Why infantry combat in CMx2 is so different

[abneo3sierra]

Just want to say that there have been many battles fought with heavy losses. It is only relatively recently that casualty counts have been low. In the battle of the Bulge there were units on both sides that lost 70%+.

In CMSF, I have fought some battles with very high attrition,and have fought some with almost none,just finished the first battle in the German campaign with the loss of one vehicle and 2 WIA 0 KIA. When I first got the game, played that same battle and lost 10x that. We learn,and improve...and sometimes also it is just luck,as it sometimes is in RL.

[LongLeftFlank]

Don't go there with JasonC on any kind of WWII macro data. You may receive a riposte that feels akin to that grenade barrage Elvis' Army just ate.

On that topic, I just wanted to note that even given the point blank ranges in the AAR, grenade volume seems a little prolific, as was often the case in CMSF CQB up to the last patch -- it seems better now. That could use some ongoing tweaking.... perhaps relate it to the number of men in the squad? (i.e. more chucking opportunities)?

[abneo3sierra]

ha,thanks..but I have told him before that I will gladly trade macro data. At West Point we had a great many courses in military history,not to mention a family that served (on the other side) in that particular battle that I used as an example. WW1 and WW2 both saw incredibly heavy losses on both sides. Sure, if you count an entire division's strength,including those sitting safely far behind the lines, the losses may not seem too high a percentage, but in the combat battalions and units from Regiment on downward,very many were reduced to almost nothing. In Stalingrad in September's battles alone, the German 6. Army had 9 Battalions that were reduced to 15% strength or less...granted,that is over a month...but they generally were not in action 24/7 every day of the month, so those losses generally were several men here and there, and the rest concentrated around each of the 3 major pushes that month,each of which did not involve all of those battalions, and each of which only lasted as long as several CMSF battles. Granted,that was one of the worst examples, but the same thing happened in the Bulge battle in 44,where German and American casualties both were ,as I said, near 70% in more than a handful of the committed COMBAT units...again,not counting REMFs...who are not counted in CMSF,either.

In short, I just don't see how it is "unrealistic" to show that,without proper training and hard work, casualty rates really ARE incredibly high...this is the entire reason why the Army and Marines pay us as officers, to train to use every advantage, to keep the cost as low as possible. I would think it was an unrealistic game if someone with no knowledge could walk in and get through a battle with a handful of casualties...as long as you can learn,and improve,it is realistic...and in CMSF it has been awhile since I lost more than 5% KIA and WIA combined...so it is a question of tactics,as it should be,rather than a flawed game system.

[vincere]

Well thanks to Jason C for prompting an interesting side discussion. I probably would have tended to agree with him in an philosophical way about abstractions having some merit over trying to engineer a collection of variables as best as possible. Still see some merit of the idea; but boy does it work well in CMSF most times. I'm assuming that at today's tech level abstraction is stronger the higher the level, as in strategic games. And CMx2 still uses abstraction where prudent. Just wanted to say it's interesting to read the differing views- moon's clinched it in that I don't want pre-ordained outcomes.

Back on topic. I really notice the squads ability to fire at different targets. Tac AI works well at this and is better than manually targetting most times.

[Battlefront.com]

Our experience with tactical warfare in CMx1 was that there was a tipping point where an outcome based system became more difficult to maintain and had quicker diminishing returns than an engineered system. At that tipping point we were basically obligated to cease trying to improve the feature because from a technical standpoint it wasn't likely to succeed. Or at least succeed enough to justify the investment of time and the side effects (ripple effects) of making the code more complicated.

We found the opposite with engineered systems within CMx1. It might take longer to get it working, but long term it was easier to maintain and the results were more likely to be consistently acceptable. And when they weren't, the chances were fairly good that an engineered solution could be implemented. Obviously "cost:benefit" equation still applied here, but the cost was typically low enough that the benefit didn't have to be extraordinarily high. Put another way, we generally were able to eek out a slightly better end result or add a non-esscential behavior at a reasonable cost.

These experiences with CMx1 laid the groundwork for CMx2. Anything that seemed to be likely to benefit from an engineered solution was built that way from the start. Anything that appeared capable of working well with an outcome based solution got one.

