Artillery Effectiveness against Infantry too high ? tests included

[Duckman]

Interesting experiment. The difference may be partly due to the lack of foxholes in CM, like others have said (probably the biggest missing piece in the game in my opinion). The WWI-style trenches should be a lot more vulnerable to artillery.

As for improved field fortifications, simple overhead cover is probably not a big issue since airbursts were rare in WWII (only the US had proximity fuzes, from late 1944 and in limited numbers). However given time troops could of course construct log bunkers and the like which are much better protected.

[YankeeDog]

CM is intended to be a low-level, tactical combat simulator, modeling situations where the enemy is close and contact is imminent, and/or has already happened. When you attempt to use CM to model things beyond this scope, such as operational-level artillery concentrations, etc., the game model shows weakness because you’re using it to depict something it wasn’t really designed to represent.

CM is not designed to model the state during the vast majority of the time on the front lines of a conflict like the ETO in WWII, when neither side was attacking and lines were more or less stable. Even heavily engaged units often went days or even weeks between “hot” combats. Front line infantry especially spent the vast majority of their time trying to avoid contact with the enemy, hiding out in whatever holes, trenches, or bunkers they could construct or find, being as quiet as possible, and generally trying to keep their exact whereabouts concealed.

Ultimately, the comparison you’re trying to set up is largely out of the scale of CM. If CM were a were a hedge trimmer and PC were a lawn mower, you mowed the lawn with both, and discovered that the hedge trimmers don’t mow the lawn as well as the lawn mower. Big surprise there. Try to trim the hedges with the lawn mower and you’ll get a similar result going the other way.

While the game isn’t really designed for it, if you really want to simulate how units would typically be deployed on the operational scale when not prepared to repulse an immediate enemy attack, in addition to spreading units out both laterally and in depth, using bunkers to represent overhead cover, etc., the majority of units should have “Hide” commands, with a few OPs, pickets, and perhaps small patrols up “watching the wildlife”, as you put it, to keep a lookout for enemy activity. Only when significant enemy activity is detected should the majority of units go “heads up”, and “man the parapets,” so to speak.

All this said, there is a tactic that I do often use in CM that is similar to the operational-level deployment conditions described above: When defending, after I get my setup arranged, I often select my entire force by double-clicking on the most senior commander, and then Hide everybody. Then I go back and selectively “Unhide” specific units that I want to stay up and watching for enemy approach. I unhide other units once my OPs have seen the enemy coming and have an idea of the plan of attack. IME, this is a very effective way to reduce casualties from any prep artillery concentrations or other speculative area fire by the attacker. It also prevents my heavy weapons from revealing their positions too early by opening fire on enemy scouts.

I do agree that it would be great if the Computer Player had the ability to execute tactics like this, as it would make the computer a much more challenging opponent. I think this would be a difficult piece of AI programming, but I'd love to see it.

But I do not see any great need to change how the Hide/taking cover system works right now for a Human player. The commands are already in the game for me to tell my men to shelter from incoming fire, or stay “up on the parapet”, as the tactical situation demands.

While I don't see any urgent need for change, I can certainly imagine tweaks that would further improve things. For example, Steve has hinted at the eventual addition of some sort of “Ambush” command. This type of command could be way to tell units to “Auto Un-Hide” when the enemy gets within a certain distance, which would be a very useful thing.

But I certainly do not want my units to do the reverse, and “auto-hide” as soon as artillery starts coming in, as a default. If enemy tanks are in sight, I generally want my AT Gun crews to stay “eyes up” and shooting at the tanks, artillery be damned. Same goes for MGs if enemy infantry is approaching. If there is no need for a unit to be up and ready to fire, I’ll hide them myself. And if I lose bunch of men to artillery because I ordered them to stay up ready for contact when they should have been down and sheltering, that's my fault, not the game's.

[akd]

This does not help the AI player, which just exacerbates the ease of playing against the AI. Although the original comparison is irrelevant, there is a valid takeaway: the AI does not react very realistically to incoming artillery when in field fortifications. They should be taking cover (i.e. cowering) in their fortifications much earlier, rather than after taking 20-30% casualties. However, that is complex to separate from the more general suppression system (which leads to breaking or staying under cover (cowering) even after the artillery lifts.

