v1.05 Curious how you feel about small arms accuracy now.

[Guest Guest]

deleted per user request

[c3k]

My very limited experience supports the slower ROF for longer ranges.

At 300 meters, prone, it was VERY difficult to even be sure that I had just seen a pop-up silhouette rise up. This was on rolling terrain covered with light scrub. There were no markings or lanes, hence no known ranges. There were a lot of torso pop-ups which would pop-up at random. I was using an M-16 with iron sights. At 300 meters the front post completely occluded the torso.

My ROF certainly dropped as I concentrated on aim and wind correction. (No misses. )

Regards,

Ken

[Secondbrooks]

Originally posted by JasonC:

Secondbrooks - thanks for the test, and I agree that the represented troop concentration may well be too high, and that may be part of it. Interesting observation on high casualties from single grenades - I've notice that HE weapons "nuke" squads in CMSF, well beyond what seems realistic (e.g. single HEAT rounds from rocket launchers nearly wiping out 7 man units etc).

On the issue of average accuracy in WW I, though, I think your figure is off. 20-30 is more like the number of heavy artillery rounds produced (by all sides) for each infantry casualty suffered (by all sides). Small arms expended over the whole war exceeds infantry hit from all causes by many orders of magnitude. It was normal for single nations to supply their force with 100 million rounds per year and up, while losses ran single digit millions over the whole war - with the lion's share caused by artillery.

Yesterday i did test with two US stryker troops (42 men was strenght for both sides, consisting standart platoon, CO and Batallion commander). It was blue vs blue, two long trench lines having 200 meters wide no-man's land (standart sand) in between.

After 4 minutes of intense firefight both sides suffered about 6-10 casulities (usually 1 dead while rest were wounded). Casulities seemed to been coming in steady pattern during battle, no lulls or heatpoints.

AI which played as red side and was set to active used alot more ammo, about two times as much as my guys. They had used almost half of their ammo, while my men used 1/4 of their ammo. And usually they caused also two times bigger casualities to me, i lost 10 while they lost 5 or 6. This is usuall pattern which tells me that AI leader managed it's fire better than me and so gained advantage over me. I gave my troops just target arches which covered whole enemy's trenchline and then i let them shoot (or not) for 4 minutes.

Ofcourse test was pretty bizare. "Hey. Lets shoot enemy over there!" I wonder how human mind would have thought about that. I quess most would have pulled their heads down quite soon after first casualities and not revealing themselves after that, but for short moments when firing few rounds or bursts.

About WW1: Thanks for straighting that up, i found those figures bit hard to believe myself.

[ December 27, 2007, 04:55 AM: Message edited by: Secondbrooks ]

[Zemke]

All small arms accuracy data would be easy to get. Take the percentages of target hits for each weapon from soldiers in basic training or untrained soldiers them compare that to trained soldiers, just go to your locale basic training post and ask for the numbers. Accuracy will go down as stress increases for less experienced troops, also, as well as suppression levels. Earlier I read in this post that hits from 200 meters were right on, that is crazy, if true then the game is way over modeled. Most US soldiers cannot hit a man's head at 200 meters for all the tea in China, and from what I have seen people in the middle east cannot shot for crap. The armies of that region tend to think weapons make the man, not the other way around, in other words if they have good weapons they will do as well. But I would put any western force armed with WWII weapons against them, and I'll bet the results will be almost exactly the same; their armies have poor marksmanship training. Most of small arms casualties take place at 100 meters or less, unless you happen to be charging across the open ground covered by stabilized MGs.

[JasonC]

Zemke - no, it is not easy to get, because it has absolutely no relation to what happens on firing ranges. Nor is it army specific marksmenship training effects. Personally I do not miss at 200 meters or less, with an ordinary M-16 and iron sights. I rarely miss at 250 meters, beyond that bullet drop gets significant and my misses rise. Give me a full rifle round and I can extend that to 300 to 400 meters. But this has nothing to do with the question, how many rounds would I fire on actual battlefields to hit even one enemy man?

The fundamental theorem of operations research is that the average weapon system never accounts for even a single opposite system of equal power, as a portion of side power. A rifleman (unscoped) is a below average weapon system (arty is much more effective etc). Ergo, the average rifleman never hits a single enemy. Over his entire period of service, with all the rounds he is ever issued, in all the roles he ever plays. We know this, to a mathematical demonstration. No hypotheticals about what *should* happen or be possible, can alter it.

