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

[SgtMuhammed]

Originally posted by Sgt.Rock of Easy Company:

As for why theres like 250,000 or 400,000 rounds/kill, I would say that urban environments would be extremely similar to jungle in terms of LOS and whatnot. That and your so point blank sometimes (within 100m is pretty damn close, let alone room to room or buildings) that I would assume people would shoot as many rounds as it took before they are sure that the enemy is down or dead, whether they see them or not.

Is your dad a Marine? They do longer range stuff on the KD (Known Distance) range. I don't doubt his qualifications but he will tell you himself that range and combat shooting are as alike as driving and Grand Theft Auto.

The high ratio of rounds to casualties is because most fire isn't aimed at anyone. The vast majority of fire is aimed at locations where the enemy might be or could be or that's where I would be if I were him. Aimed shots at a definite target are few and far between.

[Drusus]

I am going to use paintball as an example in this post. I know that in many aspects it has little to do with real warfare, but the point I am trying to make here is about the concept.

First, some background. I was active paintball player for some five years, playing tournament games both here in Finland, and also in Europe (Millennium series). So, I know something about paintball, and this is why I am using paintball here.

The basic setup of the game when using those air filled cones as cover is that there are some places of perfect cover and then the rest is without any cover. In the game it is extremely important to keep your enemy's head down. This is why it is easy to use 1000+ balls per game. If you can manage this, you can maneuver. And when you maneuver you can hit your opponents.

Many times the game is resolved when there isn't seemingly happening much. Both teams are trying to suppress the other. The better team either manages to force the enemy to keep their heads down, or if the worse team tries to shoot back, they will get hit when trying to do so. The inexperienced players are easy to spot in the game. They don't know how to take proper cover ("my eyes are behind the cover, thus I am safe") or they take huge risks trying to shoot back when they are "dotted".

Now, how does this relate to modern warfare? I think the concept is somewhat applicable. That is, in modern warfare you use enormous amount of bullets to make your enemy keep their heads down. If you manage to do that, you are free to maneuver, and thus able to kill your enemy. In modern warfare you have also other tools for the problem, heavy weapons being the example.

The crucial thing here is that in paintball, it doesn't matter how much you shoot when your enemy is keeping his head down. I mean, you can't hit him by just shooting more when he is in cover. On the other hand, if he isn't in cover, you need just a few balls to hit him if you can shoot freely.

In this game there isn't such cover that is able to withstand _huge_ amounts of fire. At least I haven't found such cover. Now, the claim is: in real combat there exists such cover. And I think very much the same thing should happen in combat than what is happening in paintball: the other "team" is able to suppress the other team and thus can maneuver for the kill (eliminate in PB terms).

It is also worth mentioning that there should also be those spots of no cover. I do not want the whole map be like the paintball field, where there is lots of spots with perfect cover. But there should be those spots also.

How the game should represent these spots of cover is up to Steve & Charles. As JasonC pointed out, it most likely has to be some sort of abstracted cover model. If not, excellent. Of course, it might be that they decide the game is good as is. That is good, too. I can keep complaining...

By the way, what comes to the range shooting and combat shooting, in paintball the accuracy goes as follows: Before game 9 hits out of 10, in game 1 out of 10. The target being the same, of course. Hitting moving targets is really hard when the target is visible only for the 5-10m as is usual in paintball. And the range is something like 20-30m!

In real combat you have one additional thing to worry about: if you get shot, you can get killed. I imagine that would drop accuracy a bit more.

Back to the topic of my post: if CMSF would try to simulate a paintball game, the game would be decided by just sitting behind those back field bunkers and shooting until every opponent is eliminated. No matter if he is in cover or not.

The same problem is present also in CMSF's simulation of modern warfare. You can win by just shooting enough bullets at your enemy, no matter where he is.

[Lethaface]

In my opinion the problem here is the lack of AI using cover adequately... I can understand that this is hard, if not impossible, to code correctly.