Interesting experiment. The difference may be partly due to the lack of foxholes in CM, like others have said (probably the biggest missing piece in the game in my opinion). The WWI-style trenches should be a lot more vulnerable to artillery.

??? There are foxholes in CM.

[pord]

You are the commander of all your troops. You are supposed to tell them what to do. Entrenched troops implies they should be hiding as well (i.e. utilizing the entrenchments). If you as the commander do not tell them to hide (use the entrenchments) then the troops sustain high casualties. Makes sense to me.

As an analogy, say you are fighting a tank battle. You put all your tanks on a 20m fire arc. The enemy does not. All your tanks get blown to bits. You could argue the tacAI should override your orders. But the tacAI cannot be as smart as the human player. As a smart human player, you give the orders and your units try to obey them.

I suggest you rerun your test, and as soon as you realize your troops are going to be under artillery fire, you order them to hide. Compare your results then, since this is how the game is designed to work IMO.

[womble]

This does not help the AI player, which just exacerbates the ease of playing against the AI.

Fair point, that. I've been playing solely PBEM recently, and had forgotten about vsAI games.

I've run a test like Siffo did: 4 slightly overlapping 400m diameter missions. I took 7 KIA and 3 WIA. Didn't count the "lightly wounded". I'm obviously doing something different.

Here's the scenario I used in DB.

https://dl.dropboxusercontent.com/u/79012974/Artillery%20vs%20trenches.btt

[siffo998]

Fair point, that. I've been playing solely PBEM recently, and had forgotten about vsAI games.

I've run a test like Siffo did: 4 slightly overlapping 400m diameter missions. I took 7 KIA and 3 WIA. Didn't count the "lightly wounded". I'm obviously doing something different.

Here's the scenario I used in DB.

https://dl.dropboxusercontent.com/u/79012974/Artillery%20vs%20trenches.btt

cannot open your file. which version of CMBN do you use ?

[womble]

cannot open your file. which version of CMBN do you use ?

It's FI v1.01. I'm pretty sure I'd take less casualties in Normandy because I could use bocage to shelter one side of my trenches.

[Vinnart]

The arty seems fine to me. When CMBN first came out it was WAY too overpowering. One round was taking out whole squads all the time, but since the fix I have no complaints. When the arty starts coming down your two best options are to either run FAST away from it, or HIDE if you want to try to stay in position. Doing neither will always get you more casualties.

[Brindlewolf]

One of the first things i learnt when playing CMBN-stay mobile-any resting on your laurels and you were doomed from above.The first spotting rounds i see now,and my guys are getting the hell out of dodge.Either defending or attacking i keep my guys on the move.

[DavidFields]

Having weighed in on this from the beginning, I think that things are better than they were on release. In my last game, I called in off-board artillery on an enemy 75 meters away caused a spotting round to end up hitting my own troops, with a devastating effect. I was pleased with the improvement in the inaccuracy of the artillery--if not the result.

I am even caving on the linear fire plans: though the reality was probably rectangular boxes, the linear fire plans simulate that without having to increase the shells from off-board units--it is a game mechanics decision.

My opinion has always been that it is hard for CM2 to simulate micro-environmental issues--rocks to hide behind, depressions in the land, that a soldier would desperately use. Given that CM2 tracks every piece of flying metal, the lack of improvised "body armor" is an issue.

I already put everyone on a covered fire arc, then cancel those I want not to have it. It is a useful suggestion to have everyone on hide.

One could have as a default CM2, that when troops moved "Quick"--which is what I usually use when moving troops in a battle space, they would automatically end there turn in "hide", and you would have to counteract that if you wished. On balance, I might even prefer that.

Or, to torture BFC, they could have a "some hide" command, so that everyone is not sticking their head out.

