Kaunitz

Nobody wants to die. If you can't detect source of incoming fire, especially sniper, this can seriously reduce motivation of personnel to advance. 
Here for you the video with almost similar situation. 10 Aug 2014 - the probe to enter in Ilovaisk. Fighters of "Azov" advance through the sunflower field and turn under fire of enemy snipers and MGs. During all video fighters ask each other "Do you see them ? From where is fire?" They are shelling ruined building, the house nearby (at the end of video in that house ammo dumb is setting fire). On 6:12 famous "Azov" arniored KAMAZ guntruck "Prianyk" (eng."Gingerbread") is appearing. "Move 300 meters forward for better position!", order it commander. Later "Prianyk" will be damaged with incoming fire and will leave the battle (not filmed). Single armor on video - BMP-2 of 51st mech.brigade, which covered "Azov". It's continuously is breaking during the battle and at the end failed completely and was abandoned by crew. "Azov" tried to advance under fire, but lost 3 KIA. The same situation was with "Donbas", which adcanced left from "Azov". They encountered with heavy fortified positions and masked enemy riflemen in tree-plants, which couldn't exactly detect. Thus, again several snipers, MGs and 1-2 82 mm mortars have stopped advance of 200 men. In this day UKR forces lost 12 KIA (4 from "Donbas", 3 from "Azov", 3 from 40th territorial defense battalion "Kryvbas", 2 from "Right sector"). 
Big trouble of CM - almost immediate uncovering of position of groups, which open fire. As you see on videos, which I upload here, in most cases, soldiers can't detect point of fire and forced just to shoot in suspicious places. And of course if you will try to play like in real life in order to save own pixeltrooppen, you will not throw them through the fields on enemy invisible MGs.    
Better to see from 1:23
 

I had the below paraphrased summary in my notes.  I located the original link (bottom of this post) that I took the information from so it could be viewed in its entirety.  Thought you might find the last paragraph of the attached post interesting.  It is from 2008.   
CM's Action Spots are squares for the most part, customized shapes in special circumstances.  By default soldiers treat the direction ahead as the threat area. When the unit advances into an Action Spot it chooses positions that would be defensively good if the enemy were directly ahead of it. This can be overridden by a FACE Command.  In this way you can move a unit north and have it set up positions to defend to the south instead of the north.  Flavor Objects provide cover but not concealment. The degree of cover is dependent on what the object is. A soldier will get a lot more cover lying behind some tires instead of standing up behind a sign post 
 
 
 
 
 

As area fire plays a big part for deliberate grazing fire, I thought I could also point out a few things:
Lateral aim
From my experience, when an infantry unit (that occupies only 1 square) is ordered to area-fire on a square, each individual soldier in the unit will select a point within the target-square (which must be visible to him?) and fire at it. Naturally, this means that if the target-square is very close to the unit, the resulting "cone of fire" will be very broad/large. If the target-square is very far away, by contrast, the cone will be narrow. You can try it out very easily in the editor - if you let a unit area fire at a very very short distance (to the adjacent square), the soldiers will be firing randomly somewhere within 180° to their front. In some very rare situations, you might make use of this to affect/narrow or broaden the size of a beaten zone of a MG. By setting the target square closer, you broaden the angle of your fire and vice versa. Note that soldiers do seem to switch their target points within the target square from time to time during a turn. So the soldiers are not firing at a single point for the whole turn.
Vertical aim: 
Soldiers who are area-firing aim at ca 1m above the ground. (Depending on the angle between the height of the muzzle and the target point, bullets might travel on much farther after they have passed through the aimpoint) Still you can see some shots going too low or too high. I suppose that this is just due to random deviations from the perfect elevation. Soft factors (unit experience) might play a role here.  Rate of fire:
Among other factors (clip size/reloading intervalls), the output of fire depends on the distance to the target square. It's because the more distant the target is, the longer soldiers aim before they fire. So, as the duration for the "aiming" action increases, the intervalls between individual shots/bursts increase. Experience and other soft factors (and optics/weapon type?) probably play a role here as well.  

I like the thinking behind these ideas, but they are not that simple to use, and they are not really enforceable. I've seen good, intelligent, and honest opponents forget way simpler house rules.
Instead, these good design decisions should be implemented into the actual game system, forming the core of a new, optional difficulty level above Iron. Something which many players have politely requested for years.
One funny thing:
This is clearly an oversight by Battlefront. There's a whole realism level (Elite) only dedicated to turning the contact icons into "plain infantry markers" to prevent the player from knowing which type of support weapon is where. But they forgot that we can still click the enemy icon and see if it's a mortar or MG etc...
This oversight has persisted until the point where people are starting to make elaborate house rules around it - why not just fix it?
 

