Does different terrain add any protection from small arms at all?

[DaveDash]

It's definately 'niggly' trying to get your guys to move to the edge of a crest. Telling them to slow move to the crest results in guys all over the place - if they move at all.

The biggest problem is not everyone lines up on the crest like they do in real life, meaning the firepower of your unit is heavily reduced. To get them all exposed means some units are way over exposed and usually take heavy fire.

It would be nice if there were formation commands, column, vee, line, echelon left/right, wedge, 'scattered' etc.

[skelley]

The soldiers spread out on purpose. Unlike the olden days of muskets and pikemen, soldiers aren't supposed to be all clumped together in tight lines Sure, there are situations when this is an accidentally optimal alignment, but generally skirmish lines are uneven pretty much by design.

The pattern I get is generally something like this:

______Ridgeline______

0

00

0 0

0 0

0 0

then it goes somethin like this: First guy gets killed, while second row doesnt see anything. Move them up further, they get popped, and so on and so forth. I think a more optimal pattern would be maybe something like this:

0 0 0 0 0

0 0 0 0

At the end of a waypoint generally speaking you would want your firepower on somewhat of a parallel line to the enemy rather than perpendicular.

[skelley]

Hmm, my examples don't really look like I intended. The top circles should be spread out a little more and zig zaging.

-------0-------

-----00--------

---------0 0---

-----0 0-------

--------0 0----

That is how it looks:

--0---0---0---0---0--

----0---0---0---0----

[ December 10, 2007, 07:22 AM: Message edited by: skelley ]

[aka_tom_w]

the real issue is that a ridge line is not treated the same way a wall is or a roof top is.

Guys take up good positions along a roof top

send a squad up to a low wall and they spread out.

But a ridge line is a tricky thing for the AI to figure out and say "Oh this is (ridge line) where I need to spread out in a line at the crest because I may or may not have LOS and LOF to something that can shoot at me".

So the suggestion of a line a breast formation option or button (somehow) is not unreasonable IMHO.

I would like to suggest there may be a few changes in some other aspects of the game (not this exact example) that will mitigate how bad this is now when they release v1.05.

[skelley]

Yeah some kind of formation button would be helpful. I like the 1:1 but I do miss not having to think about all of this in CMX1 especially in RT.

[Battlefront.com]

No chance of a formation feature. That gets into micromanagement which, as skelley points out, is a detractor from the game. Better for us to work on the TacAI. And TomW is absolutely correct... some terrain features are subtle and it is more difficult for the TacAI to figure out what to do compared to other terrain types. Check out v1.05 and see if you notice improvements (you should).

Steve

[DaveDash]

Well I've noticed that giving your units the 'face' command in "open" terrain makes them spread out in a lineish pattern. I might test giving them a 'face' command at the base of a ridge then move up up onto the edge of the ridge and see what happens.

[MarkEzra]

Originally posted by Battlefront.com:

[Check out v1.05 and see if you notice improvements (you should).

Steve [/QB]

You bet! Happen to have a link?

[skelley]

I've tried the face command and it didn't do much. But I have noticed that if you have infantry that are stuck you can get them unstuck by clearing orders then giving the face command, then reissuing orders. This works pretty good for the bug that makes the remains of a squad run the opposite direction of the waypoint you set.

[DaveDash]

Actually you're right. The face command only puts them in a nice line during setup, not afterwards. Problem is during movement they all bunch up into a group and lose their nice line formation.

I did have much better luck by splitting the squad.

Full Squad:

Split Teams:

Ahh that tip is extremely handy. It's frustrating to have an entire squad stuck at different levels of a building, including outsid, and refuse to move. I resorted to splitting the teams and merging them up later to get around it.

[Dirtweasle]

Originally posted by Battlefront.com:

.... Check out v1.05 and see if you notice improvements (you should).

Steve

I tried but only 1.04 appears available to DL at the moment.

[skelley]

Spliting them looks like it works pretty well. That middle team looks like its in a perfect row. Wonder why the outer 2 aren't.