A slight abstraction in using cover in semi open areas while being prone, is not un-argueable. this seems to be in place at the moment, but perhaps it's values could be tweaked a little bit as to give a more satisfying result, when compared to Real Life situations.

The idea to make some random soldiers be like 'union road workers' - or whatever goverment paycheck non-do'ers are supposed to keep up their name to , does appeal to me. I've been informed that, although the rate of fire fight particpating soldiers has increased firmly since WWII, there are a lot of soldiers not actively firing during modern day firefights.

[Guest]

Just a thought: has anyone noticed troops becoming suppressed by fire that did *not* hit one of their number? I have not, just wondering if anyone else has.

And as for playing as RED... indeed. It's hard to win with green, untrained troops, but enormously fun. A bit odd to see one of my men pop somebody 100m away through a window, though, while under fire, in a low-light (not night) situation.

An unlikely shot, some might say.

Edit: Having US AI "infiltrate" my position at night and beat the crap out of me from 10m away was enormously fun, though, especially with the new dynamic lighting.

[aka_tom_w]

Does anyone have a test file or save game or test scenario that clearly illustrates an "accuracy overmodeled problem" that is a sure thing and repeatable??

There are a lot of issues in this thread, Jason C mentions cover status at length. But for me its the 200m+ shot that hits guys running in the open on the first volley or first round that gets me, or the Syrian unit that picks off the head of an opposing unit at 178m with the first round volley as they crawl over a crest line or berm into LOF. :confused: :eek:

[ December 18, 2007, 11:23 AM: Message edited by: aka_tom_w ]

[Battlefront.com]

Some quick comments...

CMx1 and CMx2 both simulate "crossfire". This means the more points of fire, and the greater the differences in angles, the less chance the more the target will suffer. To picture this, 12 guys firing from one spot vs. 6 guys firing from two significantly different spots is no comparison. The 6 guys will probably do more damage than the 12, all other things being equal. The math in the game is vastly more complex than that, so please don't take this example as statistically meaningful.

There is no such thing as being "immune" to enemy fire. Well, not in the sense you guys are talking about. When a unit improvises cover it will do so unevenly. Some will find excellent cover, some hardly none at all, and everything inbetween. This depends heavily on the available cover, of course. As the enemy unit pounds away the chances that this cover will be compromised increases over time. Guys get curious and stick their heads out, a leg becomes exposed, a ricochet finds its way in, chips of rock do what bullets can't, etc. This is true even for prepared defenses such as bunkers and trenches.

To use Druss' paintball analogy, I've definitely seen squad sized encounters where one side stops dead and basically stays put behind thick cover (I only played in turbulent New England hardwood forests) only to be whittled away over time. And that's against guns that are not very accurate and have no chance of punching through thick brush or thin trees like bullets do.

What I'm getting at here is that when a squad drops down in otherwise empty terrain, and it is being shot at by competent riflemen with plenty of ammo and nobody shooting at them, expect the squad in the open to eventually get cut to pieces. Even at significant range, though it should take a bit longer.

Steve

[Guest]

Playing Al Huqf last night, against a Conscript/Poor red militia battalion. I was advancing on the big building while they were doing the same.

I had maneuvered up to the adjacent houses and destroyed nearby red elements, established overwatch and supporting fires, and was pouring a whole squad's worth of fire into the first floor of the big building.

I moved a three-man flank element across the wide street that the big building looks down, maybe 100m from the building itself.

As all this fire is pouring into the building's first floor, I see a single muzzle flash come from one of its windows. The middle man of my three-man group, who had been Quick'ing along quite happily, went down.

Now, this was awesome to see (and also raises the question: can we *force* buddy aid? That guy died because his two buddies, once the coast was cleared, refused to help him), but it seemed just a little improbable. Am I wrong?