The proof is simple - men walk off the field alive. The portion of men actually hit in war can approach one in the long wars (with some men hit multiple times, healed and returned to duty and hit again etc, while others go through the whole war unscathed). But it takes the whole war to get to that figure. Typically it is well under one on the winning side, and one side or another wins. The average is therefore under unity.

Well, it is the above average systems that are doing most of the hitting. Up to 70% of infantry wounds from shell fragments in WW II, e.g. A modest portion of those are grenades and some are direct fire cannons on tanks, but a solid half is indirect artillery fire, which is by far the most effective overall killer and wounder. The reason being very simple - the weapon can fire for long periods without being is any real danger from the enemy, and without needing to locate him too precisely.

Those are the above average weapons, and everything else is perforce below average. The average is less than unity. With the arty backed out of both sides of the "hitting and being hit" balance sheet, the "being hit" column is much larger than the "hitting" column.

Even within that remainder, rifles are below average weapon systems. The belt fed MGs are the leading bullet killers, by miles. They have the volume of fire to hit whole areas, the full power to reach to long ranges, and are important enough to be manned continually, remanned frequently even when operators are hit, etc.

Scoped rifles used by specialists with the highest training are also far above average systems. Besides their technial superiority (optics e.g.) they have stealth and reach to fire with near impunity, and long loiter to wait for an engagement since the specialists involved are not given other missions and can focus exclusively on a good attrition rate. They get well above unity.

Back those out of the hitting and being hit column, and precious little is left on the being hit side, to ascribe to the ordinary riflemen. Who are, however, hit in great numbers. Making a plain rifleman a far below average weapon system - and also implying that it is a rare outlier of a plain rifleman, who shoots even one enemy before being shot or shelled himself. Despite all the rounds he is issued and uses - 70 a week is a typical figure.

Ergo, to a certainty, the typical rifle shot occurs under such horrid conditions of visibility, enemy location knowledge etc, that the chance of actually hitting anything is well below 1%, and is actually probably well below 0.01% or one chance in ten thousand. Nothing like this is seen on any rifle range, regardless of the training standard involved.

Ergo, nothing seen on any rifle range has any relevance or bearing on the question, whatever.

You don't have to like this conclusion for it to be true.

[Omenowl]

I have seen demonstrations with .50 bullets fired into brush and it still alters trajectory of the bullet. To say concealment does not give cover would be somewhat imprecise. Brush won't stop the bullets, but it will deflect them and it is much better than nothing.

[Battlefront.com]

Adam1,

I don't think the opposite is true, I just don't think it is the nature of the problem we're talking about.

Well, as I've said all along this is a complex simulation, therefore it is highly unlikely that there is just one thing that needs to be adjusted. One of the main sources of complaint, which I tend to agree with, is that as of v1.05 targets at fairly long ranges are being hit too easily. One reason for that was a slightly unrealistic volume of lead hitting the target. It's a simple matter of physics... the more lead flying at the target, the greater the chance of hitting something. The less, the lower. So we reduced the volume of fire based on sound principles, therefore casualties at further distances are reduced based on that. Since this is part of what people found to be inaccurate, I think they will agree that we made a good improvement.

Now, that is not to say that there isn't anything else to look at. So if you're concerned that we made this one tweak and that's it, please don't be. We found a definite area that needed improvement so we improved it. To ignore the issue would have been detrimental.

I think that crew served weapons fire sporadically enough at range as is.

Crew served weapons have a greater RoF so the differences should be less noticeable with them vs. squad weapons where 300m is getting up there in terms of effective range.

That seems to me a design problem. Rather than keeping it and dropping the solution, why not tweak the way that design works? (If possible, if not oh well.)

Because I don't think your "tweak" is a good idea I already outlined my reasons, but the biggest one is that what you've suggested would require 1:1 LOS to make any sort of sense, and that's a practical impossibility. So without 1:1 your idea actually makes things worse, not better. But that's just my professional opinion speaking

In any case, if it doesn't work, no worries. The main thrust of it is that if "micro-skulking" is the problem rather than penetration modeling through cover, it would be better to make a change like that than compromise the accuracy of the penetration model.