After nearly 5 active years of experience with the CMx1 code and another 5 years with CMx2... we know, for sure, that we made the right choice. JasonC is welcome to disagree with this, of course. As soon as he shows us the game he made that employed outcome based coding we can evaluate who is more correct. Until then it is just academic theories compared to our 20+ years of practical experience.

Steve

[vincere]

I have a recurring thought about CM, (internal dialogue "damn I hope this doesn't come across as fan boy")

An engineered and scientific creation yes. But also a bloody creative peice of art. That really shows through in SF. Sooo many games (and other entertainment genres for that matter) show much promise but let down in the final analysis. You guys, with the help of the testers make at least hundreds of decisions about balance. Some of which are critical. And you obviously make the right calls on those most of the time.

Ok, fan boy mode off.

[JasonC]

First on Normandy supposedly being unlike the Leyte description because it was supposedly a case where both sides fought to the death - um, sorry, not remotely.

German losses in Normandy pre breakout or during, but excluding Falaise, are variously given as 85000 (Zetterling and a low figure) to around 115000 - with another 90000 lost in the encirclement fighting during the withdrawal, plus large PW hauls in Cherbourg and the channel ports, amounting to 200000 PWs all told. Which is a clear case of the battle ending in retreat and surrender on an operational scale, if ever there was one. Meanwhile we can estimate the company * days of engaged German infantry formations at something around 250 companies average for 50 days from D-Day to breakout (starting lower, some not being there for 50 days etc, to arrive at that figure). Which puts the losses in KIA and WIA per company*day around 7 to 10, slightly above the supposedly dull and inactive US Leyte figure. For comparison, US divisions averaged 25 combat losses per day in action for the whole campaign, and the heaviest losses in the period, divisions actively attacking daily in the push to St Lo, lost around 250 per day. From a force of 9 infantry battalions - some engineers, losses in other arms, etc, but most infantry. Which is 11 per company per day.

Any way it is sliced, then, a normal loss rate for an infantry company in reasonably heavy action, to heavy action, ran 5 to 11 per day. Call it 8 men with an error bar for the tactical stance and how heavy the action is. Naturally, this average was spread unevenly across its days in combat.

Stoex comments that gee this actually sounds like pretty heavy losses if you try to keep it up for any length of time, and he is exactly right. And they did keep it up for lengths of time on the scale of weeks to a month and a half - burning out in the process, and requiring refit off the line, replacements, etc. The level of losses an infantry formation can afford and sustain and remain combat effective, in other words, tops out around losses of 25 men in a heavy action, even if it is the only heavy action seen for several days (of rotated roles, better days, etc).

In the AAR we are all reading over in the sticky thread, that is about what the US commander expects to get in a minute of heavy action at the pointy end. Maybe in crossing one field - or at the utmost, in the course of taking one tiny body of woods with most of his force.

We are mashing units together vastly harder than they practically ever get mashed together in reality, in other words, with butcher bills that extrapolate out to the first day of the Somme or worse. Meaning the greatest outliers of bloodshed in the history of modern infantry warfare. We expect this as normal every time we engage. In the name of the supposed realism of a 1 to 1 engineered depiction of firepower.

Meanwhile in the after action reports of real units in Normandy, instead of the electronic ones in this game, you read things like one of the platoons of the company was caught by 81mm mortar fire crossing one field, and took 25 wounded - check on the potential lethality of the weaponry - and the sequel was? The entire battalion called it a day and broke the whole thing off as a disaster. The rest of the company evac'ed the wounded shambles and held the frontage, the neighboring companies went to ground, the plan that depended on the mortar'ed company fell apart, commanders bickered over what to do about it, etc.

Which is why that realistic 25 men lost in a few minutes of mortar fire was not compounded into 20 or 30 minutes that bad, all within one hour, until something like 70 to 80% of the battalion were casualties. Which we do see in electronic games, but didn't happen in reality - and assuredly was not the normal outcome of a tactical infantry firefight in Normandy.