I thought I could upload the current state of the map here if anyone is interested in experimenting with how the game plays out when greater distances are involved. WARNING: This is by no means the final state of the map and it's not a proper scenario. All I did is to put some troops on the WIP map to get a feeling for how the game plays. It's only meant for experimenting, and you need to play both sides. 
https://www.dropbox.com/s/bk9rpz0bwzvr4d2/vast_map_test.btt?dl=0

Via Lindy Beige's latest video on tactical psychology https://www.youtube.com/watch?v=jRB9IU3Vt2g I came upon the book "War Games"  by Leo Murray which also featured an account of the battle at Longdon hill, which, in turn led me to this interesting video:
 

I observed the same thing with trenches. Under an artillery barrage, some troopers would leave the safety of the trench to seek cover in a freshly created shellhole. Needless to say that running through an artillery barrage is not that clever. 
I had to put my Gerbini scenario on ice because of soldier-placement issues and the lack of protection provided by defensive structures. It was impossible for me to provide the attackers with proper cover against artillery:  http://community.battlefront.com/topic/125278-highlanders-the-battle-of-gerbini/?do=findComment&comment=1748396 (also check out the links here: http://community.battlefront.com/topic/125278-highlanders-the-battle-of-gerbini/?do=findComment&comment=1747999
My impression is also that the engine uses some rather refined method to simulate the effects of cover. My go-to story is that I once created a very sophisticated MG position in CM:BS that provided excellent protection from small arms fire and direct HE fire. The MG was placed on/close to a log (a prop placed on the map). The soldier firing it was in a shell-hole and behind the log (similar to the position you can see here: https://youtu.be/ke4SbGZ7LJE?t=337), the surrounding area rather flat (to prevent direct HE from hitting close). Even though his position took a lot of fire, the firing team survived. The slightly more exposed MG, however, got knocked out. Up to this point, i didn't even know that MGs could get knocked out like that. (In this snippet, the MG hot knocked out shortly after the firer: https://youtu.be/ke4SbGZ7LJE?t=484)
 
 

haven´t read through all of this thread, but here´s my quick 2 euro cents.
All depends on general setup and location of HMG´s. If employed correctly on a map that offers it, have them more rearward so they can exploit there maximum ranges. Have them covered frontally for desirable cross fire opportunities (kyeholed). Have them supported by other friendlies that step in for support if they receive enemy fire. Move to switch position if necessary (enemy brings heavy support fire to bear on position). Protect flanks and approaches to them with other freindlies. And so on. Question is how to deal with this properly setupped HMG, not a single haphazadly placed singe one. This is parts of german WW2 tactical doctrine. Other nations may differ, but i rather think not so much.

From quite a lot of experience on messing around with fortifications (I think i may have some sort of dug in troops fixation) I am pretty sure that the troops do need to be actually inside the foxholes to get the cover bonus. Most casualties seem to occur on men left outside the foxholes or moving out of the foxholes on their own  so I am pretty sure it is important to try to get the men actually in to the foxholes to give them any extra cover. What is happening in the picture: I would say that the road bank is overriding the foxholes as preferred position for the men. Much like the already mentioned elevation slopes and other terrain features. I have argued in the past that the problem seems to be the prioritization by which the pixel troops select their positions inside action squares. Number one priority is to place troops next to buildings or walls/hedges and this is why you need to be really careful when using foxholes or trenches close to buildings or walls. Second comes the priority to have troops lay down along any ridge or bank or such elevation feature inside the action square and this is why foxholes don't work well in steep terrain but rather need relatively flat action squares. Third comes the trees: foxholes and trees don't mix very well since the troops seem to prefer to seek shelter behind trees rather than in the foxholes, although this usually affects only some of the pixel troops in a group. And then there are the shell holes which might actually be on the top of this priority but i am not sure. So all in all any terrain feature that has a overriding effect on default troop placement inside any given action square usually has a higher priority than the foxholes which them selves seem to also have some overriding effect vs the default placement of troops.  The problem is that the foxholes should have higher priority than the other overrides.  So to sum up troops place them selves by: 1) walls/buildings 2) ridges/banks 3)trees (and maybe flavour objects) 4) shell holes (might actually be the no.1) 5) any fortifications 6) default placement. 
Using the facing command, as has been pointed out, does help in some cases but often the troops might still crawl back to the wrong positions on their own. 

As before, even if terrain or fortifications do provide a "cover save", I think it's wise to assume that they don't. I've certainly had the impression that individual soldiers die faster out of foxholes than in, but can't confirm that.