Also... given any thought to allowing us to move our wounded? I'd think moving wounded out of the cone of death would be within CM:SF's scale. Another thread, perhaps.

Cheers.

[PeterLorre86]

Just had my sagger crew wiped out from 300+meters by one squad(on the move no less). My saggers were crawling up to the crest of the hill, and killed imediatly after cresting. Wasn't able to get it on video though.

Close range the explanations given by Steve make sense.

But 300+meters, crawling to the top of a grassy rise? How could they see my men, much less acurately kill all 3 of them with thier first volley? While running mind you!

Still, close range works great IMO. 4 of my recon squad walk up a hill and meet face to face with about 30 US soldiers. One of my guys bolted, two went straight for the grenades and managed to throw one each, the last was killed with his jaw dropped in suprise. The whole thing was over in 3 seconds. Caught that on video, i would upload it to utube, but it exceeds the size limit. I was very impressed, exactly how i would expect things to work in the real world.

But 300+ meters!? Come on now.

[JasonC]

Steve - thanks for the feedback about how it works inside. And I appreciate that crossfire effects are moduled, that is clearly important.

I'll explain the sort of situation I've seen that I think calls for correction, you can decide if it does, and whether something like my suggestion helps with it or not.

A platoon of infantry is in a set of trenches, not bunched up.

A single vehicle with MG main armament pulls up to a spot 150 to 250 meters from the trenches. It sees men in them, but has no "enfilade" nor any crossfire from other units. The infantry in the trenches have no weapons with which to fire back effectively, at that range.

The vehicle gunner depresses his MG trigger and leaves it there.

Every man in the trenches dies.

A few hit and all suppressed, no problem. That suppression enabling maneuver by other arms, no problem. Even just another vehicle at a closer range and for another angle, getting several more kills, no problem.

But when fire alone can solve any tactical problem - while it fits the firepower dominated, attrition attack tactics I personally favor and that I'm best at - I think it makes the tactics less interesting and less realistic.

I expect to need grenades (launched however) or mortars, in that situation. Or infantry getting quite close. If instead it is building cover, a tank main gun will serve. But just lots and lots of ordinary bullets, at men who should already be fully suppressed and aren't in the open but are in cover, I do not think should be able to shoot them down to a man. No matter how long I fire.

FWIW...

[Guest]

Exactly. If all problems can be solved (eventually) by fire, why maneuver?

Like I said before -- I don't see much *suppression* when heavy volumes of fire are flying about, I see it when casualties are taken. Steve's description of crossfires causing more "damage" lends itself to that observation as well. Crossfires shouldn't cause more casualties (seriously?) they should cause people to want, even more, to get the hell down and stay down.

I'm not sure that damage modifiers are really needed at all, seems like suppression modifiers would do the same trick and model a firefight more accurately. Why go to the trouble of modelling realistic trajectories when you're going to add *damage* when bullets are coming in from different angles (unless I'm misreading, of course)?

I think direct fires should cause more suppression, less damage.

['Card]

Originally posted by Phillip Culliton:

Just a thought: has anyone noticed troops becoming suppressed by fire that did *not* hit one of their number?

Sure. I've had that happen. Last night I had two Syrian guys... an anti-tank team up on a rooftop, in fact. A squad of US troops got to the second floor of a building a few streets away and put a mess of fire on top of my AT team. They were pinned for awhile, and then their status changed from OK to Cautious to Shaken, then I took out the American squad (tank round into the building) and the AT team eventually returned to normal - all that without either one of them taking any damage.

[Guest]

I stand informed. Thanks! I do tend to see more death and destruction prior to suppression, though. Hmm.

I'm amazed those two Syrian guys weren't killed in the first volley, actually. What quality were they and the Americans? Were you playing against the AI?

[Battlefront.com]

We are looking into specific circumstances that may not be correctly modeled, or are perhaps blessed with a bug Long range accuracy and combos of firing with movement involved are top of our list.