In the Leyte fighting, the real AARs mention things like a platoon hit by all of 2 hidden Japanese machineguns losing 17 men in a matter of seconds, as an outlier example of what sometimes happened. Check. Meanwhile the Japanese were not inflicting 17 men per 4 of their shooters every 30 seconds or minute, but instead were losing 20 men per man they took out.

85% of the Japanese losses were to artillery fire. The American higher ups were exasperated with their men because the men had a very personal approach to the war - they wanted to live through it, the rogues. So when they got shot at, they hit the deck, tried to locate the shooters, and called for artillery fire on them. Occasionallly American junior officers or NCOs charged Japanese machinegun nests single handed and got medals for taking them out - frequently posthumous medals. The men had this crazy disregard for the shiny bits however and an unhealthy focus on the "posthumous" part, didn't take the hint, stayed down, and continued to call for artillery fire. They figured when their side in the war had every conceivable advantage and the overall strategic outcome was entirely certain, it was kinda pointless to go get killed for no reason, or just to make it go faster or something.

And somehow, engineering approaches to simulation as opposed to design for effect, never seem to get that sort of thing quite right. Lots of kudos to BTS for going way farther in that direction than most, certainly - men cower and men run in CM, 1 or 2. They don't stay pinned as deep or as long as in reality, no doubt, but at least the system is trying. It just is never able to simulate the range of such things, their variance, and their average impact, with anything like the fidelity it can bring to a 75mm AP shot booking downrange.

In my opinion nobody can or ever will. Which doesn't mean wargames will always get it wrong. But does mean design for effect starts off about 50 miles ahead of 1 to 1 engineering of each physical bit, in overall realism and predicted rates of loss.

[dan/california]

So what you are really saying is that most CM battles should be over in about the first 15 minutes because one side or the other would give up trying so hard after the first bad mistake. That is how the U.S. lost Vietnam, I am far less certain it is how we won WW2.

In any event you aren't arguing nearly as much for changes in game mechanics as you are scenario design.

That or you think games at this scale are pointless and we should all find another hobby.

[stikkypixie]

So what you are really saying is that most CM battles should be over in about the first 15 minutes because one side or the other would give up trying so hard after the first bad mistake. That is how the U.S. lost Vietnam, I am far less certain it is how we won WW2.

In any event you aren't arguing nearly as much for changes in game mechanics as you are scenario design.

That or you think games at this scale are pointless and we should all find another hobby.

I think he's saying in CM the soldiers are not reluctant enough to go into their death. It's far too easy for the armchair general to send them on suicide missions.

[dan/california]

You can probably get a lot of that behavior now if you set morale and motivation low enough.

[dan/california]

I do feel Steve's comments about how things usually work out at the NTC are relevant here as well. And there are times when the other just keeps coming. The relatively recent battle of Wanat is an example.

[stikkypixie]

You can probably get a lot of that behavior now if you set morale and motivation low enough.

The thing is, they will still *go*, they just won't stay long with low morale and motivation.

[Thin Red Line]

A good way to play more realistically regarding losses (withdraw or stop fighting if the nut is too hard to crack), is to play meta campaigns.

Because you know the next phase you've got what you're left with after this battle.

Just my two cents.

[Michael Emrys]

A good way to play more realistically regarding losses (withdraw or stop fighting if the nut is too hard to crack), is to play meta campaigns.

Because you know the next phase you've got what you're left with after this battle.

Just my two cents.

Or, for those who do not wish to play campaigns, increase the victory points for keeping your force intact. What would be wrong with that? (Serious question.)

Michael

[undead reindeer cavalry]

I am sure I'm not the only CMx1 player who quite frequently found myself looking at a unit and thinking "they are dead if they move, so they might as well fight it out where they are".

"reverse advance" would be / have been great!

The other rule of thumb is that the more consistent the fidelity is within the simulation, the more consistent the results will be.

have there been any significant improvements to the CMx2 engine what comes to this (consistency of fidelity) and CMBN? for example the fidely between fire and movement is quite stark in CMSF - are there any new major improvements, fixes or tweaks in CMBN?