CM's ballistics modelling is great. Individual bullets are modelled, and where they impact matters, on a 1:1 basis. This would imply that the physical representation of the soldier matters here.
Artillery seems to be an explosion, followed by a random number of  invisible projectiles that are drawn from that point. This means that whilst fortifications might offer little - or even zero - cover against the initial blast, they may offer cover against the fragmentation.
In foxholes and to a lesser extent in trenches, quite a large amount of a solider is exposed. This does mean that a 1:1 representation may result in higher casualties than expected. The AI targeting centre of mass will shift this somewhat in the other direction, but perhaps not enough to counter it.
Some more generic fortification thoughts:
Tests I've run seem inconclusive as to how effective they are as direct-fire fighting positions, but they certainly provide good cover against indirect fire, when hiding. A 1:1 representation means that cowering or hiding soldiers will be hidden fairly well from direct fire, and nearby plunging fire. This representation also implies that foxholes are probably more protective at a higher relative altitude - since less of the man will be exposed. If you have to put them in an exposed position, stick them on a hill. Foxholes and trenches provide cover where there previously was none. Some cover is definitely an upgrade over an empty field, and allows you to shape the terrain, rather than letting the terrain shape you. Foxholes or trenches on a reverse slope defence make for a powerful position- they'll provide protection from indirect fire  (pre-battle or TRPs, probably), so they have a purpose there even if they provided no cover from direct fire at all. The lack of this kind of reverse slope position was one of the major flaws of the Argentinian army in the Falklands - the forward-slope trenches could be effectively reduced by HE and Milan ATGM fire with relative ease.  Split up your squads. 4-5 man teams find it a lot easier to make use of fortifications in general. So. I think fortifications are primarily anti-artillery measures, or used in situations where they're better than nothing. Ideally on reverse slopes, or to shape the terrain to your advantage.
One thing I'm still unsure about is actually how best to use Wire, by doctrine or otherwise. 

haven´t read through the thread or watched the vid but the first frame of it already shows that a road ditch is very near. There´s always troubles when there´s either road ditches, shellholes or small steep slopes in the same (or adjacent) AS as the FH´s. Seems the game engine considers these as at least same good cover than the FH themselves, so chooses for the best one appropiate for the single pixeltroopers individual situation (suspected or spotted enemy LOS/LOF).
Edit: just few more things coming in mind. Think the game engine also considers final positions for the team/squad leader in order for him to gain visual or voice contact to higher HQ if applicaple. Also found the squads lMG team combo having issues to deploy properly and in timely matter. This bugged me like hell in pre patch game version, as lots of precious time gets lost, before a team/squad gets itself sorted out for final deploy. With regard to FH´s it´s oftenly the leader and the assistant lMG gunner to leave a FH for the above mentioned reasons. Worst is some german squads that have 2 inherent lMG teams, like the Panzergrenadiers in my PBEM with HeirloomTomato. But that was not in FH´s, it was inability to deploy properly in houses. Really got to check the new patch now for a fix or improvement hopefully.

I don't want to belittle casualties by any means, but I don't think it's reasonable to just throw around numbers like that as an argument. How many troops took part in the 1 month of fighting in the battle of the Buldge? According to wikipedia ( - best source ever), it was about 705.000 men for the US forces and 23.000 of the casualties were "missing". In any case, even if you admit the fact that the casualties would be spread very unevenly amongst the troops, you can get much worse casualty figures than this in a 1-2 hours' Combat Mission game. 
For example, when two battalions of the Durham Light Infantry brigade launched a night attack across the Simeto river in Sicily against the defensive positions of the german paratroopers, the scene was also described as "hell's kitchen" and casualties were "severe": The 6th battalion suffered 120 casualties, the 9th battalion 100 casualties. (ca. 25-35% of the front strength, I suppose?). In Combat Mission H2H quickbattles, you often get a casualty rate of 60% and more within an HOUR (you just need to check out some youtube AARs).
I stand by my opinion that the engagements in Combat Mission are often too bloody, for the reasons stated above ("hard/ambush" is the most common type of contact) and because of the "player recklessness" (which is just natural for a game) that Jack Ripper has mentioned.

I've described in my post above that I don't think it's just due to the "recklessnes" of players. In my opinion, it's also a matter of map-design (primarily quick battle maps) in combination with a lack of properly functionable defensive structures. So, in my opinion, it's not just because players don't break off the engagement, but also because the rate of casualties inflicted during the actual engagements is relatively high. 