True, if the soldiers are not taking adequate precautions against fire, that is something that needs to be tweaked. However, tweaking is all that's needed. For example, Charles can change some cover values and/or he could make changes to the TacAI to alter individual Soldier behavior. I'm not saying these tweaks would necessarily be quick to implement (anything with the TacAI is a chore due to it being AI), just that the underlying model is very, very good. No system is perfect, though we will try to get it as close to that as possible.

Bipod, you're correct. The thing is I am seeing the PK shooting 25-50m to the left and right of the target at the same spots respectively over and over, like an abstraction, alternating between that and shots lined up on target which miss in a more natural way.

I've made a note for Charles to check into this. The shots, even at that range, should still be spread out quite a bit when on a bipod. A good gunner and slower RoF can help reduce the spread, but a MG on a bipod is really a weapon for suppression, not for felling specific targets (though of course they can do that quite nicely too).

JasonC covered the issues of figuring out exactly what "accuracy" is when simulating a combat environment. It's a lot of black magic and voodoo, unfortunately. However, we always start with the "inherent" accuracy of a weapon and then degrade it's actual performance based on a whole host of factors. This includes everything from a snub nosed AK to a GPS guided munition from a fixed wing aircraft. The inherent accuracy gets us a nice place to start off with, but no more than that.

Omenowl also pointed out that it doesn't take much to deflect a round. Angle and impact surface can do all sorts of things, even to very heavy rounds.

Steve

[ December 27, 2007, 08:53 PM: Message edited by: Battlefront.com ]

[Thomm]

As I am sitting in front of my laptop, I imagine how tall the silhouette of a human being 1.8 meters tall would be on the wall opposite me, at a distance of 2 meters. It would be 1800 * 2/200 = 18 mm if the man stood upright at a distance of 200 meters. That is awfully small. Factor in movement (short dashes) and guys being prone and I feel that - shaking from fear or stress - I would have a hard time hitting someone.

Best regards,

Thomm

[JasonC]

If you can't hit a man sized pop up target at 200 meters within about 3 seconds, from a prone position and from a position standing in a foxhole and braced against the side of it, with a zeroed M-16 with iron peep site - then you won't pass basic rifle marksmenship in the US army, and will never graduate from basic training.

The specific targets are at 50 and 100 meters head and should sized, 150, 200, 250, and 300 meters man sized. The longer shots have a couple more seconds, and you can pass even if you miss all the 300 yard ones. But you have to hit 65% of the time at the whole mix of targets, with one round for each, and frequently 2 targets up at a time.

You can use rounds "meant" for the longest shots for second chances at the intermediate, if you like. To get an "expert" rating you have to hit 90% of the time, which requires hitting the longest targets (250 and 300) at least half the time.

Those who fail once through can try it again, and some people take 3 times, though even that is rare. Most people can hit between 2/3rds and 3/4 of the shots. This is after a couple weeks of rifle training and hundreds but not thousands of rounds fired - nothing extensive.

It simply is not that hard to orient a rifle along a fine trajectory. It is what the bleeding things are built to do. The Marines fire 500 yards at static targets, rather than popups - at those distances they are being tested on their understanding of bullet drop and windage adjustments, effectively, rather than just basic sight picture and squeeze control, which is all you really need to pass the army BRM course.

But it is all rifle range stuff, not combat reality. In combat, such shots will occasionally be made, because a riflemen happens to get clear LOS to a visible enemy for a few seconds - especially if the target is stationary for those seconds, typically because he doesn't know he is being observed. But it is rare to have the opportunity, and most combat hits instead come from heavier weapons, vehicle MGs, or SAWs and such firing bursts to get single hits.

[molotov_billy]

Originally posted by JasonC:

Ergo, nothing seen on any rifle range has any relevance or bearing on the question, whatever.

You don't have to like this conclusion for it to be true.

I wouldn't say it has no relevance - why do armies emphasize training with small arms? - it would be just as false to say that any individual piece of the puzzle is useless.

At the very least it gives you a comparative analysis between individual weapon systems - X gun is more or less accurate than Y gun, given that outside factors are the same between the two.

[molotov_billy]

Good example of long range rates of fire.

[Secondbrooks]

It's quite obvilious that concealment weighs nearly nothing, which then again has part in lethality of small arms fire.