There wasn't much point having individual soldiers simulated (even non-graphically) since each terrain tile was internally identical. Which in turn meant ballistics couldn't be simulated directly since there weren't 13 points of fire for a 13 man unit, but rather 1 point of fire. Since there is only one point of fire there's no reason to simulate individual weapons firing on their own. Instead a volley of fire, made by "adding up" the individual weapons, was the only way to go. With no terrain variations or soldiers to use them, why bother having direct simulation of cover and concealment? So on and so forth.

i am not 100% sure what you mean ("each terrain tile was internally identical" :confused::confused:), but the way i read it is that you say that one 20x20m tile in CMx1 would not contain any variation within it. to me that seems like a counterfactual statement, since there can be variations not only in both elevations and terrain types but also in LOS/LOF "rules" (LOS/LOF may change even when elevation or terrain type does not) within a single 20x20m tile.

If this had been CMx1 the terrain wouldn't have had any affect unless Jon's guys moved from one 20x20m tile to another which offered situationally better cover/concealment, or perhaps a significant elevation fold in the middle of the tile. Meaning, as Jon's soldiers started moving they would usually have been just as vulnerable with each meter moved.

but that's simply not the case in CMx1? :confused:

for example you can only see 26 meters into woods (combined, between tiles - e.g. if you are 10 meters into a wood tile and you are looking at woods across an open field, you can only see some 16 meters into it, or if you are in middle of woods tiles you can only see 26 meters around you) or how in urban combat you can only see half way into those heavy buildings (or not see opposite corner of a light building etc etc) and then you have those tiles with multiple internal terrain types (e.g. a single 20x20m tile can contain such different terrain types like "open", "scattered trees", "road", "crater", "trench").

in optimal conditions all it takes is less than one meter of movement in CMx1 to break LOS. in CMSF you can not make such high fidelity moves - the smallest move you can make is 8 meters (when you move non-diagonally, for diagonal movement it's of course more). does CMBN allow these kind of high-fidelity movements we have in CMx1 or are there any plans for improving the system towards that direction at some point in future?

Keeping in mind the two rules of thumb noted above, the CMx2 engine was based around the 8x8m Action Spot system instead of the old 20x20m Tile system.

well...

Hope that long read was interesting

it was. sorry for the CMx1 nitpicking, but i want to be sure i haven't misunderstood something.

[LongLeftFlank]

JasonC, in each of the examples you mention (Normandy, Leyte, and maybe Goose Green too), you seem to have one or more platoons -- perhaps a whole lot of them across a broad front -- probing forward to locate the enemy MLR. One hits an ambush and gets badly torn up. The commander quickly calls a halt to the other forces (or they halt themselves depending on FoW) and calls in the arty. Everyone seeks cover. The savaged unit extracts itself and its wounded as best it can.

After the bombardment lifts, either you (a) call off the attack ( resume the advance with a new set of platoons © direct fire on and/or assault the now located enemy positions. Maybe there's options (d)-(f) too, but I'll let you tell me what those are.

The assault is a VERY different animal from the probe. Over the top, boys! The safest spot is in his trenches; if you go to ground in the beaten zone you're dead meat, and all the veterans know it. The Foy attack in BoB is a good example. And if wave 1 fails, send wave 2, etc.

In WWII even a successful charge against a damaged enemy could result in terrible casualties, knots of unlucky men struck in twos and threes. Or worse, the poor kids get hung up on some obstacle and slaughtered in heaps WWI style -- plenty of examples of this nightmare in all armies, even the Americans.

Now would a given infantry unit be capable or performing that kind of desperate charge day after day? Not bloody likely, especially given the losses from doing it once. Not even sure WWI forces were capable of that in spite of the desires of their leaders -- I don't know enough to say.

In CMx2 terms, the very act of doing such a charge once would likely Exhaust a unit at least for the remaining duration of a scenario -- it could defend its gains and that's about it. Any further advance would have to be by another wave. Except perhaps for a handful of Wigram's "gutful men" who are willing to slog on and keep killing the enemy.

These types of actions lie buried within the "average casualties per division per day" statistics. But their intensity and lethality are quite believable, and documented. And tend to be the focus of CMSF scenarios.