Hi guys!
Good suggestions! I also like command friction and the idea that units should not react to occurances they don't know of. I have no idea myself how it could be translated into a system that could be managed by the game/pc though.  
In my opinion many realism-aspects affect the context/setting of the typical CM match: 
Night battles. People often underestimate how many attacks were carried out at night. (This is also linked to point 2: If the enemy is in defensive positions, you don't want to engage in daylight, as it will be easier to detect your appraoch and stop it - by means of MGs and artillery, usually) Play fewer meeting engagements as this would be an extremely rare situation. Play scenarios with clear attacker and defender roles. The big problem here is that CM lacks the fortifications that would be neccessary to portray such an engagement (lack of proper trenches & dug-outs). The trenches we have right now don't provide sufficient cover against artillery and small arms fire. In most cases, a defensive line can simply be "bombed out" instead of requiring a capture by fire & manoeuvre.  Many maps are a bit crowded. Of course it depends on the historical region (--> bocage in Normandy is an exception, obviously), but there is a lack of more open maps that allow heavier weapons and some defensive assets to play out their advantages.* For example, this is also true for the bunkers which are available in the game right now: In order for their cover to be of any help, the distance to the target need to be really big; otherwise, too many bullets will hit the bunker's opening. Also, MGs were preferably used at ranges at which ordinary rifles could not return fire and at which it was not so easy to pinpoint the location of the MG nest. Because of the short lines of sight (= high lethality of weapons) that are so common in CM, I feel that spotting is more important than it should be, requires you to fiddle around and micromanage a lot, and it also adds randomness to the game. Another factor related to map design is that in many terrain features (e.g. woods) have a rather small footprint or are cut off by the edge of the map which turns them into an "obvious" position that can be neutralized quite easily. It's very easy to saturate a clump of trees with arty or HE. By contrast, a 1x1 km wood is a bit more difficult to neutralize. Generally speaking, from a realism point of view, Combat Mission matches strike me as way too bloody. The amount of casualties is insane. The reasons for this are probably a mixture of the things listed above: the typical CM engagement is an engagement in daylight at very short range with no proper means of defence for the defender (trenches, dug-outs). The one  (BIG) advantage that the defender has is the short range of the engagements which enables him to trigger devastating ambushes (spotting the enemy first, destroying him before he can react).
This is what I've described as a lack of "soft contact" in Combat Mission games. You usually only get "hard contact", with one asset getting out immediately after being spotted, with few if any chances to react. In some cases, it can be extremely frustrating to play under these circumstances. 
I'd also prefer if an infantry unit that gets fired upon while moving with the standard movement-order would go prone immediately, not run to the next waypoint.  
------------------
* Once I've finished my Catania scenario for Command Ops II (https://forums.lnlpublishing.com/threads/the-battle-for-catania-primosole-bridge-sicily-july-1943.5326/)], I will continue my work on a large map (http://community.battlefront.com/topic/133505-thats-one-vast-valley-hard-edged-realistically-scaled-map/).

If you give them a facing order, they'll usually reorient along the terrain piece, which includes fortifications. That's how to fix it, usually.
They're clearly taking up spots within the action spot, and finding a best fit for your move, rightly or wrongly.
It's hard to tell whether there are actual "cover bonuses" or "saving throws" in Combat mission.
In general, I think it's wise to assume that there isn't, and any bonus from cover is the physical interaction between projectile path and terrain piece, which includes tree trunks, rocks and foxholes. Since there's a lot of variability in flight path, you'd be able to produce a reasonable average with enough tests ("This terrain protects 80% of the time"), but that's all it would be.

That means that foxholes are going to offer good defence against direct fire unless you're some distance away (since geometry is a thing - same reason why halftrack gunners survival increases with distance, and it's not just the range of incoming fire), or you're hiding/cowering inside them them.
 


 

I've described in my post above that I don't think it's just due to the "recklessnes" of players. In my opinion, it's also a matter of map-design (primarily quick battle maps) in combination with a lack of properly functionable defensive structures. So, in my opinion, it's not just because players don't break off the engagement, but also because the rate of casualties inflicted during the actual engagements is relatively high. 