Guys hiding in grain field and still they can be spotter and shot dead from 100 meters away. And there was nothing but grain in between them. Basically one chould no tbe visible when he's lying prone, hardly even if he's kneeling (acoring CMSF guys head is visible). One minute was enough to get 25-33% of men killed or wounded. Hiding was not a option, they couldn't get hiden from enemy but it kept shooting them with deadly accuracy.

Yes. Again i felt like testing

This time: Militia platoon (regular experience and normal morale etc..) against same kind platoon. Testing ground was totally flat, only ground type and such changed. Platoons had distance of 100 meters in between them.

I also tested these types of terrain:

Rocky ground. Effect was almost zero (just as with grain field). After same amount of time, about same amount of casualities were produced as earlier.

Brush and three bushes per square in grass soil. Effect was almost zero. 100 meters was piece of cake. So after same amount of time same amount of casualities. I tested bush models A and B, no differences seemed to be.

In grass soil, brush and three times tree model E per square. From 100 meters no contact to enemy, so i started to advance (with hunt order). Enemy opened fire from aprox. 80 meters. Fire was pretty ineffective: basically one man went down from opening fire. I could hit the dirt, lose contact and fall back if i chose to. if i chose to advance i sacrificed my platoon for nothing. If unit is standing he is as good as dead. Tree model C performed the same.

Brush and three times tree model D per square. From 100 meters no contact to enemy, so i started to advance (with hunt order). Enemy opened fire from aprox. 40-50 meters. Fire was pretty ineffective as in previous example.. Due shorter ranges it seemed that effectivity of crossfire increased. Surivability from front was supprisingly good but fire from side could wipe out several guys in short moment.

Overall i think that concelament and cover is way too low in all these examples. Even in last two examples, altough they are closest to reality.

Overall i've noticed that movement in prone position has little of effect to stealthness of unit and same applies to lying prone in certain ground types (guy shouldn't be visible if there is 100 meters of bushes or grain field in between. Or 80 meters of forest). It should usually be possible to get much closer by crawling than with walking, but right now it seems that this is not the case. It has only slight effect on things.

I know that i sound like prick, but could Charles look into this aswell Or is there logical explanation for such low concealment values? Ofcourse it could and most likely will break missions if this thing is tweaked drastically.

EDIT: ****e. We are having marksman conversation. Raising weapon to shoulder then douple shot inside 5 seconds to pop up target (or 1 shot in 3 seconds) from distance of 150 meters with AK type weapon... No problem. We trained alot same things as US army seems to be doing, these so called snap shots.

The thing itself ain't difficult, problem from what i've seen in allkinds of combatexercises is that you usually won't see enemy. Just hear gunshots or see enemy briefly and then it's gone. And this happens under conditions when i don't need to preserve my life (same applied to opponents). Usually when i saw them and was able to nail them was because they were acting stupidly... And same applies to me when i got hit: "It's too wet to be lying in the ground" or "i don't feel like running" or "I have to smoke a gigar".

molotov_billy: And that was from unsuppressed unit. What if there would have been tank or 12.7mm HMG firing at them? Like one of our famous WW2 veteran said: "When enemy stops firing and flees then even the most self-preservant can act like hero." Not to be taken offencively by anyone, this is what he thought after his 4 years of war.

[ December 27, 2007, 04:29 PM: Message edited by: Secondbrooks ]

[Guest Guest]

deleted per user request

[Guest Guest]

deleted per user request

[Secondbrooks]

Only area where hiding or being prone was truly effective was trees + brushes (it made units much more hard to kill and made them invisible to opponent. Bushes seemed not to give concealment at 100 meters, which was biggest supprise to me. Also grainfield's zero effecticity in concealment at 100 meters was quite a supprise.

Also brushes itself doesn't offer concealment, so basically in earlier tests with 'trees and brushes' only effecting factor were trees. Just minute ago i did same test as before (with 100 meters distance) with only brush filling the map and effect of brush seemed to be zero, enemy opened fire and killed most of platoon with in minute. Men were hiding (lying prone) in there and still enemy spotted with out problem and also shot them without problem.

Basically this far from brushes, ground- and foilagetypes only trees (expacely D model) are limitting visibility and effectivity of fire: hit the deck and you are relatively safe if distace is enough long... Or something like that.

EDIT:

About tests where there were no trees placed on map:

Has these 'problems' always been there? Or am i missing something: like how these haven't been "figured" out earlier? Or have they?