JasonC,

The problem with text examples is that I always have more questions than answers. However, in general I would say that the guys in the trenches should have "gone to ground" and stayed there if they were indeed protected from fire.

Phillip,

Crossfires shouldn't cause more casualties (seriously?) they should cause people to want, even more, to get the hell down and stay down.

Yes, but sometimes crossfire doesn't give them the opportunity to bolt. Druss brought up paintball so I'll give an example...

Our team found ourselves facing an equal sized force in a parallel line running down a hill. We sat there and exchanged a lot of paint and only a few people were knocked out. I was getting tired of this and, because it isn't "for real", I took it upon myself to go commando on their asses. I bolted through a weak spot in their defensive line and kept on running. I fired from the hip but that was just for suppressive effect. I had a big tree probably 40m behind their positions as my destination. I got there, turned around, and started firing at their backs. I picked off a few guys and the others, realizing that they were COMPLETELY exposed, had to turn around and fire at me behind a tree. I got another couple. While this was going on they were not firing at my comrades, which I screamed at to get their asses moving. They did, and the defenders broke. As they got up to run away they were cut down either by the bulk of our force or by me (best damned day I've ever had ).

We wiped out their entire force in probably under a minute. It was something like 20 of them knocked out, not a single casualty taken on our side. Had we stayed in the positions we were at it was anybody's guess which side was going to slowly chip away at the other. But because I got in back of them, adding a new angle to the fight, their positions became untenable. And because they didn't knock me out right away *and* the rest of my team charged, they had little chance of doing much more than picking a few of us off. They didn't even do that. I was hit several times but the balls didn't break on me. Which is why I would never try such a foolish stunt if I were a real infantryman

Steve

[Taki]

Originally posted by Battlefront.com:

I was hit several times but the balls didn't break on me. Which is why I would never try such a foolish stunt if I were a real infantryman

Steve

Ha! I knew you where cheating. Next Time play without "Godmode"

[Thomm]

Test result from yesterday: soldiers hiding in a trench cannot be hit from infantry at a distance of approx 150 meters.

They *can* evade by sneaking along the trench, however, when changing from 'Hide' mode to 'Slow' mode some of the soldiers went into the kneeling position and were promptly fired upon (until they dropped to prone again). Whether this is intentional or not I do not know.

At distances of 100 to 150 meters prolonged firefights are possible between prone teams over open ground. Seems to be realistic to me.

Best regards,

Thomm

[Guest Guest]

deleted per user request

[Battlefront.com]

What are the army advisors telling you guys about this?

That customer expectations are not inline with military reality This is coming from people that have seen hostile fire in person.

Now, this is not to say that there isn't some sort of combo of things that is out of whack. In fact, I've asked the lads to test a couple of things in particular. Instead, what they are saying is many of the things they see customers complaining about don't look wrong to them, just wrong to the customer.

We've outlined the reasons for this in other threads. In general, however, I'd say the lethality issues customers have are skewed by the following factors:

1. Experiences based on WW2 settings where Squad lethality is definitely a LOT lower than Modern. This is probably the main issue as I often see people comparing things in CM:SF to CMx1. Even comparing to CMAK isn't valid due to the differences in firepower and accuracy.

2. Comparing real life with a game experience. Over the past 10 years we've had countless, and very interesting, discussions about why game lethality is always higher than real life. CMx1 was absolutely no exception to this either, having casualty rates in nearly every battle that exemplified a handful of real life "outliers". Tons of issues come into play, including poor tactics, virtual soldiers too willing to die without question, generally worst case matchups (i.e. both sides have a chance to slaughter the other), artificial game motivation ("dude, I'm going to kill you Tiger if it is the last thing I do!"), God like knowledge of the battlefield, Borg like control of units, etc., etc., etc. Meaning, no matter what we do casualty rates will ALWAYS be higher in CM:SF than in real life because pretty much all wargames have the same problem.