On the one hand, my inclination is to defend BFC for difficult game design, and marketing, decisions. I might argue that the simulation is already too tough for most new, casual, gamers. On the other hand, any idea that they are absolutely right is.....well, this has been a great franchise because it has evolved. (? darkest hour--the CMSF release, I think. We are now mostly dealing with judgment decisions) Effectively toning down the artillery, without relying on the gamer to glean stuff from the Boards, might...and I don't know for sure...be wise.

[womble]

My opinion has always been that it is hard for CM2 to simulate micro-environmental issues--rocks to hide behind, depressions in the land, that a soldier would desperately use.

That's why it's not simulated. It is, however, accounted for by a "terrain save" abstracted mechanism.

[Duckman]

...a "terrain save" abstracted mechanism.

I love that term, it reminds me of the old D&D saving throw.

[siffo998]

That's why it's not simulated. It is, however, accounted for by a "terrain save" abstracted mechanism.

well first i`ve just had the time to rerun your scenario and i`ve got 8 KIA and 8 WIA which is in my opinion still a lot of casualties for a entrenched force thats randomly shot at by four 10,5cm pieces. In % these are exactly 10% casualties (out of 160 Men total).

second there are some differences between our tests:

1) you are using CMFI 1.01. Iam using CMBN 2.01. Dont know if theres something changed regarding artillery but there are definitely differences regarding environment and ground conditions.

2) my force consists of 164 men (a small difference )

3) and thats the most important in my opinion: you are centering most of your men in the middle of the map directly in a village with the trenches beside the buildings. so most of the company in the middle does not get any casualties because the buildings around the trenches protect them from shrapnells... etc...

[Duckman]

Is there any way, using the in-game camera, to see exactly how the casualties happen?

I'm thinking if we can settle the issue of whether they get hit because they don't keep their heads down (hide) or if it's something else.

[womble]

well first i`ve just had the time to rerun your scenario and i`ve got 8 KIA and 8 WIA which is in my opinion still a lot of casualties for a entrenched force thats randomly shot at by four 10,5cm pieces. In % these are exactly 10% casualties (out of 160 Men total).

Again, it's only hastily entrenched. Replace the trenches with wooden bunkers to get some overhead and sideways cover, representing 'proper' entrenchment and casualties will go down further. And 150 rounds is a lot of HE to chuck.

second there are some differences between our tests:

1) you are using CMFI 1.01. Iam using CMBN 2.01. Dont know if theres something changed regarding artillery but there are definitely differences regarding environment and ground conditions.

I don't think anything changed with arty terminal effects. Ground conditions have been determined to have no detectable effect on blast/shrapnel, though it'd be good if they did.

2) my force consists of 164 men (a small difference )

You kept the Jeeps, I'm guessing?

3) and thats the most important in my opinion: you are centering most of your men in the middle of the map directly in a village with the trenches beside the buildings. so most of the company in the middle does not get any casualties because the buildings around the trenches protect them from shrapnells... etc...

Well, duh. It's a test to see whether we think something is realistic. So I've used (what I consider to be) realistic deployment patterns, and to a cursory extent, realistic siting of the trenches. I've made some effort to avoid siting trenches under trees, too. And if it was Bocage country, the out-of-town platoons would be snuggled in to bocage lines.

Oh, and one platoon out of four, plus the CoyHQ and 2 MG teams in no way constitutes "most" of the men. Two rifle platoons are in the rough terrain out of town and the mortar teams and 4th Plt command elements are in the backfield out of

(direct) harm's way, where they ought to be.

If I knew what was coming (or if it was one 105mm battery firing harrassment across the entire position for two hours), the town guys would be in the houses, largely immune to anything but a direct hit; they're in the trenches because that's where is safer if there's a direct concentrated stonk on the town or the threat of DF HE. They'd move up a bit once infantry contact became imminent.

And that's still with the troops not Hiding. While I agree that the AI might need some help in issuing Hide orders before units take enough casualties and suppression to make them cower, the fact that casualties when troops are hiding are very low indeed simply shoots your "artillery is too effective" generalisation in the face.

