Has this happened to you?

[Nidan1]

...............

I would be interested in seeing how the game played if we were able to split squads into fireteams and fireteams into buddy teams.

You can basically do that now with most German infantry formations and some Russsian ones. How small would you want to go? You can have two man teams now.

[JasonC]

Of course it is a fixable problem. There is no reason for a single short burst to hit 4-5 men if realistically it would only hit 1.

[sburke]

Of course it is a fixable problem. There is no reason for a single short burst to hit 4-5 men if realistically it would only hit 1.

everything is "fixable" given unlimited resources and time. Neither of those does BF have. Whether the current positioning of units in an action square is "fixable" is something that I think only Charles is qualified to answer. Whether he agrees that it is so inherently broke to require he work on it I highly doubt.

beyond that I agree with what a number of other posters have said. Basically it isn't broke enough to lose sleep over. If you want to micro manage more, you can break units into teams further limiting the issue.

In the above mentioned example from Pelican Pal. You have a building with two windows. So maybe one guy shoots from each window? Is that the proposal because then only one guy from each window could be hit....would we be happy with that scenario, an entire squad on the 2nd floor and only two guys shooting? I know I would take the current dilemma over that option.

[Pelican Pal]

In the above mentioned example from Pelican Pal. You have a building with two windows. So maybe one guy shoots from each window? Is that the proposal because then only one guy from each window could be hit....would we be happy with that scenario, an entire squad on the 2nd floor and only two guys shooting? I know I would take the current dilemma over that option.

Well really the number of windows/doors a building has should limit the number of shooters it can support in some way. An infinite number of men cannot shoot from a building concurrently.

But like I mentioned above I think this problem is too big for BFC to handle with their current system. The action square system has some pretty significant limitations, but those limitations let the game run as well as it does.

[YankeeDog]

You have a building with two windows. So maybe one guy shoots from each window? Is that the proposal because then only one guy from each window could be hit....

Several problems with this:

- In many situations, more than one guy can shoot through the same window. Proper MOUT technique is to set up firing positions inside the building, ideally several meters away from the window openings, and shoot through the window from a distance. Not that difficult for two soldiers to do this through the same window, especially if shooting at different targets.

- Especially in lighter construction, loopholes and other non-window openings are used when possible. A simple two-course brick exterior wall can be loopholed fairly quickly with a couple bursts of MG fire. And, of course, there is the fact that buildings in combat areas tend to develop additional apertures...

- Many common types of construction are not proof against full rifle-caliber fire, so even if some of the soldiers are not actively shooting out of the building and are assumed to be sheltering deep inside the building, they cannot be assumed to be completely safe from even small arms fire. Nevertheless, some buildings would have shelter areas impervious to all but the heaviest coming fire (ex: basements), but units sheltering in such areas would also usually have pretty much zero situational awareness, so really they should have no LOS/LOF out of the building and also probably suffer severe C2 penalties (how would a unit hiding in basement communicate info and receive orders from a nearby HQ?).

Overall, urban combat and particularly combat involving buildings is extremely complex, and the game model is going to have to involve a lot of abstraction in this area for the forseeable future. Overall, I think CMx2's current fairly abstracted system for buildings (but considerably higher fidelity than CMx1) works fairly well, albeit with some rough corners here and there.

I would eventually like to see some improvements, but I think steady incremental improvements is the way to go. For example, I think it's worth examining whether an explicit modeling of basements (with all their advantages and disadvantages) can be added to the game engine relatively soon ("relatively soon" meaning within the next 3 major game engine releases or so).

[JasonC]

There is nothing wrong with the action spot system as long as its idealizations are taking into account in other aspects of the game system. The designers already recognize this in issues like slightly "nerfing" HE effect. They know that a literal kill zone radius applied to the literal presentation of soldiers on the screen would result in too high a lethality per shot fired by HE weapons, and that the result would be tactically less realistic than their nerfed version of the casualty chances within a given blast radius.

Being engineering literalist about one thing when you can't be equally literalist about another thing that is causally linked to it in reality doesn't result in heightened realism. It results in decreased realism - it "stresses" the idealization you are forced to use about the second item. Clear issue, but not one without a solution. Just one that actually needs to be seen, acknowledged, and addressed.

How must it be addressed? Answer, preservation of the tactical relationships and realism about them is fundamentally more important than literalism about weapons effect measurements exterior to the game, that only interact with the game through this subsystem. A bit of design for effect must be inserted to keep the lethality correct, when it is being artificially boosted by having a denser target than it actually would have in reality, by the limitations of the action spot system.