3. Bad tactics. This is partly wrapped up in #2, but in a game (as opposed to dry test scenarios) people tend to make very costly mistakes without knowing it. This could be the result of any number of things, such as a lack of familiarity with a weapon system, misunderstanding of how much time it takes to pound an enemy squad into submission before moving, overestimating the relative safety of a vehicle on a particular street, etc. Some of these things can be "blamed" on the user, some on the user interface, some of the game mechanics (#2 above). But in the end it comes down to doing something that probably would not be done in real life, resulting in something that would not likely be seen in real life.

4. Comparing the wrong "real life" to the game. It's natural to compare what is going on in Iraq and Afghanistan over the last few years with what CM:SF is designed to simulate. The problem with that is the two share very little in common. There is a huge difference in "full tempo" conventional (or asymmetric) warfare and low-medium level insurgent asymmetric warfare. The latter tends to produce a steady stream of casualties over time, but no large number at one time. The former, however, can vary from producing few casualties at all to massive casualties in certain circumstances. It really depends on the nature of the fight. In OIF the Iraqis basically gave up before the shooting started, in CM:SF's Syria they decided to fight it out. In other words, CM:SF is a "worst case scenario" to start with, so higher casualties should be expected.

Somewhat valid examples to look at in Iraq are An Nasiriyah and Fallujah I and especially II. There are others, but these have been well documented and the US losses were significant in terms of lives, wounded, and heavy armor. Tons of small scale skirmishes are equally valid when the setup is similar in CM:SF (i.e. more asymmetric than conventional).

Steve

[ December 20, 2007, 12:07 AM: Message edited by: Battlefront.com ]

[Guest Guest]

deleted per user request

[Omenowl]

I thought the average for all wars before vietnam was 1 kill per 10000 rounds. This does not include wounded, etc. A sniper was at 1:1.7. Vietnam was 1 kill per 100000. That said we are also looking at a lot more bullets being shot by vehicles (no one ever said exactly which rounds were being fired).

I don't think the problem is with the weapons being too accurate, but rather the soldiers calmness under fire. It is very easy to shoot on a range. Try running with 50 lbs of gear for a mile and then shooting. Very different results. Most soldiers don't fire their weapons past 125 yards. So I think being accurate in combat is more the exception rather than the rule.

[aka_tom_w]

Originally posted by Adam1:

What is out of whack then?

I am not sure of this but if it is possible to isolate these specific factors (IMHO) for patching or fixing or tweaking purposes in the game engine but this what I see as out of whack:

i) First shot accuracy to acquire and hit a moving target

(that might be two issues)

ii) Meaning moving man size targets at 200m are not given enough "credit" in the game system for being harder to hit.

iii) More Chaos less calmness under fire.

iv) Sniper units could be more accurate

Good point here:

I don't think the problem is with the weapons being too accurate, but rather the soldiers calmness under fire. It is very easy to shoot on a range

More things like chaos and blind fire, and spray and pray, and soldiers simply not firing by pretending to just reload while seated with their back to the wall on the roof, (just not participating in the fight) would all add to the realism IMHO.

If the first round volley did not connect or hit (at ranges 100m and over) with such accuracy and lethality at medium to long range (more suppression fewer actually hits) there is a chance the target might have a better response or reaction to shoot back or find cover uninjured.

Its possible many of the complaints have to do with close range Urban "street fights" where close range, poor tactics and fully automatic weapons combine to increase lethality to much much higher level then might be expected. I would suggest the accuracy at close range in Urban environments (less then 100m) does not need to be fixed, UNLESS you would like to argue that and Soldiers (US or Syrian SF) with body armour should die less and end up with yellow bases far more often, but this aspect of the simulation might be OK now too. )

I believe it is possible that a little bit of fine tuning (where possible) can sort out some of these issues.

