High casualty rates in CM games

[arpella72]

Comparing the outcome of real war actions with the ones modeled in the game,I always notice that the casualty rate is usually quiet high in the game compared with the real stuff.I mean experienced players not playing boldly and trying to do it as close as to real combat procedures.A part from the AI behaviour in one side,I think there is a reason for that.In my opinion,infantry in the game don't have the same chances of survival as in real life due to the poor modelling of defensive works or the,sometimes,difficulties in getting apropiate cover in terrain features.You can't find parapets in the game,fortified houses and underground basements in buildings nor underground shelters; it's told the protection in shallow trenches and foxholes is abstracted but I m not quiet sure at which degree.What's more,some weapons seem to be overpowered,like the SMG which are deadly accurate and lethal beyond its teorically effective range

This issues should be improbe(probably in a CM3?) in order to make the game more challenging and realistic.This game,differently from others,is focused on infantry which I just like it so,why not to care more about our dear pixeltruppen?.What do you think about it?

 

Edited October 5, 2020 by arpella72

[SimpleSimon]

Many of the scenarios are designed around puzzle solving and play out almost every time like a siege. There are many symptoms of these conditions, but one of the most prominent is of closely matched headcounts or map populations between defender and attacker and this invariably promotes high lethality within a given slice of map. Force-to-space ratios with a high density of units promote extremely meticulous play and are often excessively relied upon to make up for a perceived passivity of the AI on defense. 

As long as you're making reasonable decisions you have a right to expect reasonable results, and the game should score you fairly for that. Unfortunately, I think the way that we score the player's conduct in most of the scenarios is sort of poor and in a minority of cases, egregiously unfair or abusive. 

 

[Erwin]

Many have noted over the years that CM games seem to suffer a much higher casualty rate than in RL no matter how careful one is with the pixeltruppen.   My understanding is that in normal combat taking over 10%-15%casualties will make a unit "combat ineffective".

There are two counter arguments.  1) CM scenarios depict the unusually desperate situations in which much higher than normal casualties may be suffered.  2)  Historical combat casualty statistics are for the whole larger formation - including rear echelon chaps etc.  So, the poor bastard frontline combat troops may suffer massive casualties but the overall rate for the battalion or division may still be under 10%.

Many scenarios for CMSF when it first was released heavily penalized friendly Allied casualties and one did learn to win while losing very few if any men.  Later scenarios made CMSF more bloody.  And CMBS is very bloody.  Actually, if CMBS reflects real modern warfare vs a peer enemy, it's hard to understand why anyone would want to fight as CMBS would seem to teach it would be a massacre for both sides.

[A Canadian Cat]

There are lots of explanations that contribute: poor play on our and the computer's part, a misunderstanding of how different casualty rates were at he tip of the spear vs the over all numbers, various (mostly incorrect) theories about fortifications and modeling etc.

The biggest difference is that we don't act like real commanders. We press far too much and don't withdraw when we take casualties. In a real action units rarely stay and fight on and on. They pull back to fright another day or bring up extra fire power or wait for support etc.

We are playing a game so we don't do that. If you play a game with someone like @Bil Hardenberger or another professional they will tell you when they would be done in real life and trust me it is way way before most people stop. Including myself.

[Glubokii Boy]

A few reasons i can think of that might contribute to higher casulties...

- time...in many scenarios the time limit might be shorter then what it would have been IRL.

- lack of micro management...especially in larger scenarios we might not split the squads up as much as desirable to avoid micromanagment.

- carelesness...we might not take the time to make sure we have base of fire/suppresion and otherwise security in place to a simular degree as we would have done IRL.

- WEGO...the AI might not react to unfortuness events during the one minute turns as well as a RL human would.

- Abstractions....things like fighting in and around buildings does not really give the playrer quite the same level of control as IRL.

- IRL a squadleader would handle the squad...a platoonleader would handle the platoon etc...in CM you as for example a battalion commander is forced to handle every little detail yourself of your reinforced battalion...every turn. The finer detailes might be missed every now and then ...

