I accidently flipped a couple of values.

Share this post


Link to post
Share on other sites

It will change the distance that the AI spots you and how accurate their shots are i should think.

You can change this stuff in difficulty in the options menu...

But this is a more customizable way of doing it. But i wouldnt reccomend fiddling with this unless you know what you're doing.

Share this post


Link to post
Share on other sites
It will change the distance that the AI spots you and how accurate their shots are i should think.

You can change this stuff in difficulty in the options menu...

But this is a more customizable way of doing it. But i wouldnt reccomend fiddling with this unless you know what you're doing.

Its broke,

I have lowered the enemy slider all the way to 0 and the spotting and enemy accuracy is exactly the same.They are still dead on and spot immediately.

Share this post


Link to post
Share on other sites

Hi all

does anyone know if the modified values in PlayerName.ArmAProfile override the settings in mission editor (AI skill slide..)

Cheers

Share this post


Link to post
Share on other sites

Hi, what I am searching for is an intelligent (as good as AI can be) enemy but with a lower precision in shooting like it is now and sure I don't want to have bots with RADAR!

I think the slider in option menue cuts down both values at once, getting a dumb AI. Not very realistic.

Glad to see that the skill of AI is split up.

Share this post


Link to post
Share on other sites

"They are still dead on and spot immediately. "

I`d be ok with this, if my weap reaction movement times were a tad faster. Physically, I can react fast. But the weap onscreen doesn`t react to my movements as fast/smooth as I`d like.

Share this post


Link to post
Share on other sites

what makes you assume sarani troops would be less trained than US troops? i can even say the opposite as sarani troops are in a known environement, while US troops are actually in a foreign place.

Share this post


Link to post
Share on other sites
what makes you assume sarani troops would be less trained than US troops? i can even say the opposite as sarani troops are in a known environement, while US troops are actually in a foreign place.

Not to mention that if you play as OPFOR, skillEnemy would probably affect the US and vice versa.

Share this post


Link to post
Share on other sites

my settings:

skillFriendly=1.000000;

skillEnemy=1.000000;

precisionFriendly=0.650000;

precisionEnemy=0.650000;

Why? They shouldn't be blind, or have problems with spotting me. They shouldn't kill me every time they shoot though, that's why I set precision lower than 1. I'm gonna test what will happen when the settings are set above "1.0", like "1.5".

Share this post


Link to post
Share on other sites

I was playing with the numbers above, and they seem pretty good all around. I'm wanting to do some tweaking to them though. I'll get back with results soon enough.

@funnyguy:

For any semblance of balance whatsoever, the AI should have roughly balanced abilities with a human as far as spotting goes. I've got damn sharp eyes, but over a certain distance, things get pixelated, you have frame rate drops to contend with, a narrow FOV due to monitor limitations, etc. Not to mention that in combat, you don't have perfect awareness, ever. You have tunnel vision to contend with, mud/sweat/blood of your friends in your eyes, adrenaline, etc.

Spotting a stationary prone man in the bush 500 yards away at night after he fires a single shot and takes the top of your freinds head off with or without NVGs when you weren't even looking in that direction simply isn't realistic by any measure.

I found that on .65 precision (or thereabouts) the AI too often pegs you when you're attacking from an advantageous position with the element of supprise in the first shot. .45 is still pretty unforgiving. If you mess up, you're still pretty dead. Just sometimes takes an extra shot or two.

what makes you assume sarani troops would be less trained than US troops? i can even say the opposite as sarani troops are in a known environement, while US troops are actually in a foreign place.

Not to mention that if you play as OPFOR, skillEnemy would probably affect the US and vice versa.

I agree with Eda Mrcoch here and that's why on my settings they're the same for both sides. In order to be generic as possible, no matter what nation is fighting what nation (addon considerations) and who's side the player is on.

As to my comment that the Sarahni troops shouldn't be as skilled as US soldiers: I swear, some people just get all offended if someone mentions US military training in a positive light as compared to someone elses. It would be understandible in some way if you where from this nation, but since it's a work of fiction...