[DaveDash]

I did a test under combat situation (a house with an enemy squad over the rise) and it still ended up reasonably bad no matter what way I did it. Splitting however I managed to get enough fire to bare upon the bad guys to eventually surpress them after taking about 3 casualties.

I tried throwing smoke to get onto the crest, but the wind blew it away and half my squad got chopped up while the other half couldn't see the enemy to shoot back, apparently.

Having a full squad means less casualties initially, but youll never get enough men on the ridge to surpress the enemy, so in the long run you'll either have to stay down there or suffer more casualties as your guys move onto the ledge 2 at a time and get cut down.

[reel-why]

I've been using face to organize troops in a line and its worked reasonably well.

I think the smoke thing might work well if you split your teams and have each pop smoke. Although I'm not sure all infantry teams have enough grenades for that. I've done it with one of the HQ or specialty squads though.

Or alternately you could always just take the hollywood option. Tell your troops to book it as fast as they can over the ridgeline so the heroic team leader and spray wildly on full auto. Works like a charm everytime.

[DaveDash]

Yeah I had each team pop smoke. The problem was it blew the smoke away and I had to shuffle across to get in the smoke. Then, the smoke wasn't really where it said it was during the orders phase, so one team ended up exposed while the others sat around twiddling their thumbs wondering why their buddies were yelling out.

[Drusus]

The real problem seems to be that when the ridge line is sharp, then a 1m difference in position will change a soldier in a good position to a dead soldier.

This problem is made worse because soldiers can't be given exact locations to deploy, and actually the squad will position centered to the 8x8 action spot and this might be just on the wrong side of the berm.

A neat way to solve this problem would be to have a "hull down" command for infantry. Give a forming point near the ridge line intended ("take position here") and a target point ("you must be able to lay down fire to this point"). Then the soldiers would automatically take good positions along the ridge line.

I think this would be a valid command for a "real life" commander to use. However this could result in gamey usage, as the information available to the player can be used to take unrealistically good positions.

As an example in CMx1 it was possible to position your tanks so that they can see pretty much only the single enemy tank from their hull down positions. This is of course unrealistic, as the tanks getting in the hull down position shouldn't have any clue where the enemy tank is, so ending up just in the perfect position should be unlikely.

Also, implementing "hull down" for soldiers is probably PITA. And I am not sure the terrain is actually fine enough to enable this kind of a position for a soldier. So another proposal:

Maybe it could be made so that soldiers near a ridge line should get a good bonus to the cover the terrain gives. Calculating when a soldier is near a ridge line is probably another PITA. And when a soldier is covering near a ridge line, I think after 5 seconds or so of covering the soldier should have gotten into a place of really good cover, that is behind the ridge line.

Actually, I think in general when a soldier has been covering for a while he should be considered to have good cover. The amount of time needed should be variable, pretty much instant perfect cover when in a trench, fast and good when in wooden terrain or rocky terrain and slow and bad cover when on a road.

Of course experience should have a big factor on both the time needed and the amount of "extra cover" gained.

Actually I think the absent of good cover is what is behind the complaints of too much accuracy.

Foe example even I could hit a non-moving target the size of a football from 150m given 30 seconds of time and a few bullets to use. Now this means that a soldier just "covering" in the open should be easy enough to hit, hence the accuracy isn't wrong. But I think that at least a well trained soldier (US soldiers, that is) would not be exposed for too long. He would find a place giving good cover. Actually I would also claim that an experienced soldier would do very much to not get into a position where he can't find good cover fast enough. This means that when he is moving he constantly moves so that good cover is near enough.

I am having a hard time imagining how to implement this in the game so that the TacAI would actually calculate the positions with good cover and move into there, especially given that the squad will want to be in the same 8x8 action spot. To solve this I think an abstraction should be used. The soldier should be covering for a while, and then end up in a position where he has very good cover. There could be some graphical feedback that the soldier has found good cover, but the soldier shouldn't necessarily move during this covering.

I know the cover in this game is already abstracted so this wouldn't be something breaking the WYSIWYG of the game. Also computationally this should be easy, and I imagine this would be rather straightforward to implement. Of course not knowing the code means I can't really know how hard implementing this would be, nor do I know how much time it would actually take to implement...