Edited October 5, 2020 by RepsolCBR

[nik mond]

More often scenarios focus only on the action: the companies involved in heavy fighting, and don't focus on the sister companies luckily positioned a few bounds away engaging the lesser threat, because those guys would be a boring one sided affair to play. Other gamey things such as units might get gunned down trapped on the edge of map, or forced to rush/hold/deny an objective for points.

Historically sure, some engagements were fought to decimation. It happened, accounts of units being demolished, battalions rendered combat ineffective. Its just that it happens more often in game scenarios because you deserve a good fight.

Its also rare to recreate battles utilizing tactical snail doctrine, although I would like to see more of those because I think people could appreciate that.

[Vergeltungswaffe]

People or the AI pushing an attack when it's getting ground up or a defender not pulling back immediately upon being flanked, rolled up, etc, definitely contributes, but to me the one biggest factor is that when you play, you are playing in an area (likely one of the few) where the two sides came into close contact.

As Erwin touched on, your units are only part of a much larger whole, and though your company may have gotten chewed up today, the rest of your battalion and the two others in close proximity, may not have been involved in much of anything, or fired a few shots and the other side got the wind up.

I would definitely like to see the ratio of KIA to wounded be more realistic than now, but actual combat where both sides commit is going to be bloody.

Edited October 6, 2020 by Vergeltungswaffe
nik mond said it before I got around to hitting submit

[Bulletpoint]

While I agree with the above comments about tactics and players pushing too hard, here are a couple of very specific reasons why we see such high casualties.

• The HUNT command doesn't work as it should - troops don't go prone immediately when fired upon. There is currently no way of really advancing cautiously with infantry, even when you know enemy contact is imminent.
   
• Real life troops can go "hull down" behind ridges, but in CM, since it's a square-based system, it's either you stay where you are or you advance 8 metres into the field of fire. No middle ground. This makes slopes more difficult to deal with than in reality.
   
• Troops can't use corners of buildings for cover. Yes, there was a "peek around corners" feature added recently, but it doesn't work reliably, and even when it does, it just places one guy from the team in the street. He's not peeking, and he doesn't seem to be getting any cover bonus from the corner of the building.
   
• Broken troops can still be ordered to advance, and even though they are less effective than fresh troops, they can still engage. They never reach the point where they refuse to follow orders. In reality, there's a limit to how far you can push people.
   
• Many buildings offer less cover than they should. Especially the barns in Normandy are shown as being made of stone, but they offer cover values as if they were American barns made out of thin wooden planks. I've seen many opponents place MGs etc. in such barns, not realising they are death traps.
   
• When placing troops in shellholes, some guys will often simply refuse to take cover in the crater, and instead they will sit on the edge in full view. In CM1, this was abstracted, but in CM2, placement of individual pixeltroops really matters.
   
• Gunners in most open vehicles are placed way too high compared to where the sight of their weapon is.
   
• When sending a team to run or hunt along a low wall, individual troops will often decide to cross the wall and walk for a while on the other side before crossing back. This exposes them to fire and makes it more difficult to use the cover.
   
• Troops don't go to ground when they see artillery starting to land nearby. They only go prone if the enemy shells hit very close to their own location. In real life, if you saw a mortar shell hit 100m to your right, and another 80m to your left, you'd be smart enough to realise that you're in a barrage, and that you should get down. Even if no shell actually landed on top of you yet.

Edited October 6, 2020 by Bulletpoint

[Freyberg]

15 hours ago, Erwin said:

There are two counter arguments.  1) CM scenarios depict the unusually desperate situations in which much higher than normal casualties may be suffered.  2)  Historical combat casualty statistics are for the whole larger formation - including rear echelon chaps etc.  So, the poor bastard frontline combat troops may suffer massive casualties but the overall rate for the battalion or division may still be under 10%.

Agreed.

I recall reading that casualties in intense combat always have been and remain quite high, but the relative number of troops on the cutting edge at one time has changed over time.

CM casualty rates are high, but not unrealistic, given all the variables.

[Glubokii Boy]

A few more reasons...

- CM makes it somewhat more difficult to properly judge LOS/LOF compared to RL.

- Inability to areafire at certain areas of the map because of some terrain feature...that IRL only would provide minor problems.