The reasons, if you know your military stuff, are:

Firstly Sarahni uses old soviet military doctrines which emphasize quantity over quality. In the words of, I believe, Joseph Stalin, "Sometimes quantity is a quality all of it's own." We're talking about individual training levels here, not the army as a whole.

Secondly, Sarahni isn't exactly a large place with tons of money to spend on the military (probably partly why they use soviet doctrines anyways). If they can only afford to give their soldiers X number of rounds to fire for training purposes then that's what they get. They can't afford to do large scale multinational combined arms excercises like the USA does. Hell, they're not even going to be able to afford enough rounds to keep their troops well versed in their heavy weapons.

Thirdly, Sarahni is a new pair of nations, and that's the whole premise of the game. Presumably, they where a territory of some eastern block country, and would have been garrisoned by foriegn troops.

I would pit any modern professional force from a large country against the Sarahni army man for man any day.

Whatever your opinions of the USA's national policies (I'll keep mine to myself as this is no place for politics), it's hard to argue that the US military isn't one of the finest forces of professional soldiers man for man on the globe.

If you dropped any infantry division/regiment from the USA, Germany, Canada, Great Brittian, etc. you would wipe out their whole military in a matter of days.

As to the Sarahni being in a known environment and the US in a foreign environment:

That's not how the game starts, the situation is reversed as north Sarahni invades the area under protection of the US army.

That's also a strategic, and in some cases, a tactical advantage. Not a training advantage.

Share this post


Link to post
Share on other sites

CSL here with just a quick post, since it's christmas and all that. During a little free time, I set about testing the skill and precision values in a more organized manner.

The figures posted above seem to work pretty well for close quarters fighting *but* (and in ofp there's always a but...):

With these values when using the default skill levels, as they are when you place the preconfigured groups (squad leader 0.466667, team leaders 0.333333, privates, 0.200000):

<table border="0" align="center" width="95%" cellpadding="0" cellspacing="0"><tr><td>Code Sample </td></tr><tr><td id="CODE">

skillFriendly=0.650000;

skillEnemy=0.650000;

precisionFriendly=0.750000;

precisionEnemy=0.750000;

The soldiers will detect and (and this is the important part) *identify* a standing walking enemy at about 300-350 meters, and their fire will be effective around 300 to 350 meters and closer. Since most armies (that I know of) do weapons qualifications at 300 yards/meters, I feel this is probably a fair approximation, when compared to what a generic soldier should probably be likely to do.

Of course, with increased skill levels in the editor, their fire and spotting is better, but it doesn't really seem to have all that much effect anymore.

The serious drawback of using these values is apparent when you want to simulate something along the lines of professional soldiers going up against untrained/poorly trained conscripts/rebels/angry mobs. If you're familiar with such things, it would be clear to you that these groups would need dramatically reduced values for precision, although their skill value shouldn't be dropped that much.

The reasons for these thoughts are two fold.

1: If you've ever paid attention to news reports of gunfights between untrained groups people, you would know that these engagements are mostly spray and pray close quarters fights with little or no casualties on either side until they get into personal combat range. As an example, if two rival gangs in a major city start shooting at each other, usually there's between 0 and 3 dead "gangsters" and 0 and 3 dead innocent bystanders before the police arrive and everyone scurries away. In fact, in most of these cases, the first couple of bullets are the shots that do the dammage to the opposing side. The rest just tend to effect non-combatants.

2: Precision determines weapons ability, and skill only determines their ability to spot targets.

Given that people with little or entirely without combat training don't have fire discipline, and sometimes don't even fully understand their weapons beyond "pulling the trigger makes it go boom, point the hole towards the person you want to kill", their precision values should be no match for the precision value of any average soldier from any army.

Since skill only determines spotting ability (and to some extent communications ability since a unit spoted for an individual is effectivly spotted for his whole squad), these should be roughly the same, regardless of training. While trained soldiers will undoubtedly have learned some things about spotting camoflaged units, etc, basic human vision and hearing doesn't change depending on training.

These values are also of questionable worth when night fighting...