Once more: I really do hate that so many times I see a veteran US squad covering for long periods of time, and the whole time the soldiers are getting killed one at a time. Now, I am pretty much sure that in real life, the soldiers wouldn't just wait in a position where they can be hit. On the other hand the terrain in the game isn't perfect enough to actually show the places of perfect cover, nor is the TaCAI / infantry representation advanced enough to actually find these spots and maneuver into them.

I don't expect the engine to be able to represent all this in WYSIWYG way. Neither does BFC AFAIK. So, just allow the soldiers to find the spots of great / perfect cover when they are covering. And I do think that the current way of covering soldiers just waiting to be shot is broken.

I would really like to get some feedback from those who know more about all this, that is those who know how this works in real life...

DISCLAIMER: It is 5 AM here...

[Battlefront.com]

The way things worked in v1.04 is not relevant to the way they work now. When you get ahold of v1.05 (you wise arses above know who I'm talking to ) you may not see things working as you want them to, but I know you will notice improvements. There are a couple of things that you guys are struggling with at the moment:

1. Some problems with Soldiers and Teams being "stuck", either for good or temporarily. This can be the source of guys being separated too much, though it could also be combat conditions making some guys stick to where they are instead of maneuvering to be with the rest (this is realistic and obviously would be wrong to prevent).

2. Some doors are busted in v1.04. It is extremely hard to say which ones, though, because for the most part the different types are not visible to anything but the guts of the code for purely code reasons (i.e. not gameplay). This can further screw things up in terms of guys going where they shouldn't.

3. Guys don't do a good enough job trying to get into LOF when they need to be shooting, out of LOF when they need to avoid being shot.

The latter is where the crux of the "ridge" issue lies. It is also where a number of other possible issues might be too. The problem in v1.04 is that all soldiers in an Action Spot are theoretically in LOS with a particular target, but not necessarily in LOF (this is a big issue which I've discussed before). Soldiers not in LOF tend to stay that way more than they should, hence why you can get a problem with soldiers on a ridgeline not uniformly attacking or hiding from the enemy. This is a TacAI issue that has been worked on and no doubt will have to be worked on some more over time.

So the answer to the problem of guys not pulling their weight is not to introduce formations, nor is it to put in new Commands. These involve additional micromanagement and in some cases won't fix the problems satisfactorily either. What CMx2 needs is for individual soldiers to get into better firing positions or cover, situationally dependent. Hopefully you will notice improvements with v1.05 and more in the future.

Getting back to terrain for a sec here...

Remember that 8m by 8m is not a big space. It's smaller than the footprint of most of the houses you guys are sitting in right now. Sure, out in a field or a span of desert there is some degree of cover within that space. As I said, CMx1 and CMx2 simulate such things abstractly because there is no way we can simulate them literally. However, keep in mind that there is only so many places within that 8x8m space that 5-9 men can be. 3 might be in the best bits of cover that are provided, 2 might be in fairly poor cover, and another 1 or 2 might effectively have no cover at all. When the area is swept with fire you should expect the ones with no effective cover to become casualties, the ones in poor cover to possibly suffer some, and the others to be OK for a space of time (at least). I mention this because it would seem that some of you think that there should always be enough "good" cover in an open space for everybody to happily hide behind. Maybe in a churned up floor of a temperate forest there would be, but in an arid environment... not inherently, no.

Steve

[hoolaman]

Excellent thread so far, with a few issues I have been wondering about. Drusus is right, a supressed enemy hiding in a solid trench or foxhole, whether natural or manmade will never be damaged by a distant shooter if they keep their heads down. You would have to manouvre to HW range, or more likely in the US case, bring massive firepower in and level the whole place as I've seen in many Iraq videos.

Maybe suppressed guys should get a cover bonus.

Originally posted by thewood:

How are the small undulations in terrain handled. I don't think the editor is fine grained enough, plus the major pain in the arse it would be to draw them. I am talking the small rise and fall in ground maybe just enough to conceal a prone person.

This is a big issue if you ask me. I think the 8x8 tiles could be made to give satisfying results, but as others have said, that will show the game as far from WYSIWYG.