Hi guys!
Good suggestions! I also like command friction and the idea that units should not react to occurances they don't know of. I have no idea myself how it could be translated into a system that could be managed by the game/pc though.  
In my opinion many realism-aspects affect the context/setting of the typical CM match: 
Night battles. People often underestimate how many attacks were carried out at night. (This is also linked to point 2: If the enemy is in defensive positions, you don't want to engage in daylight, as it will be easier to detect your appraoch and stop it - by means of MGs and artillery, usually) Play fewer meeting engagements as this would be an extremely rare situation. Play scenarios with clear attacker and defender roles. The big problem here is that CM lacks the fortifications that would be neccessary to portray such an engagement (lack of proper trenches & dug-outs). The trenches we have right now don't provide sufficient cover against artillery and small arms fire. In most cases, a defensive line can simply be "bombed out" instead of requiring a capture by fire & manoeuvre.  Many maps are a bit crowded. Of course it depends on the historical region (--> bocage in Normandy is an exception, obviously), but there is a lack of more open maps that allow heavier weapons and some defensive assets to play out their advantages.* For example, this is also true for the bunkers which are available in the game right now: In order for their cover to be of any help, the distance to the target need to be really big; otherwise, too many bullets will hit the bunker's opening. Also, MGs were preferably used at ranges at which ordinary rifles could not return fire and at which it was not so easy to pinpoint the location of the MG nest. Because of the short lines of sight (= high lethality of weapons) that are so common in CM, I feel that spotting is more important than it should be, requires you to fiddle around and micromanage a lot, and it also adds randomness to the game. Another factor related to map design is that in many terrain features (e.g. woods) have a rather small footprint or are cut off by the edge of the map which turns them into an "obvious" position that can be neutralized quite easily. It's very easy to saturate a clump of trees with arty or HE. By contrast, a 1x1 km wood is a bit more difficult to neutralize. Generally speaking, from a realism point of view, Combat Mission matches strike me as way too bloody. The amount of casualties is insane. The reasons for this are probably a mixture of the things listed above: the typical CM engagement is an engagement in daylight at very short range with no proper means of defence for the defender (trenches, dug-outs). The one  (BIG) advantage that the defender has is the short range of the engagements which enables him to trigger devastating ambushes (spotting the enemy first, destroying him before he can react).
This is what I've described as a lack of "soft contact" in Combat Mission games. You usually only get "hard contact", with one asset getting out immediately after being spotted, with few if any chances to react. In some cases, it can be extremely frustrating to play under these circumstances. 
I'd also prefer if an infantry unit that gets fired upon while moving with the standard movement-order would go prone immediately, not run to the next waypoint.  
------------------
* Once I've finished my Catania scenario for Command Ops II (https://forums.lnlpublishing.com/threads/the-battle-for-catania-primosole-bridge-sicily-july-1943.5326/)], I will continue my work on a large map (http://community.battlefront.com/topic/133505-thats-one-vast-valley-hard-edged-realistically-scaled-map/).

Hi guys!
Good suggestions! I also like command friction and the idea that units should not react to occurances they don't know of. I have no idea myself how it could be translated into a system that could be managed by the game/pc though.  
In my opinion many realism-aspects affect the context/setting of the typical CM match: 
Night battles. People often underestimate how many attacks were carried out at night. (This is also linked to point 2: If the enemy is in defensive positions, you don't want to engage in daylight, as it will be easier to detect your appraoch and stop it - by means of MGs and artillery, usually) Play fewer meeting engagements as this would be an extremely rare situation. Play scenarios with clear attacker and defender roles. The big problem here is that CM lacks the fortifications that would be neccessary to portray such an engagement (lack of proper trenches & dug-outs). The trenches we have right now don't provide sufficient cover against artillery and small arms fire. In most cases, a defensive line can simply be "bombed out" instead of requiring a capture by fire & manoeuvre.  Many maps are a bit crowded. Of course it depends on the historical region (--> bocage in Normandy is an exception, obviously), but there is a lack of more open maps that allow heavier weapons and some defensive assets to play out their advantages.* For example, this is also true for the bunkers which are available in the game right now: In order for their cover to be of any help, the distance to the target need to be really big; otherwise, too many bullets will hit the bunker's opening. Also, MGs were preferably used at ranges at which ordinary rifles could not return fire and at which it was not so easy to pinpoint the location of the MG nest. Because of the short lines of sight (= high lethality of weapons) that are so common in CM, I feel that spotting is more important than it should be, requires you to fiddle around and micromanage a lot, and it also adds randomness to the game. Another factor related to map design is that in many terrain features (e.g. woods) have a rather small footprint or are cut off by the edge of the map which turns them into an "obvious" position that can be neutralized quite easily. It's very easy to saturate a clump of trees with arty or HE. By contrast, a 1x1 km wood is a bit more difficult to neutralize. Generally speaking, from a realism point of view, Combat Mission matches strike me as way too bloody. The amount of casualties is insane. The reasons for this are probably a mixture of the things listed above: the typical CM engagement is an engagement in daylight at very short range with no proper means of defence for the defender (trenches, dug-outs). The one  (BIG) advantage that the defender has is the short range of the engagements which enables him to trigger devastating ambushes (spotting the enemy first, destroying him before he can react).
This is what I've described as a lack of "soft contact" in Combat Mission games. You usually only get "hard contact", with one asset getting out immediately after being spotted, with few if any chances to react. In some cases, it can be extremely frustrating to play under these circumstances. 
I'd also prefer if an infantry unit that gets fired upon while moving with the standard movement-order would go prone immediately, not run to the next waypoint.  
------------------
* Once I've finished my Catania scenario for Command Ops II (https://forums.lnlpublishing.com/threads/the-battle-for-catania-primosole-bridge-sicily-july-1943.5326/)], I will continue my work on a large map (http://community.battlefront.com/topic/133505-thats-one-vast-valley-hard-edged-realistically-scaled-map/).