Comments?

[Battlefront.com]

JasonC,

If your argument is that it is too easy to do things in CM than it is in real life, and that then produces more casualties from small arms (and other direct fire) than was common in WW2, then I agree with that position with certain caveats:

1. It has nothing to do with engineered or effect based designs. We could EASILY make our engineered soldiers not move when ordered to, run away and STAY away, not fire weapons, etc. Piece of cake to do that. Not very marketable, but easy to do none-the-less.

2. Combat Mission simulates the exceptional combat situations. When you look at the macro view of the entire war there might have been... I dunno, a couple thousand encounters over the entire Western Front that would make an interesting CM battle. So Combat Mission is obviously not about simulating the average battle. Therefore, comparing it's results to the average battle's results is an apples to oranges comparison.

And since there's no commercial wargame I know of that simulates the sorts of things you're talking about better than Combat Mission, I don't really know how that relates to engineered vs. effect based designs. In fact, I'd say that actually reinforces my points about why going engineered was the best way to go for Combat Mission.

Steve

[Battlefront.com]

In any event you aren't arguing nearly as much for changes in game mechanics as you are scenario design.

Correct. He is also arguing for something which isn't really a game but more-or-less a spreadsheet full of battle results broken into rather large and ambiguous chunks of data.

That or you think games at this scale are pointless and we should all find another hobby.

That is the problem with JasonC's position in a nutshell. The only way to have a truly realistic tactical wargame experience is to sign up for an armed service and hope to get into combat. Not so much a game as real life, but that's the crux of the problem. Real war isn't fun, so the closer a game comes to being truly realistic the less chance it has of being fun. Finding the balance is where good game developers are separated out from the bad.

have there been any significant improvements to the CMx2 engine what comes to this (consistency of fidelity) and CMBN? for example the fidely between fire and movement is quite stark in CMSF - are there any new major improvements, fixes or tweaks in CMBN?

Before I begin answering your posts, note that your questions don't have much to do with the topics being discussed. Either the original one, which is why infantry modeling is better in CMx2 than CMx1, nor the argument that "design for effect" is better than "engineered".

I'm not sure I know what you mean. "Fidelity" is the degree of detail found in a particular part of the game. "Consistency" is how frequently the results are within acceptable bounds compared to being outside acceptable bounds. What do you think in CM:SF isn't behaving that way? We certainly haven't added any more factors, equations, or other changes to the underlying infantry modeling. TacAI tweaks, adjustments for using terrain, etc... sure, but those are different things that what I've been talking about.

i am not 100% sure what you mean ("each terrain tile was internally identical" :confused::confused:), but the way i read it is that you say that one 20x20m tile in CMx1 would not contain any variation within it.

Correct, there was no variation of the terrain within a 20x20m tile. Elevation changes, and their effects, are a different aspect. If you wish to consider this an exception to "no difference" that's fine with me. It doesn't change what I wrote.

for example you can only see 26 meters into woods (combined, between tiles - e.g. if you are 10 meters into a wood tile and you are looking at woods across an open field, you can only see some 16 meters into it, or if you are in middle of woods tiles you can only see 26 meters around you) or how in urban combat you can only see half way into those heavy buildings (or not see opposite corner of a light building etc etc)

This is a good example of the difference between engineered and effect designs. The reduction of LOS in CMx1 was an arbitrary limitation "designed for effect" and not one that was engineered. CMx2 also uses this same basic concepts because it's simply impossible to account for everything in the real world directly within the game. Therefore, this is one of the areas where CMx2 retains some design for effects features.

The difference between the two game systems is that CMx2 is doing this in smaller increments (8x8m), with a greater variety of variations within each "tile", and some elements within that are directly simulated. Boulders, trees, haystacks, walls, doors, windows, etc. are all directly simulated now. CMx1 didn't directly simulate any of that stuff. Heck, it didn't even have most of it at all. So CMx1 was pretty much purely designed for effect, CMx2 leans more towards engineered and has a bigger "pallet" to work with.

and then you have those tiles with multiple internal terrain types (e.g. a single 20x20m tile can contain such different terrain types like "open", "scattered trees", "road", "crater", "trench").