Both sides were given order to hide in scenario editor. AI was controlling red with 'active'-order, so he opened fire if they spotted my men. While i just issued 'hide'-order to my men during setup, i didn't give them orders to open fire (and they didn't open fire). This is how i conducted most of the test. I also tried to give my men target arces and gave them liberty to shoot if enemy is spotted, but after short moment i gave them 'hide'-order to hide away from enemy fire. No difference to earlier example, both sides spotted and shot each other with few seconds delay after mission started. And these guys were regular experienced militia troops all having binoculars (like they would have any effect in that kind terrain).

I also tested it in hotseat, and ordered both sides to hide in setup, in 'bushe+brushes'-map and just 'brushes'-map which both i tried. In both maps within 30 seconds most (60-80%) of hiding opponents (=individual men) were found out, and this was from longer 150 meters distance. They didn't do anything else but tried to hide from enemy.

EDIT... about... 4: Sorry for editing this post so many times... This must be actually be something like 6th editing already.

[ December 28, 2007, 05:43 AM: Message edited by: Secondbrooks ]

[Topo]

Thanks to Secondbrooks, interesting tests.

I think cover/concelament must be refined too.

Saluti

[JasonC]

Secondb - could you please clarify a point here? Are all your tests at exactly 100 meters?

I realize for the heavier woods they don't see each other at 100 meters and you moved them closer until they did. The question is, is there some other distance, beyond 100 meters, at which brush or wheat cover will block line of sight.

In reality, fights like that become "gophering" sessions. Meaning, in order to see, a man has to raise himself above the level of the brush or wheat, and when he does so can only see the enemy if they are "up" at the exact same time. Men then reposition, prone or crawling, to *not* be where they were the last time they "popped up". Because the enemy will "area fire" at the "base" of pop up locations, if they see any. But going prone in a position no one has popped up from, will in practice provide near perfect safety. Fire will not cover the whole area. The volume of fire is low, because most of the men are spending most of the time crawling, prone, or resting, and try to avoid revealing themselves.

So there are several issues here. First, prone movers in brush or wheat should be much harder to spot. It takes an elevated position to do so. More generally, prone movers should be more like prone stationary in spotting terms, and much less like upright movers - the reverse seems to be the case now.

Second, spots are lost much more readily than the game depicts. Once a unit is seen in CM, it is quite hard to hide again. Only at quite long ranges is it effective. Taking incoming fire, the unit typically either fires back or tries to move, and both maintain the enemy's spot. In reality, dropping prone and sometimes short prone movements, rapidly break LOS contact successfully. This is a sighting and hiding behavior issue.

Third, part of the reason for the unrealistic effectiveness locating enemy infantry in this sort of concealment, is the squad blob grouping forced by the game units. I mean, in reality it is one or two guys who are seen, and there it no real assurance that there are 5-6 other guys right next those seen. The "gopher" spotters in fact avoid popping up near friendlies.

It seems to me the last calls for something like the vagueness shown by sound contacts, in the sense that the spotted location can be "off". Not because you think the man you actually see is somewhere else, but because he is not dead center of his squad at the time he is seen. Spots with any kind of concealment involved, should have some "error cloud" like sound contacts, "smearing out" the reported position. Not as far as the 200 meters or so used for sound, but more like 20-40 meters.

This would tend to discourage over use of area fire in such situations, in a realistic manner. I mean, you could hose the sighting and everything around it and would be able to hit things that way. But the ammo used would go way up and the firepower per unit time hitting the right target, way down. It would also become less reliable, since sometimes you'd fire for minutes at the wrong location and simply miss etc.

Coupled with the first - ability to move at crawling rates unseen - I think this would make the use of concealment cover types dramatically more realistic.

[Zemke]

Jason,

I guess I should have been more clear. Bottom line, the accuracy in the game for small arms is too high. Jason it is very unlikely you will hit anything at all at 200+ meters while under stress, but I think you are saying that also.

What I was trying to say is, take the data for small arms fire, then plug in the variables of stress, experience and training, then reduce again due to less than ideal environments, and you get better results or more realistic results than the game currently allows. The US Army has tons of data on rounds expended and BDA to support this, and BFC should take advantage of this information to build a better model.