Addition/correction to previous post:: Test 1 / tank / tall grass:  460 / 650 / full lane (second value was missing)
Conclusions from the tests described in the post above
1. Weather and daylight conditions don't seem to have an effect on LOS per se. For dawn (05:00) and hazy conditions, the LOS is the same as for clear conditions. At night (00:00), there is a hard cap on visibility (400m in this case - I've read that CM titles do consider the moon phases, so the exact value may vary with the date respectively). As there clearly is an effect on LOS in adverse light an weather conditions, but LOS as given by the target command stays the same, it seems as if units might rather receive a "hiding bonus"?
2. I think that my theory from last year (based on observations in CM: Black Sea) is not too far off the mark: There must be at least two values for each terrain: 1) density/LOS blocking value, and 2) height (either as in an acutal hitbox of some sort, or a z-value for the whole action spot). Different densities must be the reason why the range of full lines of sight vary with terrain. For example,  forest terrains (105m full LOS) are not as dense as crop terrains (60m full LOS).
The tricky part is to explain the "reverse slope" line of sight zones and the difference of results between tanks and infantry. It's much easier to explain my theory in pictures so here we go: 

Eyes below terrain height (e.g. prone infantry --> creates limited "reverse slope" LOS)


Eyes above terrain height (e.g. tank --> creates unlimited "reverse slope" LOS)


Explaining hull-down results with different terrain heights


 
--------------------------------------------------------------------------------------------------------------
If things work indeed as described in the diagrams and when you consider the test results from above, we end up with these terrain characteristics:
clear: no effect on LOS crops: large height, large density forest: medium height, low density tall grass: small height (but still higher than prone infantry), medium density  