I just ran the test with the dogfaces told to keep their heads down and got 2 and 3 KIA/WIA. And that's with a T-shaped barrage, not a cloverleaf, so not covering the "proper" area that your PC comparison would have scattered the incoming across, and 176 rounds rather than 140 (though we still have no idea how many rounds are involved in the PC mission).

[Pak40]

While I agree that the AI might need some help in issuing Hide orders before units take enough casualties and suppression to make them cower, the fact that casualties when troops are hiding are very low indeed simply shoots your "artillery is too effective" generalisation in the face.

Ditto. This is the real issue here. Of course, Siffro's tests are further invalidated because he doesn't order his units to Hide after the first minute of of a 10-15 minute bombardment, which any sane player would do. The TacAI actually does a good job of having units go prone for a few seconds when under bombardment, so the real fault here is that the commander(player) doesn't issue a HIDE order after the first minute (in real time he could do it immediately).

I'm actually under the opinion now that the TacAI shouldn't order units to HIDE - it should be left up to the player because there will be cases where the player will take that chance of having his units exposed so that they can spot and return fire. If the TacAI starts issuing HIDE orders then where does it draw the line between who should and shouldn't be forced to hide? Does a unit that is 50m away from an exploding shell get forced into a HIDE order? And for how long? I can foresee many complaints that many units will HIDE too easily and for too long, especially at the end of a barrage when the impending enemy infantry attack will come.

[Battlefront.com]

Well, duh. It's a test to see whether we think something is realistic. So I've used (what I consider to be) realistic deployment patterns, and to a cursory extent, realistic siting of the trenches. I've made some effort to avoid siting trenches under trees, too. And if it was Bocage country, the out-of-town platoons would be snuggled in to bocage lines.

This is the correct methodology. Not only is it useless to compare two games together, it is equally useless to compare some real life information with a completely different situation within the game. This gets me back to my favorite bone to pick... people who use real life test range results and expect the game to replicate them consistently. Or they do things like test for Morale effects and neglect to tell everybody they are comparing the behavior of Veterans (in game) to Conscripts (in some historical battle).

People MUST approach these sorts of comparisons using standard scientific methodology. That means setting up carefully considered tests which compare against a reliable data set that is directly applicable to the game setup. And also to keep in mind that Combat Mission never, ever, ever assumes it is engaging in a "test range" because that would be silly. Therefore, test range data can never, ever be used without caveats and a great deal of allowances.

Failing to follow simple scientific methodology gets an instant "bzzzzzzzt" sound from the judges

Ditto. This is the real issue here. Of course, Siffro's tests are further invalidated because he doesn't order his units to Hide after the first minute of of a 10-15 minute bombardment, which any sane player would do. The TacAI actually does a good job of having units go prone for a few seconds when under bombardment, so the real fault here is that the commander(player) doesn't issue a HIDE order after the first minute (in real time he could do it immediately).

I'm actually under the opinion now that the TacAI shouldn't order units to HIDE - it should be left up to the player because there will be cases where the player will take that chance of having his units exposed so that they can spot and return fire. If the TacAI starts issuing HIDE orders then where does it draw the line between who should and shouldn't be forced to hide? Does a unit that is 50m away from an exploding shell get forced into a HIDE order? And for how long? I can foresee many complaints that many units will HIDE too easily and for too long, especially at the end of a barrage when the impending enemy infantry attack will come.

Bingo. This is EXACTLY why the AI doesn't mess around with this sort of thing. It's so very, very hard to get such behavior even remotely close to what people expect, therefore we go with the best "middle of the road" approach. Works most of the time, but definitely runs into problems with this sort of setup.

I have a few "short cuts" in mind that could help the TacAI, but we haven't tried to implement them yet. In short it is making the TacAI aware of being deliberately targeted by artillery, and it's mission parameters, and making a decision based on that information (with randomness, of course!). This would work equally well for both Human and AI Players and would never be explicitly shown to the Human Player. The concept here is to remove the need for complicated, brittle AI programming and just allow the AI to figure out what even the most basic Human Player would guess is going on. The end results will be pretty consistently sensible and yet it won't take a month to code.

My hope is for v3.0 to have significant TacAI behavior improvements such as this one.