I think Steve is right on target with these comments and issues:

Tons of issues come into play, including poor tactics, virtual soldiers too willing to die without question, generally worst case matchups (i.e. both sides have a chance to slaughter the other), artificial game motivation ("dude, I'm going to kill you Tiger if it is the last thing I do!"), God like knowledge of the battlefield, Borg like control of units, etc., etc., etc. Meaning, no matter what we do casualty rates will ALWAYS be higher in CM:SF than in real life because pretty much all wargames have the same problem.

3. Bad tactics. This is partly wrapped up in #2, but in a game (as opposed to dry test scenarios) people tend to make very costly mistakes without knowing it. This could be the result of any number of things, such as a lack of familiarity with a weapon system, misunderstanding of how much time it takes to pound an enemy squad into submission before moving, overestimating the relative safety of a vehicle on a particular street, etc. Some of these things can be "blamed" on the user, some on the user interface, some of the game mechanics (#2 above). But in the end it comes down to doing something that probably would not be done in real life, resulting in something that would not likely be seen in real life.

[ December 20, 2007, 07:52 AM: Message edited by: aka_tom_w ]

[Guest]

A quick note before running my 6-year-old off to school:

I freely admit to not being an expert. I do, however, have a good dose of modern military reading under my belt, and a similar amount of common sense and knowledge of probability and physics. It seems to me that "outliers" should constitute a hit on the move at *50m*, not a series of kills on the move at *300m* (if I'm remembering that example correctly).

If the experts say otherwise, then obviously there's no accuracy problem. If not, then scaling back seems reasonable; the overall EFFECTS on which the experts comment may be right, it's the details I currently have an issue with.

Cheers.

[aka_tom_w]

OK I am not so sure 50m is a good starting point.

Now I could be wrong about this, but at 50m if a squad of 9 guys ran across the 50 yard line of an NFL foot ball field and you put an AK47 in my hands semi-auto, firing three round bursts, I would bet even I ( I would consider myself a reasonable shot, but I have no police or military training what so ever) could hit at least one of them before they got across the field. (I have never actually done that, of course, but it might be fun to try, so long as no gets hurt )

So my point is a squad of nine well trained soldiers opens up on a 9 man sized targets running laterally across their position at 50m there will be some hits for sure IMHO.

Now the same 9 guys firing at the 9 man squad running across their field of fire at 200m (thinking 2 NFL Football fields down range) should be a WHOLE other story, the result there should be MOSTLY supression with a low odds of a hit. The real question is what is "low odds" 1 hit in 10 shots? or 1 hit in 100 shots? or 1 hit in 1000 shots?

Its also possible I have no idea AT ALL what I am talking about, given breath and depth of military knowledge and first hand experience that routinely shows up here in these threads like this. (From soldiers, veterans and experts with first hand experience.)

[ December 20, 2007, 12:43 PM: Message edited by: aka_tom_w ]

[Guest]

Sorry, clarification: people firing while on the move, not people firing at moving targets. There was an example where a fire team on the move hosed a group of reds at 300m, IIRC.

I don't think that invalidates any of your points, just more food for thought I think.

[Omenowl]

My father had a friend who was a platoon leader in Vietnam. He said his unit was in a gully and they were fired on from 25 meters away by AK-47s. Not a single person in the unit was hit despite having the enemy shoot down the gully.

I have also seen demonstrations with tompson submachine guns, 7.62 fully automatic rifles and the ak-47 at targets. Only the first 2-3 rounds in a burst would hit the target at 30 yards.

I don't know about SAWs, M-16 or M-4s for their automatic fire accuracy, but I definitely think automatic fire is probably overly accurate at longer distances unless it is braced or setup.

[Neepster]

One game playability note to interject on this. The rest of the comments are very interesting and I tend to agree lethality is probably a bit overmodelled.

However, without this, given that squads and individuals do NOT surrender (unlike CM1), I anticipate the scenarios lasting crazily long times if the lethality is significantly reduced.

Thoughts?