Wow so much feedback! Thank you, guys!
Scenario vs. Quickbattle:
It's true that in order to implement the retreat-idea, I need to create a scenario rather than a quickbattle (@Bud Backer, @sburke). I think it's perfectly possible to create two versions of the map. One for quickbattles, one for H2H scenario-play (no AI!). I also wished that more (slightly modified, if neccessary) scenario-maps were available for quickbattles (@JulianJ), especially since most scenario-maps are really well done and often scaled realistically. It's just a pity that scenarios have a rather limited H2H appeal if both the forces and their deployment are pre-determined. With free deployment, it's more interesting. I think the appeal of scenarios could be greatly enhanced by adding more randomness: e.g. giving players the option to choose between sets of pre-selected troops, adding a random factor to the arrival time of reinforcements, etc. 
A scenario also gives me the option to select troops for the players. The selection of forces by the players is a good thing, but sometimes I think it really favors some set-ups while discriminating others. 
I'm also intrigued by the idea to add reinforcements for the defender. I can even imagine to make him start the battle with only infantry (against some armored support for the attacker). He would then get reinforcements (Panzergrenadiere in halftracks!  ) to relieve the infantry or launch a counter attack. But maybe a proper counter-attack is better represented as a separate mission on the map (as part of a tiny campaign).
Retreat-idea:
Indeed I think it will be tricky to set up the objectives in a way to make the defender ponder whether it is better to retreat or not. As you've mentioned, domfluff, It's easy to give the defender incentives to run away. You just need to give him exit objectives. Giving him some incentive to stay, however, is more complicated. In any case, I'd like to make a retreat an option once the defender knows he's going to lose the terrain objective. Instead of making a final suicide stand, I'd prefer if he could withdraw his forces to limit the extent of the defeat or perhaps even get away with a draw. So therefore, I think that the defender should not be awarded too many points for destroying the attacker's units - this would just reward the suicide-stand rather than the retreat. I assume it's more reasonable to create a balance between "preserve own troops/exit the map" and "terrain" objectives. In order to make the decision more interesting, there need to be several smaller terrain objectives, not just a single large one.
Another important aspect here is that I think players should be allowed to know how the outcome is calculated (by adding the info to the briefing...). They need to know that at some point, a withdrawal can be an interesting option for the defender. 
Scouting:
For me, scouting is just not within the scope of CM. In a H2H battle, it's rather boring to exchange 50 turns of doing "nothing". Rather, I would like to add the information that has been gained by recon before the battle to the briefing, or perhaps even on the map (by using "landmarks"). But this again cannot be done if you allow free deployment of forces. But then you can still set the "intel filter" in the scenario editor to give some information to the players.
Villages, sizes of fields:
It's certainly true that in the 1940s, field sizes were smaller in general. However, there is still a lot of room between the standard QB-map field sizes and a properly scaled field. And you can get a pretty good picture by looking at the paths that are displayed running along larger fields on contemporary maps and also by taking a look at photos of aerial recon.
As for the villages - maybe I'm confused by the maps. On many contemporary maps, houses seem to be spaced out quite a bit. But probably their footprints are displayed in an artificially distorted (also too big) way in order to make the layout of the village clearer. 
---------------------
Generally speaking, I need to point out again that I don't expect battles on a more realistically scaled map to be more interesting/tactical by themselves. I think they will just play out a bit differently and also a bit more "relaxed", with what I'd like to call a "soft" contact. Units will become aware of each other at longer distances which means that they're not knocked out instantly and can observe the enemy a bit more. And, for that reason, MGs can for example lay down fire when the enemy's rifles are still way out of their range. You will quickly learn how much of a "close range" weapon ordinary infantry really is. Casualties tend to trickle in more slowly and more "accidently". I also hope that the suppression-system will shine in a bit more nuanced way. If my units get suppressed, they're usually dead very soon anyway and their suppression bar is maxed out. I rarely see medium levels of suppression for sustained periods of time. WIth a larger distance between the contrahents, I hope to see more nuanced levels of suppression at work. In the same vein, I think that armor values will become more important (at point blank ranges, anything goes).
On the current QB maps, by contrast, contact is very "hard", spotting leads to immediate catastrophical results. This fosters a kind of un-relaxed (many people would say: more exciting! ) gameplay and inculcates the typical "paranoia" in CM-players. Every freaking ridge, every corner of a house is just a death trap that potentially leads not to 1 casualty, but to the wipe-out of the whole team. It's just so unforgiving. This is certainly realistic for the final stages of an modern fire-arms assault (if the defender does not withdraw!), but it misses out all the stages before the assault, where it's not as much down to instincts and reaction time, but to slightly more deliberate decisions.

Here is a list of tiny details I've been unaware of and that are not described in the manual. I came upon them more or less by accident. Even though they're of minor importance, I think it's a good idea to note them down, also to show how detailed the game really is! Feel free to add your own tiny discoveries!
When buying  the mechanized/motorized version of a force, the organic (i.e. bought as part of a formation) trucks sometimes carry special extra ammunition. If part of an AT-gun unit, they carry extra AT-gun ammo (e.g. for the Soviet 45mm AT-gun: 4x HE, 16x AP, 2x cannister). If bought for other unit types, soviet trucks labeled as "(weapons)" come with additional extra ammo for the light mortar and the AT-rifles (24x 50mm mortar rounds, 50xAT rifle rounds).  Even though you can't tell from their ingame appearance/UI, soviet AT-hunter teams (3men) do have AT capabilities. They are listed as carrying "6 grenades". What the game does not tell you (also no special symbols in the equipment UI-slots) is that 4 of these 6 grenades are anti-tank-grenades. The team leader carries two ordinary HE grenades on him while each of the remaining two soldiers carries 2 AT grenades. You need to be really close to use them (within 10m?). So they're like breach teams with detonation charges, just cheaper. On the Soviet SU 76M (self propelled AT gun), you get some standard extra ammo that is automatically shared with infantry units close by (710x 7.62mm rounds). There is no need for the "acquire" command. Maybe it's meant for the weapons of the crew - the manual mentions that they can somehow fire their weapons in self-defence? The enclosed SUs also have extra ammo - I need to test if they also share it.