Conclusions so far:

Just looking at this, it seems that the numbers immediatly above are a good starting point to customise it to your preference. But, I feel very strongly that mission makers should probably specify what their reccomended settings are in the readme, if nowhere else. A player with different values, will undoubtedly have wildly different results than the mission designer intended if the values are different enough. I honestly have some pretty mixed feelings about taking these settings out of the hands of the designers. It's nice on the level that you can adjust your game to your preference, but it's bad in that it leaves a *lot* of room for missions to break with different values. As an example, suppose that your mission calls for a platoon to cinematically ambush some guys from some trees. If the settings are set to a certain range, the ambush may not be effective at all, breaking the mission. If the settings are in another range, it's likely to be TOO effective, also breaking the mission.

As a generalisation though I would say that AI's react to well in close quarters fights, and not quite well enough at long distances under normal visibility conditions. Under reduced visibility conditions, the AI reacts to well in nearly all cases imho.

Share this post


Link to post
Share on other sites

I agree that the AI-soldiers get their aim on you to fast and that their shooting is to accurate. It is not realistic (or fun). They can see you through grass and bushes. The can see you in the dark and from very far away, even when you are prone, and they hit you in the head with their first shot. Not cool. This must be fixed.

To solve it I also have tried to adjust the skill and precision values, but as far as I could tell that had little effect.

My solution was to dePbo the config.bin so I could increase the dispersion-value of the enemies' weapons (AK74 and PK). As a solution this sucks a bit, because now it is no longer interesting to seize enemy weapons when I am out of ammunition (instead I loot my dead comrades). However I feel gameplay is now more balanced and realistic.

My 2 cents.

Share this post


Link to post
Share on other sites
My solution was to dePbo the config.bin so I could increase the dispersion-value of the enemies' weapons (AK74 and PK). As a solution this sucks a bit, because now it is no longer interesting to seize enemy weapons when I am out of ammunition (instead I loot my dead comrades). However I feel gameplay is now more balanced and realistic.

So there isnt an "AIdispersionvalue"?

Share this post


Link to post
Share on other sites

As an aside, how does the AI slider in the editor work, does it set the AI to a value unconnected to the game settings, or does it set the AI to a value WITHIN the game settings?

Example, if I set the game enemy skill to 0.75, then set a unit in the editor to 0.5, do I get an absolute AI value of 0.5, or do I get a value which is 0.5 of the game's setting of 0.75, which would amount to approximately 0.37?

Share this post


Link to post
Share on other sites
As an aside, how does the AI slider in the editor work, does it set the AI to a value unconnected to the game settings, or does it set the AI to a value WITHIN the game settings?

Example, if I set the game enemy skill to 0.75, then set a unit in the editor to 0.5, do I get an absolute AI value of 0.5, or do I get a value which is 0.5 of the game's setting of 0.75, which would amount to approximately 0.37?

Funnily enough was just about to post a question on this?

I am wondering if one overides the other? So the Editor overides the difficulty setting as it can be more unit specific?

But then what do I know? confused_o.gif

Share this post


Link to post
Share on other sites
As an aside, how does the AI slider in the editor work, does it set the AI to a value unconnected to the game settings, or does it set the AI to a value WITHIN the game settings?

Example, if I set the game enemy skill to 0.75, then set a unit in the editor to 0.5, do I get an absolute AI value of 0.5, or do I get a value which is 0.5 of the game's setting of 0.75, which would amount to approximately 0.37?

In my experience:

AI slider in the editor *does not* override the skill levels set in the config stuff. If it did that, there would be no point in setting the config stuff at all, since every soldier placed in the editor has a skill value, and therefore they would all override these values.

If I had to guess I would think that it adds and removes from their total value. However, I know this isn't something as simple as setting skill to 1 adds .5 to their total score and setting skill to 0 subtracting .5 from their score.

In fact, the slider in the editor has little if any effect at all. It's kind of a difficult thing to objectivly test. I did a couple of quick tests and it did seem to have a *very little* bit of effect. But that could have also been simple deviation from average too. I didn't exactly run 100 of each tests using charts stop watches, etc you know?

What does this mean to the mission designer?

You either have to specify what game settings the game should be played on in the readme if you want your mission to produce consistant results.

Alternatively, look into the setSkill array command. It may or may not enable us as designers to produce consistant results regardless of game settings. I haven't tested it.

Share this post


Link to post
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
Sign in to follow this  

×