- Difficulty to get the supporting weapons to use linear or area...areafire. It's pretty much only point...areafire.

 

[BornGinger]

On 10/5/2020 at 8:29 PM, IanL said:

The biggest difference is that we don't act like real commanders. We press far too much and don't withdraw when we take casualties.

I think the reason to this behaviour is that the time frame is too short. Some, or many, large battles which historically took maybe around a whole day often has the limit of two to three hours in the game. This might push the player to send his pixeltruppen into something similar to suicide attacks.

[Freyberg]

39 minutes ago, BornGinger said:

I think the reason to this behaviour is that the time frame is too short. Some, or many, large battles which historically took maybe around a whole day often has the limit of two to three hours in the game. This might push the player to send his pixeltruppen into something similar to suicide attacks.

That's a really good point. The mention of suicide attacks made me wonder...

If the question is: is CM a realistic WWII simulation with respect to casualties? - we can look at some of the historical battles where the commanders did act like CM players, for example the Japanese on Guadalcanal, in which case, I reckon the casualty rates in CM are probably about right.

 

 

[Bulletpoint]

Forgot one of the most important reasons why we see higher casualties:

• Foxholes and trenches offer very little protection, and they only work when the player manually issues a "hide" command. This command can only be given once every 60 seconds, and if troops are already shaken, they won't accept the command. In real life, if you're in a foxhole, you can duck down in less than a second and you don't need to wait for an order to do so.

Proposed solution: Troops placed on a square with foxholes should get the protection all the time, no matter if they are "hiding" or not. This would also benefit AI troops, because the AI can't dynamimcally hide and unhide troops.

Edited October 19, 2020 by Bulletpoint

[Glubokii Boy]

5 hours ago, Freyberg said:

That's a really good point. The mention of suicide attacks made me wonder...

If the question is: is CM a realistic WWII simulation with respect to casualties? - we can look at some of the historical battles where the commanders did act like CM players, for example the Japanese on Guadalcanal, in which case, I reckon the casualty rates in CM are probably about right.

 

 

I've just finished reading David Glantz, Zhukovs greatest defeat - operation mars.

The russians that had the missfortune to participate in that operation would probably have been more then happy with CM casualyy levels...☠

[Erwin]

8 hours ago, BornGinger said:

I think the reason to this behaviour is that the time frame is too short. Some, or many, large battles which historically took maybe around a whole day often has the limit of two to three hours in the game. This might push the player to send his pixeltruppen into something similar to suicide attacks.

Agreed.  And many scenarios, especially in campaigns (eg: CMSF's Highland Games") are only 35-45 minutes long.  One cannot do anything "carefully" in that sort of timeframe.

[MikeyD]

What were the casualty statistics for the Battle of the Bulge? Roughly 200,000 casualties, 1300 tanks lost and 1800 aircraft lost. All over the course of just 5 weeks, give-or-take. How many CM battles would it take to reach 200,000 casualties? 

[MOS:96B2P]

2 hours ago, Erwin said:

Agreed.  And many scenarios, especially in campaigns (eg: CMSF's Highland Games") are only 35-45 minutes long.  One cannot do anything "carefully" in that sort of timeframe.

This is why I like the idea of allowing the player to play for extra time at the cost of VPs.  The briefing must be clear that the player has a decision.  Hit cease fire at the recommended 2 hour mark or play past 2 hours the cost of 300 VPs (or whatever VP amount is appropriate).  This also keeps a scenario from abruptly ending on a player who is having fun and wants to see his plans finish playing out. 

For replays of the same scenario It may be useful to identify what AI plan the player is fighting against.  For this purpose I sometimes use the following method in testing a scenario (when not on scenario author test mode) .  If the scenario has three AI plans: 

Three small modular buildings with no windows or doors are placed on the players side of the map.  There are at least three action squares (24 meters) between each building.  The buildings are landmarked (the red on map lettering) with PLAN ONE,  PLAN TWO and PLAN THREE.  When the player hits the red button to start the game an OpFor AI unit spawns inside the building with the AI plan that loaded.  A OpFor tentative contact icon will appear in that building.  Or if the player doesn't want to know he doesn't place a unit next to the row of buildings during setup.  So, if the player chooses, he now knows  what AI plan he is up against and if he has already played it before.