Steve

[beatmasta]

I have not played PC so i can't say much about that but I would say that CM has some flaws. Not that the artillery is too effective. I'm sure that when a unit that is hit like they do in CM they will have very high casualty rates, even higer than in CM. The problem is that the CM arty strikes seem to be too accurate and too concentrated. I think it would be very rare if you set a battery of artillery hit a point target to get as small impact area as they do in CM and as fast as it is spotted.

But what I have read about the PC and what I've read and seen about real life events I'd say that CM is a lot more realistic than PC.

[SlapHappy]

If you really want to complain about artillery modelling in the game - how about artillery effects on buildings (and occupants). I know that CM2 isn't a demolition simulator, but the

effects of heavy artillery and bombing on structures isn't even close to the severity one would see in real life. I watched two 8-inch shells directly impact a 2-story brick building and the model showed no signs of damage whatsoever. Neither of the 2 squads in the building showed any injuries. Each of those shells pack about 38 pounds of high explosive - more than enough to devastate a building even that size. Not to mention the fact those shells would likely have penetrated through the light roof of the structure and exploded internally creating massive pressure loads inside the building (and it's occupants). If getting the exacting penetrating potential of a 75mm tank shell vs. the glacis of a Tiger tank at 400 meters is so important, why isn't this?

[JonS]

I think people grossly overestimate the effects of HE on structures. It isn't hard to find images of buildings which are quite functional - as fighting positions - but which have been subjected to a heck of a lot more than 38pounds of HE.

[SlapHappy]

1 psi Window glass shatters. Light injuries from fragments occur.

3 psi Residential structures collapse. Serious injuries are common, fatalities may occur.

5 psi Most buildings collapse. Injuries are universal, fatalities are widespread.

10 psi Reinforced concrete buildings are severely damaged or demolished. Most people are killed.

20 psi Heavily built concrete buildings are severally damaged or demolished. Fatalities approach 100%.

The wall of a building exposed to that "measly" 38 pounds of TNT from 10 ft. away is over 130 PSI. Still think it's nothing? Physics say otherwise.

Human Injury

Humans have an amazing resistance to overpressure. Lethal amounts approach 40 psi which is staggering. The only permanent injury sustained is the loss of eardrums. However, an indirect effect does occur to human casualties. Due to the massive winds and pressure created by the blast wave, tons of debris is hurled through the air, not to mention people themselves. The flying debris, (and people) cause massive amounts of casualties such as impalements in the head, chest and abdomen, and broken bones [5].

[JonS]

I didn't say it was "nothing", and I didn't say 38lb was "measly". Do you often find that misrepresenting what people say works out well for you?

I wouldn't want to live there, but could I move around in, and fight from, this:

Yes, of course I could.

And here?

Again, yes, even though it took a bit more than a couple of 8-in rounds.

[SlapHappy]

"Do you often find that misrepresenting what people say works out well for you?"

Do you?

I clearly stated the likely effects relative to building damage and the effects on the occupants. Nowhere did I say anything about "walking around" or "fighting out of" such a building after the blast had already occurred. That is "totally irrelevant" as you are so fond of pointing out to other people on this forum.

[SlapHappy]

Look Jon, I don't want to get in a pissing match with you about this. I respect your opinion and knowledge of the topics we cover on this board. The numbers and info I have posted are quite accurate. Why is it so hard to accept that someone else might have a clue what they are talking about?

[MikeyD]

According to weekly accounts from Army group B in Normandy, from June 6 to July 7 German forces facing the invasion suffered on the order of 100,000 casualties. That's twice the U.S. total casualties in the Korean war compressed into a few short weeks. Carnage on that scale sweeps away the good with the bad, slaughter on a scale mere squad tactics can't defend against. If you have anyone at all surviving a barrage on a CM map you have BFC's restraint to thank for it. They've said on several occasions that they purposefully dialed-back artillery lethality to compensate for action square bunching.

Its been awhile since anyone's put out a 'demonstration scenario' of a serious prep saturation bombardment of an area, the kind where you're making shell craters in the shell craters.