Clear enough principles, and already acknowledged and acted upon in the case of HE. (Maybe not enough, arguably - it is still pretty darn lethal - but the tweak in that direction was made because the issue is understood).

The same exact relationship between artificial target density and fire effect exists for small arms, specifically in the relative importance of *aim* and accuracy, vs volume of fire. Denser targets than really exist boost the fire effect of automatic weapons and reduce the relative effect of aimed ones. When there is no correction between them, the parameter value that seems roughly right for fire accuracy overall, will still be too low for the single aimed shot and too high for the burst fire. Because there are 5 men in the target area where in reality there would only be 1-2. (I exaggerate for the purpose of clarity).

There is the same solution. Adjust the orientation of the shot - the smallness of the angle through which the shot is randomly distributed - to a narrower cone than used presently - an accuracy *increase* - but then give every bullet beyond the first a chance of missing / not being resolved, increasing with burst length. Over a long burst, the effect could be cut by as much as half (because target density is up to twice what it usually is in reality). So a 25 round burst might only resolve like 12 fired shots.

This gives full control over the importance of *accuracy* vs the importance of volume of fire, and thus a full "correction term" for average target density, with one single additional parameter - the rate of fall off of additional rounds in the same burst. Set it at 0.9 and you fire 1, .9, .82, .73, .64 etc. Set it at 0.5 and you fire bursts as 1, 0.5, 0.25 etc.

Then you tune. Get a realistic fire effect from a unit of pure rifles firing at given target, adjusting the accuracy / cone of dispersion of the shots as narrow as it needs to be to get rifles right. Without worrying about how high you are driving the lethality of machinegun fire. Find the right setting for the cone of dispersion of all shots, for the rifles, and leave it there.

Then with that setting fixed, simulate MG fire as bursts against similar targets. The lethality will be too high, because the target density is. Do not correct this by widening the cone of dispersion of all shots, leave that be. Instead, adjust the "fall off" of subsequent rounds in each burst as discussed above. Then take as many terms of that fall off series as your burst length, sum them, and *resolve only that many shots" down your random cone of dispersion.

Example, your SMG fires 10 round bursts and you are trying the 0.8 setting for the automatic fire parameter. The SMG fires 1, .8, ..., .13, sums to 4.46 rounds fired. Fire that as 4 plus a 46% chance of a 5th shot.

You are resolving fewer shots than are actually ticked down on the ammo counters. But we know those shots are passing through a denser target than would actually occur on the battlefield to "intercept" them. The two factors will cancel - and they effect nothing else. Ammo consumption doesn't go down, rifles are not adjusted as being too inaccurate, etc. The "nerfing" involved varies in strength with the length of the unaimed burst fire, because it is precisely such unaimed additional bullets in the fire cone that are unrealistically boosted in their fire effect by using too high a target density.

That is how you would fix it. It is entirely parallel to fixing overly strong HE effect from too tight clustering in each action spot. Of course it means you have to experiment to find the right tuning - adding any parameter does. But that is all it takes - no changes to the action spot system, no changes to the one to one shot resolution algorithm (you just call that fewer times). There is one extra function, in series, ahead of the shot resolution for automatic weapons fire, which just "downshifts" the number of times that shot resolution is called by a calculated amount - that is all.

In return, you get independent control of the average fire effect of individual aimed shots vs area-burst, small arms fire.

[JoMac]

Thanks, JC, for that detailed and in depth analysis of what may be needed to make the AS system alittle better. CMx2 is built around this AS System, and anything that a member can contribute is always a good thing.

As I was reading your post I had similar thoughts & ideas awhile back, but never could figure out how I would format it in a Thread or Post...I for one, would probably discombobulate things and frustrate the reader.

[JoMac]

As a follow-up on my above post...

I know that this has been discussed before in previous Posts & Threads over time, but will illiterate again...

I always thought the lethality of Small-Arms & HE in CMx2 are alittle to much and needed to be toned down abit to better reflect RL.

And I don't mean how aggressive we as players use our troops in a game, but rather the actual Combat effects that Small-Arms & HE cause at the micro-level...This is where JC explains it.

At the moment, I think SMG's are alittle to lethal, the use of Hand-Grenades are to liberal ( tossing a couple nads per squad per turn would seem better ).

I do think, however, that Rifles ( Bolt/Semi ) seem to work well ( could use alittle nerfing, but maybe not ), and the toned downed effect of HE is better.

[Michael Emrys]

...will illiterate again...

You are too hard on yourself. The majority of your posts are tolerably literate, at least no worse than average for around here.

Michael

[Nidan1]

To use JoeMc67s words, JasonC's last post has totally discombobulated me.