Granted, if he wants a different AI plan he now must exit the game and reload.  But for a player who's not comfortable opening the editor and turning plans on and off it might be something.  It's easy enough for the designer to set up.   

Hmm, I think there is another thread about replaying scenarios.  I might repost part of this reply in that thread if I find it.              

Edited October 19, 2020 by MOS:96B2P

[Bulletpoint]

22 minutes ago, MOS:96B2P said:

This is why I like the idea of allowing the player to play for extra time at the cost of VPs.  The briefing must be clear that the player has a decision.  Hit cease fire at the recommended 2 hour mark or play past 2 hours the cost of 300 VPs (or whatever VP amount is appropriate).  This also keeps a scenario from abruptly ending on a player who is having fun and wants to see his plans finish playing out.

I agree. Another way to do it would be to give the player a victory time bonus of X points if he can take an objective within a set number of turns. Then after that turn, each turn will erode the victory time bonus until it's all gone.

Edited October 19, 2020 by Bulletpoint

[SimpleSimon]

It's amazing to me how profound the mythology of "Flanders" and "The Western Front" has been on the history. It's really exemplified in many of the game's scenarios I think. Gotta seize the enemy position by T+2:00 is never really a good objective to saddle the player with. Why that position? Why am I starting where I am? Why only two hours? Usually confronted with the context of a CM scenario I often quickly end up feeling like i'm going to need far more support or better circumstances or both to match the scenario in a single save without causing a huge bloodbath. 

That's why I like the idea of things like an overtime meter just being a modest VP penalty. Usually the highest value of them all should be the integrity of your own force, in 2nd place destruction of the enemy's, and dead last the circumstantial objectives such as the time limit, captures, exits, etc. 

[Glubokii Boy]

14 minutes ago, SimpleSimon said:

Why that position? Why am I starting where I am? Why only two hours? Usually confronted with the context of a CM scenario I often quickly end up feeling like i'm going to need far more support or better circumstances or both to match the scenario in a single save without causing a huge bloodbath. 

One thing that i am considdering to include , in some of my future scenarios atleast, is some degree of 'custom difficulty setting'...

In addition to the default player-force that the scenario has been designed to be played with i'm thinking of adding some additional units (located on a small exitzone) at the rear of the player setup area. If the player would like to include some of those units he simply moves them of the exitzone before hitting go (during setup). This will make the scenario somewhat easier then the default setup.

Simularely...if he would like more of a challange he could move some of the default units onto the exitzone to make them dissapear ones he hits go.

As of now i have not considdered any VP impacts of these tweaks of the original force...it will simply be a voluntary tweak to suit different playstyles/difficultylevels. But i guess that it could be possible to indeed include some point- benefits/penalties if given a try...

 

[danfrodo]

losing 200K soldiers is easy in CM.  just give me 400K and give the enemy artillery and spotters and I'll do my usual impetuous, ill-advised moves.

[SimpleSimon]

It's incredible to me that guys often agree to go into multi or QB battles over a 800x800m map with artillery on one or both sides at all. It's not "shooting fish in a barrel" as much as dropping hand grenades on fish in a bucket. 

Edited October 19, 2020 by SimpleSimon

[John1966]

Funny I came across this thread today as the scenario I played last night had me thinking about the subject.

I was on the defence against German paratroopers and in 90 minutes I was aware that I had killed an awful lot of them.

Finished with a major victory but I'd taken a lot of casualties myself because the assault was relentless. In fact, suicidal. So suicidal, in fact, that I'd killed or wounded 632 Germans. They had little more than 100 men left as "OK".

That's ludicrous. I'd practically wiped out a battalion. Why did they keep coming at me?

Well I reviewed the map and practically all the surviving Germans were "crack". That's why they kept coming. A whole battalion of crack troops? Is that realistic?

I'm guessing that what's going on here is that as the AI isn't great attacking (there were several moments when they were coming at me without any suppressing fire), the scenario designer made them all crack to give them a chance. If they'd been regular (or even veteran) they'd have had none of it and given up the attack.