An example of unrealistic accuracy:

While playing last night, (testing the new patch) I had an American Squad should taken out in seconds after dismounting from a M2 by another enemy Squad 200 meters away. My men would have never suffered the number of hits they did, and it seemed body armor didn't do a thing, and even if wounded they still should have been able to react better than the game allows in the game. Also, I don't think the Syrians could ever shoot that well and certainly could not knowing by opening fire they will be getting 25mm chain gun fire right back. I would be scared ****-less to give away my position like that, of course ignorance is bliss.

Morale will break long before any unit is wiped out to the last man, examples of fighting to the death are rare in western armies. Some of the recent experience in Iraq and Afghanistan, has show the enemy is willing to do so, but the results have still been very light coalition casualties and the killing to the last man of the enemy, although not necessarily by small arms fire.

[ December 28, 2007, 09:28 AM: Message edited by: Zemke ]

[Secondbrooks]

Originally posted by JasonC:

Secondb - could you please clarify a point here? Are all your tests at exactly 100 meters?

I realize for the heavier woods they don't see each other at 100 meters and you moved them closer until they did. The question is, is there some other distance, beyond 100 meters, at which brush or wheat cover will block line of sight.

Yes distance is solid 100 (+- 10) meters, elevation was same solid 20 in whole map.

I think that 100 meters should be well enough to give good concealment. I'm not sure about rocky soil type, but atleast field of grain, bushes, brushes should do that very well (on flat ground as you said). Right no it seems that this is not the case. I'm having doupts also conserning craters, but i'll leave it to that.

That one other test which i did aswell using 2 player hotseat: 3 bushes + brush per square (filling whole map) and yet most of individual men from both sides hiding there, were spotted inside 30 seconds by their opponents. Distance in this test was 150 meters. In reality i can't see even 30-40 meters when lying down in that kind terrain... If it is what i think it is: The bush-hell, full of shoulder, waist and knee tall foilage.

And i agree with your post complitely.

My point hopefully should be atleast somewhat loud, space-consuming and clear, so i crawl back to my hole.

[JasonC]

Zemke - I understand, but nobody knows any of the intervening factors with any precision. They are all subjective guesswork, each of them. We know the "span" or variance of personal marksmenship on the firing range runs from nearly always makes these shots to hits maybe half of the time - a factor of two span, right up against "always hitting".

Then we know the reality is 5000 times less than that. We know that factor of 5000 difference is composed of a dozens of factors of 2, but we don't know what each of them is in detail, and which is a minor 10-20% thing and which is a factor of 3 or 4 instead of two times etc.

Well, the point is that the first off factor of 2 seen on firing ranges has nothing to do with the remaining dozen factors that actually drag down the hits per round by 12 more powers of 2. Nothing we see there can tell us anything about them - they depend on visibility on battlefields and fire discipline and opportunities and what typical infantrymen spend their time in combat doing etc etc - all of it entirely imprecise and subjective, taken factor by factor.

We can know the endpoint a bit more precisely - that about this many hits occur for this many rounds thrown. But it is not the case that you can just throw so and so many rounds (say, 10000 per hit) and get the reduced number of hits - because that average seen depends on the average visibility, average behaviors under fire, average time spent doing what, average target density etc, of real combat.

If the conditions under which you fire off some 100000 round quota are different (e.g. all at once without moving, area fire, general direction of a thin enemy), you may get no hits at all, not an "expected" 10. The real determinants whether you will hit 50 times or 10 times or 2 times, are all those subjective intervening factors.

Modeling them with some tactical realism, *without* having any precise figures to go on, is the whole art of simulation design. The controllable aspects have to at least move in the right directions, and the tactics one has an incentive to use as a result, have to be reasonable in real world terms and have reasonable average effects in terms of losses etc.

Neither BDA averages of rounds fired, nor range marksmenship figures, can do this for the designer. At best each just forms a statistical endpoint or "control". They can "falsify" some models, but alone are not "tight enough" to specify just one.

Suppose one model thinks factor number 4 is important and varies in a range of 4 times, while 6 is unimportant and varies only 10% - another model thinks the reverse. They can have the same top level best possible marksmenship and the same bottom level BDA average for hits per round fired. But they imply very different tactics, because one makes (say) concealment a dominant factor and the other thinks that relatively unimportant, while the other thinks a key issue is target movement, harder to hit moving men than stationary.