CMx1 Tiles could contain only 2 different types of basic terrain:

1. Ground

2. Something tall (vegetations, buildings, bridges, walls, bocage, and probably some other things I'm forgetting)

Roads were a 3rd possibility that was allowed for only very specific Tile types. CMx2 not only has roughly 4 tiles for every 1 of CMx1, but it also allows a vastly large variety of combinations of things within a single Action Spot (Tile). And as noted above, the features are (for the most part) directly simulated.

Trenches, craters, and foxholes were not terrain types. They were not treated as such within the game system either. They were simple combat modifiers applied to the soldiers who utilized them. They were also flagged to the TacAI as being good places to go.

Interestingly, in CM:SF/CM:A foxholes, craters, and trenches were directly simulated (engineered) into the terrain. This worked fine from a mechanics standpoint, but not so well from a gameplay standpoint. So CM:BN, and future games, now have a hybrid system where these things are directly simulated in 3D, but their effects are modified abstractly to account for the fact that they are no longer literally portrayed.

in optimal conditions all it takes is less than one meter of movement in CMx1 to break LOS. in CMSF you can not make such high fidelity moves - the smallest move you can make is 8 meters (when you move non-diagonally, for diagonal movement it's of course more).

Not quite correct. While it is true you can only order your guys to move in 8m increments, your soldiers move in sub-meter increments. When you order a Team to move some will move a full 8m, some will move only a meter or two. Depends on the circumstances.

LOS/LOF can also be broken in CMx2 in increments less than 1m. The difference is that in CMx2 the TacAI handles incremental movements within an Action Spot. You yourself can not directly specify to move 1m over here or 1m over there.

does CMBN allow these kind of high-fidelity movements we have in CMx1 or are there any plans for improving the system towards that direction at some point in future?

No, there is no plans on changing the fidelity because it is unnecessary.

it was. sorry for the CMx1 nitpicking, but i want to be sure i haven't misunderstood something.

I think what you've misunderstood is that CMx1 was largely abstracted at all levels. It was internally consistent. CMx2 is largely unabstracted at all levels. It is also internally consistent. CMx1 was low fidelity, CMx2 is higher fidelity. In some cases there is no effective difference in their outcomes, but in others (and that is why I started this thread) there are big differences.

As others have already noted, the CMx2 infantry modeling feels much more realistic than CMx1. The reason for that is we made CMx2's infantry modeling inherently an engineered system instead of an abstract, design for effect type system.

Steve

[Battlefront.com]

In WWII even a successful charge against a damaged enemy could result in terrible casualties, knots of unlucky men struck in twos and threes.

Yes, absolutely. According to JasonC the first hand account I read of an entire platoon being wiped out, with one exception (the guy who wrote the story), is a work of fiction because statistically this couldn't have possibly happened.

The story, BTW, was cited as an example of why the US Army had to do a better job with their junior leadership training. A fresh LT took control of a platoon and walked right into a trap. The SGTs tried to get the LT to see that there was a big problem with his approach, but he said he knew best and everybody but one of the Squad Leaders paid for that arrogance with their lives. If the platoon was near full strength that was roughly 40 men killed within a very short span of time (timeframe wasn't exactly specified, but I gathered it was under an hour or two).

These types of actions lie buried within the "average casualties per division per day" statistics. But their intensity and lethality are quite believable, and documented. And tend to be the focus of CMSF scenarios.

Correct. Another example is the casualties suffered by the Marines when they went to take Nasiriya during the initial phase of OIF:

Total US fatalities for OIF until Baghdad fell = 139

Total US fatalities for 6 days in Nasiriya = 29

Total US fatalities for Charlie Company on 23rd March = 18

Statistically this battle was roughly 40% of all US casualties in the whole operation. And of that, more than 50% of them happened in one day. Wounded for the battle for Nasiriya was roughly twice the KIAs.