I don't profess to be all that knowledgeable on how software code translates to actions in the CM world, but really how far down into the weeds did he go?, Are there a dozen posters in this forum that know what the heck he is talking about?

[YankeeDog]

The short version is that he's suggesting assigning a progressively increasing "to hit nerf" to the second and following rounds of bursts of automatic small arms fire, which would have the net effect reducing the overall casualty-causing effect of automatic fire vs. aimed single shots, while keeping other effects, chiefly suppression, more or less the same.

[JoMac]

The short version is that he's suggesting assigning a progressively increasing "to hit nerf" to the second and following rounds of bursts of automatic small arms fire, which would have the net effect reducing the overall casualty-causing effect of automatic fire vs. aimed single shots, while keeping other effects, chiefly suppression, more or less the same.

In short, that is correct.

[JoMac]

You are too hard on yourself. The majority of your posts are tolerably literate, at least no worse than average for around here.

Michael

Your Good, Michael, Real Good...Yeah, figured I would mispell my meaning, lol.

[JasonC]

YD understood me correctly.

My proposal would make burst fire from any kind of automatic less effective than the same number of bullets fired as individual aimed shots, by rifles e.g. It would do so without requiring any change to the action spot system currently used, and also without requiring any change to the present system used to resolve whether any individual shot hits or misses. It also ensures that firing more rounds with an automatic, is always more effective than firing fewer rounds - it just has that effectiveness "scale" less than linearly in the number of rounds fired.

And it describes how to "tune" the overall scale of this effect.

The reason such an adjustment might be needed is the subject of earlier discussion in the thread. If people get the impression from playing CM that bolt action rifles are useless - when in real life bolt action rifles are not useless at all - then there is something out of whack. Previous posters diagnose that something - correctly in my opinion - as tightly bunched troop deployments making automatic fire way more effective than aimed fire - more so in CM than it is in reality.

So, the reasoning is...

Initial symptom - players want their pixeltruppen to throw away their bolt rifles because they consider them useless.

Bolt rifles are not useless in reality. Players think they are useless in CM.

First conclusion - some of the true effectiveness of bolt rifles in reality is missing in CM.

Why does this happen? Second observation - the action spot system results in unrealistically tight groupings of the soldiers on the field. They stick right next to each other to be commanded as one unit.

When an automatic weapon burst gets close enough to one of those tight knots of men, it doesn't just get an elevated chance of hitting someone. (It gets that too, of course). It also gets an elevated chance of hitting several people. This isn't coming from automatic weapons being the best thing ever in real life. It is coming from the soldiers all hugging each other in CM - a game engine limitation of the action spot system.

Desired solution - turn up rifles while turning down automatic weapons.

Objection - they all use a similar fire resolution system in which the aim of the weapon is explicitly modeled as some angle and the actual path of each bullet is tracked. The designers don't want to redo that system - it is a considerable advance in realism and they want to keep it.

Granted.

It doesn't stop us from turning up rifles and turning down automatics, to counteract the increased fire effect from unrealistically high soldier density in the occupied action spots.

That is the point at which I got programmer-technical, and explained how to have both.

We treat a burst of 10 bullets as firing less than 10 bullets. We resolve each bullet the same, we just only resolve 4 or 5 or 6 of them, not all 10.

Where did the other 5 bullets go? Between the wider gaps that should exist between our pixeltruppen, that aren't shown correctly in the game because all our pixeltruppen are clumped onto one action spot. We know that *should* be happening - that our soldiers should be farther apart and more of the burst bullets should be missing by passing between our soldiers. So we just don't resolve those 5 shots out of the 10.

Any more clear?

[JoMac]

Altho, I still think Rifle fire works pretty good as is, and only automatics need nerfed down ( I've seen a few instances in CM where a single Rifleman hit two troopers within two turns )...But, anyways.

There are other good resaons to consider the idea for Automatics having say 5 bullets out of 10 for registering hits in an AS:

bullets will pass through units ( as you mention ), bullets might have 1-2 hits on a single troop ( not sure if modeled ), prone troops may use other dead buddies or indulations in the ground for cover ( not sure if modeled ), troops sense where bullets are whizzing by and take appropriate cover, Troops peaking around a corner of a building instead of having to be in open AS to spot, etc, etc.

Basically, Troops in RL are doing alot more ( then just the static animations you see in CM ) then just sitting in one place for any amount of time within an AS...Thus, giving overall less causualties.

[Nidan1]

edit.....

Any more clear?

JasonC

I think your second explanation was a bit more understandable for me, and I appreciate your willingness to clarify.