I want to share what might be an interesting idea for fellow scenario-designers. I've been thinking about woods and thickets for my scenario lately. Looking at the maps featured in CM, planting trees on top of forest-ground seems to be the most common method. But then I asked the internet. And the internet gave me the idea that there might be a better way to give woods a bit more love, both in terms of gameplay-mechanics and aesthetics.  
The problem I found with most woods on CM-maps is that they lack a proper woodland-edge (http://www.spektrum.de/lexikon/geographie/waldmantel/8789) - a rim of ca. 10-20 meters (1-2 squares) of very thick bushes and small trees. Basically, you want to create a rising, unbroken forest canopy, with bushes on the outside, followed by trees growing in size as you go deeper, first leaf trees, then conifers. A proper woodland-edge should block any LOS into the wood and provide excellent concealment. Once you're "inside" the wood, you'd get larger trees, i.e. no more treetops blocking your LOS. 
Right now, mapmakers seem to rely on an increased density of trunks in order to block LOS into woods. This looks and feels wrong and severely restricts the ability to fire "out of the wood". Also, as long as concealment in woods comes from tree-trunks, moving to the edge (from inside) is very dangerous as the number of trunks between you and the enemy decreases. Breaking up forests into an edge-zone and an "interior"-zone is the way to go! With those dense hedges at the edge, you can quick-move units to the edge and only let them crawl the very last square safely. With treetrunks only, quick-moving towards a forest-edge was a game of roulette.
After quite a lot of testing and fiddling around, I'm quite happy with my result (see screenshots). The most important finding was that - in CM:BS - you must not use "bushes" (foliage terrain) but bocage (fence-terrain) to represent thickets in the woodland-edge (or thickets in general!). What I've done is to simply place hedges in totally random patterns to create thickets. In my playtesting, the results were superb. Not only does the "low bocage" that I used provide excellent concealment (and still let's you see out), but also, a 2-3 square-wood-land-edge gives the enemy a much harder time when it comes to selecting suspected targets for area fire.
Also, I placed smaller random patches of bocage/hedges "inside" the wood. This was a real relevation. The combination of readily available lines of sight (because there are only high trees "inside" the wood, so LOS is only obstructed by spaced-out tree trunks) and drastically increased concealment potential (hedges everywhere) led to very satisfying engagements in which firing almost never gave away the position of a unit to the enemy. Of course, in such a setting, you simply need to area-fire, and the AI cannot make use of it in a way an actual player could. But I'm really looking forward to testing my "wood" in a H2H game!  
 
Some screenshots of the map for the scenario:

Here is a list of tiny details I've been unaware of and that are not described in the manual. I came upon them more or less by accident. Even though they're of minor importance, I think it's a good idea to note them down, also to show how detailed the game really is! Feel free to add your own tiny discoveries!
When buying  the mechanized/motorized version of a force, the organic (i.e. bought as part of a formation) trucks sometimes carry special extra ammunition. If part of an AT-gun unit, they carry extra AT-gun ammo (e.g. for the Soviet 45mm AT-gun: 4x HE, 16x AP, 2x cannister). If bought for other unit types, soviet trucks labeled as "(weapons)" come with additional extra ammo for the light mortar and the AT-rifles (24x 50mm mortar rounds, 50xAT rifle rounds).  Even though you can't tell from their ingame appearance/UI, soviet AT-hunter teams (3men) do have AT capabilities. They are listed as carrying "6 grenades". What the game does not tell you (also no special symbols in the equipment UI-slots) is that 4 of these 6 grenades are anti-tank-grenades. The team leader carries two ordinary HE grenades on him while each of the remaining two soldiers carries 2 AT grenades. You need to be really close to use them (within 10m?). So they're like breach teams with detonation charges, just cheaper. On the Soviet SU 76M (self propelled AT gun), you get some standard extra ammo that is automatically shared with infantry units close by (710x 7.62mm rounds). There is no need for the "acquire" command. Maybe it's meant for the weapons of the crew - the manual mentions that they can somehow fire their weapons in self-defence? The enclosed SUs also have extra ammo - I need to test if they also share it.

Good example, but the enemy hMG was badly positioned and isolated. In a proper defence, the approach to the hMG would be covered by other positions - e.g. behind the house to the left or in a house further to the rear/left (cannot be reached by your cover team), in the little wood to the right. It's the tunnel vision that can give you a bloody nose in CM if you try to advance by fire and manoeuvre. 
Also, I wonder whether the fire team actually fired? Trading fire with a hMG with some rifles usually doesn't end very well for the rifles. ^^ (Unless the very short range here helped the rifles. The MG is deployed within 110m of a wood that provides excellent cover for attackers to approach)
Personally, I find it too risky to just suppress the one position I'm aware of and manoeuvre against it, keeping my fingers crossed that there are no other enemy positions. It's much safer to suppress all potential positions within the line-of-sight-sector and only then start to move.  Ammo traded for fewer casualties! Of course you need much more firepower to pull this off. 
Sometimes, I get the impression that the whole "fire & manoeuvre" for infantry was just propaganda to give the infantryman the idea they can win. Yah! You can do it! If the enemy is isolated and offers his flanks, and if there is no enemy armor around, and if artillery doesn't hit you! No really, you can do it! I guess CM has spoilt me *thousand yard stare*