As a percentage of total forces involved:

Total US fatalities for OIF until Bagdad fell = 1% (rounded up)

Total US fatalities for 6 days in Nasiriya = 3% (guess based on roughly 2 battalions directly engaged)

Total US fatalities for Charlie Company on 23rd March = 10% (estimate based on 200 men directly engaged)

So the battle itself wasn't that intense for the Brigade sized Task Force Tarawa in terms of suffering casualties, but for Charlie Company the story was quite different.

Steve

[LongLeftFlank]

Yes, absolutely. According to JasonC the first hand account I read of an entire platoon being wiped out, with one exception (the guy who wrote the story), is a work of fiction because statistically this couldn't have possibly happened.

Well, I wouldn't wish caricature Jason's core argument about lethality of fire here, because I don't think he's denying these incidents happen at all. I think he's just asserting that -- fanatical banzai charges a la Bloody Nose Ridge or Elsenborn aside -- the vast majority of the bloodshed takes place in the opening moments of a firefight.

There was a time earlier in CMSF's development when infantry would go to ground in place when fired on and just keep dying and dying and dying at a steady rate for as long as you kept hosing them down. Even in decent cover. So all you had to do was just keep shooting; no real tactics required.

Whereas his view is that within scant seconds of a firefight, the surviving targets have wormed their way into solid cover even in the most unlikely places, and they will not continue taking casualties at anything like the same rate unless either (a) they expose themselves (e.g. resume fighting) or ( you can bring weapons to bear on them from a different, unshielded angle, or chuck in HE or sumfink.

So in game terms, he is arguing, I think, for a significantly enhanced cover value (against bullets anyway) for units that are Pinned, Cowering or Hiding, as opposed to merely prone (and still fighting/Spotting). And I think he has a point, although as I've said there are other workarounds available.

[abneo3sierra]

Added to what Steve has said, is still that the combat losses are a much higher percentage,when looking at them as a part of the force actually fighting,than when looking at them as just a percentage of the forces in theater, or in a division,etc...only small percentage of personnel in divisions are actually combat troops, so a division can appear to suffer small % of casualties,but of the combat companies and platoons who are actually out there slugging it out,the losses run much higher. That said, still, even divison-wide, there have been as recently as WW2 divisions that just melted away after some weeks of combat. And in the end, it is still about tactics, it is quite possible in CMSF to employ tactics that save your men and still win...what sometimes seems to happen is that people prefer the action and intensity of a more drawn out battle,and most people playing red, want a chance to win,and most people playing blue, want to at least be challenged. This leads to situations where the scenario designers sometimes give red a lot more chances than they would have had(in ww2) or would have(in the modern syria) to succeed. If one is going to play a game, it should be fun, if one is trying to recreate history...just read the books, but remember that some battles even then, swung on luck,and could easily have gone the other way, or could easily have had far more, or far fewer,casualties than they did have,with only a few things altered in each case.

[abneo3sierra]

Well, I wouldn't wish caricature Jason's core argument about lethality of fire here, because I don't think he's denying these incidents happen at all. I think he's just asserting that -- fanatical banzai charges a la Bloody Nose Ridge or Elsenborn aside -- the vast majority of the bloodshed takes place in the opening moments of a firefight.

There was a time earlier in CMSF's development when infantry would go to ground in place when fired on and just keep dying and dying and dying at a steady rate for as long as you kept hosing them down. Even in decent cover. So all you had to do was just keep shooting; no real tactics required.

Whereas his view is that within scant seconds of a firefight, the surviving targets have wormed their way into solid cover even in the most unlikely places, and they will not continue taking casualties at anything like the same rate unless either (a) they expose themselves (e.g. resume fighting) or ( you can bring weapons to bear on them from a different, unshielded angle, or chuck in HE or sumfink.

So in game terms, he is arguing, I think, for a significantly enhanced cover value (against bullets anyway) for units that are Pinned, Cowering or Hiding, as opposed to merely prone (and still fighting/Spotting). And I think he has a point, although as I've said there are other workarounds available.