So the cause of the high casualties in this case is the AI limitations necessitating an unrealistic set up right from the start.

Of course, maybe my boys wouldn't have defended so hard in RL. "These Germans are nuts, let's withdraw", would have been a reasonable response.

But that, of course, is the problem with one-off scenarios. Successfully defending the objectives was the whole point. So not trying to do so makes the whole exercise a bit pointless.

So campaigns should be the way round this. But that only works if there's a sense that not every scenario in the campaign is a do or die affair that has to be replayed to get you through with enough to win the next one. Some of the missions should be routine. Ones you can win fairly easily and the only measure of your success is whether you kept your casualties low enough to make life easier in the next one. Some campaigns achieve this, but some deliberately set themselves up to be difficult and it doesn't work.

One-off scenarios will always be do or die so you'll get bloodbaths.

Mind you, I'd also add that including casualty parameters in the victory conditions is another weapon to make a realistic scenario. Make the objectives fairly easy but you can only win if it doesn't cost too much in lives (and, indeed, vehicles).

That thought was so profound I put it in bold. 

 

[MikeyD]

Another example of real world casualties exceeding CM would be the siege of Poznan, February 1945, which  saw seventeen thousand KIA and an uncounted number of wounded over approx a 4 weeks. (I threw together a scenario covering that for Fire and Rubble.)

In CM we're witnessing the 'tip of the spear'. We don't see the logistics tail stretching back a dozen miles where basically nobody dies. Its like the NFL. Statistically, a very low percentage of NFL company employees find themselves being tackled. The few NFL employees that we see on TV, though, experience it on a regular basis.

[landser]

1 hour ago, SimpleSimon said:

 Why that position?

I often find myself having an inner debate about these sorts of things. And something like the quote above is one. Due to a combination of the scale (both time and area), the AI and how things work in Combat Mission generally, there is no prospect of forcing a tactical withdrawal. A strong point in actual warfare is a key defensive position, and there will be times the enemy will hold at all costs, but in the main, even these sorts of positions will be abandoned when a penetration occurs elsewhere along the line and the troops on the strong point are threatened with envelopment. I suppose a scenario designer can trigger this sort of behavior if enemy troops reach a certain line on the map? But in general this sort of thing doesn't occur in Combat Mission. There's little in the way of tactical fluidity, of reinforcements sealing a penetration, of withdrawals or exploiting a breakthrough if it occurs other than perhaps where a designer foresaw it. So in Combat Mission we must assault that position, come hell or high water, because they're not going anywhere and it's a victory location after all. There is little to 'unhinge' a defensive position and force it's retreat.

I try to make it a point not to criticize a game for what it is not. In most cases it isn't fair. You don't buy a coupe and complain it's not a convertible after all. But at the end of my inner debates I find myself hoping that these are the areas where the next steps in Combat Mission's evolution occur. That we eventually see a computer opponent that is able to think on its feet, to react and exploit, or to save its hide or push for victory through prudent recognition of the ebb and flow of a tactical battlefield. I sometimes feel like we have all the tools aside from any sense we are playing a human opponent because the computer is so rigid in its conduct. And AI that could at least approximate this would go far in making Combat Mission's single player a much better experience than it already is.

in terms of the topic this sort of withdrawal could go a long way to preserving force strength, but then again it would simply push them toward the back edge of the map as there's no escape, where it would likely be even easier to rack up high kill counts, away from suitable terrain and prepared positions. Perhaps there could be a retreat off map option for the computer opponent, but maybe the game loses some appeal as it's not so compelling if you force the enemy in to a pell-mell race for the exits. It might come down to the player's expectations, as some would see this as a realistic reaction to a disintegrating defense and some would lament the loss of the drama or challenge they expect from a given scenario. "All I did was drive my tank platoon through the gap and ten minutes later the scenario ended".

I don't know what the right answer is, and what I want out of the game is not what the next man wants. Until it can be worked out though we'll continue to see casualty percentages far in excess of what would be tolerable to a battalion commander. Stand and die is not all that rare in history as well all know, but in Combat Mission there really isn't any alternative.