There are scads of models that will fit the endpoints (best shot condition accuracy, and average achieved accuracy, I mean), as well as tons more that get one or the other of them wrong. But even the models accurate on those points can be right or wrong about all the intervening ones and encourage realistic or unrealistic tactics, as a result.

To me it is obvious one has to design for effect in these things, and tinker every time you find the system is setting up incentives or encouraging tactics you know don't work in reality, and then try to iterate to "walk" the model in closer to reality.

Which is a messy empirical process, and any one change can move other bits farther out of whack. It cannot just be solved by looking up some average and putting in "the right number" at one point - although the eventual tweaked thing will have involved doing that "for effect", repeatedly.

[ December 28, 2007, 11:51 AM: Message edited by: JasonC ]

[LongLeftFlank]

Jason, given this discussion has spread across 10 pages in 2 threads, could you kindly recapitulate your recommendation for a workable design-for-effect fix, or edit the below?

1. Suppression-based firepower reduction or cover modifiers for stationary units based on dyads (e.g. LOS pairings). Modifiers escalate to "gone deep/total cover" where a unit is all but unhittable (but also doesn't shoot) from a given position -- depending on cover terrain, range to shooter and shooter weapons.

2. Rapid loss of spotting to/from suppressed units.

3. "Spreading" out of area fire effects to (a) improve the defensive effect of concealment ( blunt the impact of the (hard to fix) design that keeps non-moving units bunched into a single 8x8 target rich "slab o' meat" even when under fire.

4. Other

[average]

Two things really stand out. One is that the infantry have the ability to direct fire to finely, and second, allowing for the relatively accurate modelling of weapons, infantry in game behave like reactive targets rather than infantry.

Infantry minor tactics are kind of lacking. Dudes pop up in the same spot over and over again, they skyline themselves, they like getting enflaided and they are really bad at hiding.

In terms of movement across open terrain, its a lack of individual and section drills that makes small arms fire so lethal. That and perhaps the number of dudes shooting at any one target is far too high, considering that one of your basic considerations is that if i can see that dude, then he can probably see me, and that is probably a bad thing unless ive got the better cover.

One of my major frustrations is my grunts can't pepper pot across open ground, have difficulty dispersing themselves, and move in spastic fireteam bounds. I've never seen them attempt a section movement in peels and leaps of 5m combined with copious fires and the proper use of concealement and cover.

Instead they race to get into the beaten zone and get smacked up, although they remain combat effective far too long when caught in the open.

I think JasonC covers the balance of anything I'd want to say, better than I can say it.

I don't know if this has been added in as well, but dudes can't dig in and don't know about vertical cover and strongpointing fighting positions.

[Guest Guest]

deleted per user request

[Wiggum]

Today i had a bad moment while playing CMSF.

The Squad was ordered to run into the house (blue line). Someone shot at them (small arms fire, red line) and it only takes 2 seconds and the hole squad was down...that must be a very good shooter.

Please Battlefront turn down the small arms accuracy. At the moment it makes Urban warfare scenarios to become very bad.

[thelmia]

I'm a bit annoyed at cover not mattering much. Troops out in the open seem about as vulnerable as those in trenches or behind trees. I know that's not 100% accurate, but that's my impression.

It's bizarre to me that a machine gun shooting out a window will repeatedly miss a prone man lying on the pavement two stories below and 20 meters away. Then that same machine gun will kill enemies 200 meters away running across a street. I get that running in the open is bad, but aren't stationary targets easier to hit? I get that troops taking cover should be harder to shoot, but lying on the pavement underneath a machine gun should not count.

I am beginning to come to the conclusion that the abstracted infantry fire in the original Combat Missions was closer to reality than the more detailed model in CMSF. I hate to say it, but it's a case of a simple abstraction being more accurate than a complex one that leaves out too many variables. The problem is that the increased detail invites questions like the machine gun that misses at 20 meters over and over, where a more abstract system would be less frustrating. Like maybe there's a big rock there, or something else we can't see. With a more detailed 1:1 model, we want to see the rock.

More detail makes the game feel less realistic because it invites more comparisons with reality. It looks more like reality, and suffers for it.

This is all impressionistic, and I'm sure that it can be tweaked so that it doesn't feel wrong. I'm not against the way that CMSF does things, just how I'm continually reminded that I'm playing an imperfect simulation of reality.

[ January 02, 2008, 01:13 AM: Message edited by: thelmia ]