If a map-maker wants to purposely put in lots of small lumps and bumps and ditches smaller than 8m, they cannot really do so. Nor is it clear that they can pick terrain tile X+Y to give the same effect and be sure that the cover abstraction will work out, because nobody knows how it really works.

I think people are fine with some abstraction, but at the moment neither solution seems to be possible.

Originally posted by Battlefront.com:

Remember that 8m by 8m is not a big space. It's smaller than the footprint of most of the houses you guys are sitting in right now. Sure, out in a field or a span of desert there is some degree of cover within that space. As I said, CMx1 and CMx2 simulate such things abstractly because there is no way we can simulate them literally. However, keep in mind that there is only so many places within that 8x8m space that 5-9 men can be. 3 might be in the best bits of cover that are provided, 2 might be in fairly poor cover, and another 1 or 2 might effectively have no cover at all. When the area is swept with fire you should expect the ones with no effective cover to become casualties, the ones in poor cover to possibly suffer some, and the others to be OK for a space of time (at least). I mention this because it would seem that some of you think that there should always be enough "good" cover in an open space for everybody to happily hide behind. Maybe in a churned up floor of a temperate forest there would be, but in an arid environment... not inherently, no.

Steve

Not inherently, but if a map-maker wants to make it so basically flat open terrain has a degree of lumpy cover, I understand there is no way to do it?

8x8 is not a huge area, but not what I would call small either. More often than not I guess a squad would IRL be spread out over more than one 8x8m square, and only a few men would need to find a ditch to lie in.

I imagine a "rough terrain" tile that just roughs up what is there (like shell damage) without changing the terrain type and gives a cover bonus.

Maybe a "large rocky" tile would be good too. It would also be handy to choose whether terrain slope in a certain area is smoothed or terraced.

[Other Means]

Can I jump up and shout "Look at me!" for a second again?

It might look OK if after a certain amount of time, due to experience and terrain etc, the infantry started to "sink" into the terrain a little with a concomitant decrease in chance to hit, to show they're taking cover?

Have them sink in various degrees to show the distribution of cover within the tile and it might get the point over quite well, IMHO.

[Drusus]

I admit that there shouldn't be perfect cover for everybody everywhere.

As I see it a soldier who isn't in cover will get shot eventually if the shooter isn't under enemy fire. This is modeled nicely in the game.

Also, the first thing a trained soldier will do when receiving fire is either shoot back or then seek cover. This isn't correctly modeled in the game.

As it is now, soldiers under fire will either be covering in a bad spot or then seeking cover crawling in the open. What I mean by covering in a bad spot is that the status of the soldier is changed to covering, but the soldier is actually in a place where there isn't enough cover to get behind of.

Firing back as a reaction is rare, as it takes too much time for the target to see the firer.

I still think that the proposal that there should be a new state of seeking cover (or to change the covering state to be that, too) is a good one.

The best way of course is that each soldier would individually seek the cover, that is their representation in the 3D view would actually move behind cover. However I think this is impossible due to the fact that the game can't actually (at least ATM) show the places of good cover visually nor is the infantry model and TaCAI advanced enough.

So, to me it would seem the best option is to let the soldier seek cover in an abstracted way. No need for really complicated TaCAI decisions, no need for the scenario designer to put in countless places of some cover.

Maybe there isn't enough cover for the soldiers in that 8x8 action spot. But that is a restriction made by the game, not by reality! Actually I think this is one more point for the proposal. This is because the _game_ limits the soldiers to be in one 8x8 spot. Realistically the ones who wouldn't find good cover in that spot would just go behind that rock in the next spot. The game doesn't allow the player, nor the TaCAI to do this. Making it so that it allows this seems really hard to implement. Simple solution: let the soldiers find the cover even in the next spot in an abstracted way!

By the way, I don't think that the game allows even 3 man teams to find really good cover even in the best spots of the map. Good cover yes, but still they are very much vulnerable to massed small arms fire.

I hope that 1.05 doesn't fix this, as then I would need to find something else to bitch about To be a bit more serious, I do think that the game is a good one, and I whine because I want it to be even better. And I really do think this proposal would add to the realism and playability of the game.