I am only speaking for myself when I say that your earlier post was about as clear to me as a lecture on quantum mechanics.

I understand the real life effects of accuracy vs. volume in close infantry combat with small arms, but translating those real life effects into computer software algorithms that attempt to duplicate those results is beyond my understanding.

BTW what's a "nerf", and don't tell me its a spongy dart.

[YankeeDog]

, and don't tell me its a spongy dart.

http://www.urbandictionary.com/define.php?term=nerf

[Sequoia]

I vote this the most improved Combat Mission thread ever.

[Michael Emrys]

To use JoeMc67s words, JasonC's last post has totally discombobulated me.

I don't profess to be all that knowledgeable on how software code translates to actions in the CM world, but really how far down into the weeds did he go?, Are there a dozen posters in this forum that know what the heck he is talking about?

I found it clear as a bell and a good suggestion towards a solution. I think the ideal solution would involve a change of code to get the pixeltruppen to spread out more, but failing that Jason's idea strikes me as elegant and powerful. Whether BFC will take it up, go down a different path, or ignore the question altogether is up to them.

Michael

[poesel]

While I agree that Jasons idea is good and elegant it is probably difficult to implement.

Why? Because you design software to a certain end and you have parameters in your mind when you do. When a pixeltrooper fires his gun there is aiming, calculation of the trajectory, counting down ammo, playing a sound, testing for intersection with objects and resolution of target effects. Usually all this happens all the time for all bullets. So you build efficient code that does exactly this. Especially it happens quite often.

But now you do all this but some bullets are not able to damage the target. First you need to decide if the bullet you fire right now is real or not. So you need to keep track of how many bullets have already been fired in this burst and how many were real or not. Then you need to flag each and every bullet if it is real or not. Then you have to test that again upon impact.

Doesn't sound much but it is some extra bookkeeping and branching which has to be done often in the game. All this costs CPU cycles.

Another aspect: an assumption is that all this is only necessary when you shoot at infantry. But CM does not know that before the shooting starts. So if after all the effort the bullet has hit a vehicle it will have an effect - real bullet or not and the effort was wasted.

Not to throw everything out the window - there maybe is a 'cheaper' way to achieve a similar result. Turn the procedure around and resolve the issue at the target.

Leave the shooter and target resolution as is. But when a soldier is hit and he is hit by a SMG bullet he gets a certain chance that he won't be injured by the hit. This has the same net effect than firing 'ghost' bullets but needs less effort.

I'm assuming that CM remembers the type of bullet when it hits but I guess it is like that. Different bullets have different effects so it is necessary to know that at impact time.

The advantage of this is that it has to be done less often than the other way round. It is an additional branch to consider but the number of times a soldier is hit is a lot less than the number of shots fired.

[Nidan1]

I found it clear as a bell......

Michael

I would have expected nothing less of you, since you are no mere mortal, but a man versed in all manner of things both scientific and cultural, far superior in raw intellect than the other small minded, ignorant fools, like myself, who inhabit these digital threads.

[sburke]

While I agree that Jasons idea is good and elegant it is probably difficult to implement.

LOL that was actually might point in saying only Charles really knows. JasonC has a possible answer for the situation, You have another. A third option is the placement of pixeltruppen within the AS (though based on some other items that have been noted with pixeltruppen behavior I expect that one is a really tangled web.)

Point is none of us really know the inherent issues with any of our potential answers. Doesn't hurt to toss them out there. Maybe one of em will catch BF's eye and give them a direction they hadn't thought to go in. To characterize any of them as easy or say they don't affect current code however is just opinion based on ignorance. As you noted in your comments on how the code is designed, it may change the viability of any of the ideas suggested.

Still it can't hurt. As JasonC noted MG behavior has been changed as well as HE. It is certainly possible this could be addressed as well.

[Noba]

Wrong thread.

Noba.

[c3k]

I like the idea behind the proposal. However, are auto bursts really too effective/accurate?

The solution proposed nerfs follow on shots to account for unnatural AS bunching.

Counter:

1. Green troops bunch more than veterans. How will that be modelled?

2. In a vehicle, the troops are realistically spaced, e.g. a truck cab. If we nerf auto bursts, action spots may feel right, but passengers in jeeps and trucks will gain immunity.

Those are 2 quick examples of unintended consequences.

Ken

[Michael Emrys]

I would have expected nothing less of you, since you are no mere mortal, but a man versed in all manner of things both scientific and cultural, far superior in raw intellect than the other small minded, ignorant fools, like myself, who inhabit these digital threads.

Now that is truly great sig line material!

Michael