It is correct that the initial moments of a fight cause the most casualties...it is for this reason that the attacking force will try to put a large amount of firepower at the decisive point...however, Jason is also one who before had preferred not to employ large forces in scenarios..so in reality,where if some companies would lose too many men,and pull off line for other companies behind them,in CMSF,especially if you are trying to play only a small scenario, you do not have follow on forces.This leaves the CMSF "commander" in a position where he has to keep on, in a situation where a real commander may pull back and insert fresh reserves,etc.

One of my relatives lost 78 men of the 112 that he began with at Ardennes. In that case, they were not allowed to pull off the line,but went over to a defensive/reserve position after 8 days. Most of the losses happened on their 4th day,where in that day they lost 32, or 28% of their starting strength. But that still left 46 losses divided up in the other 7 days, nearly 7 a day, or over 6% of their starting strength each day(of course, as the days passed, the % of losses of their remaining strength was actually much higher.) This was also the case with many German volksgrenadier companies, as well as on the American side.

[Battlefront.com]

Sure, something like 70%+ casualties in the Western Theater were caused by artillery fragments. Probably the majority of them during battles which would put the average CM player to sleep. But I'm still scratching my head as to what this has to do with anything in this thread until he brought it up.

Yes, it is true that soldiers love to find cover/concealment. But the problem you pointed out about deciding whether to favor that or firing is a TacAI issue. Was for CMx1 as well as all other wargames out there (whether they have something directly akin to TacAI or not). Again, it has nothing to do with inherent engineering, nor of outcomes. Flaws in TacAI produce flawed results. Unfortunately, getting a computer to be "smart" like a single Human is hard enough. Getting it to be "smart" in a group way is even worse.

An effect based solution would be to simply increase cover or decrease firepower. This gets into tail chasing, which I illustrated earlier with a tank example. Increase that cover and all of a sudden a barrage of 155mm rounds doesn't do nearly the sort of damage it should. It's a really tough problem to tackle because it is inherently difficult to determine when a soldier should risk shooting and when he shouldn't. Since even outcome based systems have to make this determination (or face other problems) there's never going to be a perfect solution to all situations all the times.

That being said, I think CM:BN does a very good job of this. Today I ordered a Squad to Assault a house. The first Team to run towards the house got cut down (in my defense I was only playing to test something... I swear ) and the remaining two Teams decided to ignore my Assault Command and instead retained their cover. I eventually got them into the house where they alternated between trading shots with the enemy and taking shots. They tried to leave the house at one point but I managed to countermand the order (playing RT). They dutifully continued to trade shots and take cover. Er, until they were all casualties.

Seemed perfectly reasonable result to me. I made a very bad tactical decision, countermanded a sensible TacAI decision, and then got wiped out. Would this ever happen in real life? Probably not. But what is the game supposed to do about this... not allow me to move the rest of the Squad into that house? How would it even know not to do this?

Nope, the weakest realism link in any wargame is the player. The strongest reason to play the game is for players to figure that out. No way to compromise one without the other.

Steve

[abneo3sierra]

......

Nope, the weakest realism link in any wargame is the player. The strongest reason to play the game is for players to figure that out. No way to compromise one without the other.

Steve

I have not of course tried the Normandy mission, but I 100% agree with this comment as far as CMSF....it is definitely possible to suffer the low impact casualties and win, and of course, it is quite possible to botch everything and have many letters to write to parents of your pixeltruppen. Exactly how it should be.

And I apologize for going off topic about the % of casualties, just trying to stress a point that having many infantry as casualties, is not as unrealistic,expecially in WW2, as it seems is being claimed.

[SlapHappy]

I think the one thing left out of the equation as far as the limitations of casualties is time and scale. Remotely competent commanders would withdraw in the face of perceived overwhelming pressure or the threat of high sustained casualties. I think one could safely argue that a commander may choose to simply withdraw forces in the face of contact in many of the scenarios which might be represented in the game. Especially when considering other factors which he may be aware of about what lies behind his immediate front (or the enemies). Trading space for time in real life makes sense, but doesn't make for much of a wargaming experience though, does it? CM missions take place in a vacuum, though, and the idea that some resistance from both combatants is assumed even though such contact might never actually occur in REAL LIFE.

I think it is those situations when both opponents chose to take a firm defensive or offensive position is where these much higher casualties DID take place.