[DaveDash]

Excellent post Drusus.

[skelley]

I agree too Drusus

[Battlefront.com]

Hoolaman,

Not inherently, but if a map-maker wants to make it so basically flat open terrain has a degree of lumpy cover, I understand there is no way to do it?

Sure there is Rocky gives a bit more protection than Dirt, but (as I've said before) only when the unit is prone. If you really want to spice things up, use the Direct elevation tool and change the heights of individual tiles all over the place. This will make the 8x8 space depress or rise relative to the surrounding tiles.

But all of this is rather academic. In real life it is rare to have "basically flat" terrain. Even in Arid environments there are at least subtle undulations in elevation over a significant area. This means that LOS/LOF isn't possible from one spot to all spots on a map at all times. Well, unless you get high enough. And when you do that the small pockmarks and what not in terrain don't amount to much cover anyway.

Druss,

The solution to everything you mentioned is to have the individual soldiers react to LOS/LOF more independently than they do in v1.04. This is an improvement you should notice in v1.05. I wouldn't go so far as to say that v1.05 will fix everything all the time, but it certainly does make a big difference. We have some ideas on how to improve things further in future patches, but we want to get v1.05 out ASAP so they will have to wait.

Steve

[skelley]

The solution to everything you mentioned is to have the individual soldiers react to LOS/LOF more independently than they do in v1.04.

That will be a huge advance in gameplay. I tried adding fallen trees and stumps and rocks and the infantry still didn't use it as cover as it is in 1.04. They mainly died or tried to crawl away.

[JasonC]

I have a practical suggestion to teach the AI how to seek cover better.

In CMx1, cover was an inherent aspect of the terrain type in the tile. This meant it was pretty easy to get a field of cover values for the AI to react to. It just saw 10% trenches and 12% stone buildings and 15% woods and 25% rough etc. And went for the nearest spot where that number was better than that number where it was.

What is missing in CMx2 is a good value field for cover. Cover is now largely a function of outright LOS, and not of the inherent nature of the terrain the unit is standing on. And LOS is not a unitary number at each point - in physic speak, a scalar field - but a binary relation between locations. If you had to calculate anything about it in real time, it would be too hard.

The solution is to calculate a cover field from the LOS binary, but to do so only once, at game start. After that, crude adjustments to the "LOS field value" can be made to incorporate some current info - and this can be made easier at the set up calculation phase.

What do I mean by calculating a cover field from LOS, once? I mean for each 8 by 8 location, one counts the other 8 by 8s that can see it. Low number means good cover - most of the map can't see that spot - high number means poor cover - most of the map can see that spot. This is a calculation that is quadratic in the number of tiles, and therefore you don't want to do it over and over again. But you don't need to really - the terrain form is not going to change appreciably. (Blown down buildings can change it some, it is true - but if the AI doesn't react perfectly to only that, it would still be a vast improvement).

What do I mean by weighting for current information? Well, the overall visibility number is only an approximation to real cover value, because it treats all locations equally. So weight the seeing locations a few different ways and store the resulting fields. A few different ways meaning (1) all equally (2) locations in friendly set up zones get no weight while locations in enemy set up zones get double weight or (3) weighted by how close they are to friendly edges or (4) weighted by absolute map quadrants or finer subsets (ten by ten divisions, say).

Then you can incorporate present info by multiplying the quadrant-weighted visibility field by some count of known enemy in that quadrant. This would effectively ignore the visibility of a piece of ground to locations where there aren't any enemies - or the cruder but simpler cases, ignore visibility from your own side of the map or set up area etc.

Within a turn, one single field is found from these precalculated global numbers for that map, perhaps multiplied by an enemy presence vector. Then this cover measure is used just like the old CMx1 visibility percentage at each location.

This avoids any need to solve arbitrarily complex LOS calculations on the fly, during the CPU crunch time. It makes use of extra "free" CPU time at game start. And it would "spot" all the typical "hollows" and "shadows" and other bits of useful dead ground, and use them as easily as the old system used blocks of woods.

I hope this is helpful.